Metropolitan Organization: the St. Louis Case. Commission Report

Home , John Ashcroft, Metropolitan county

Current Members of the Advisory Commission on Intergovernmental Relations (August 1988)

Private Citizens James S. Dwight, Jr., Arlington, Virginia Daniel J. Elazar, Philadelphia, Pennsylvania Robert B. Hawkins, Jr., Chairman, San Francisco, California

Members of the U.S. Senate David Durenberger, Minnesota Carl Levin, Michigan James R. Sasser, Tennessee

Members of the U.S. House of Representatives Sander Levin, Michigan Jim Ross Lightfoot, Iowa Ted Weiss, New York

Officers of the Executive Branch, U.S. Government Andrew H. Card, Deputy Assistant to the President for Intergovernmental Affairs Ann McLaughlin, Secretary of Labor Vacancy

Governors John Ashcroft, Missouri John H. Sununu, Vice Chairmun, New Hampshire Vacancy Vacancy

Mayors Donald M. Fraser, Minneapolis, Minnesota William H. Hudnut, 111, Indianapolis, Indiana Robert M. Isaac, Colorado Springs, Colorado Vacancy

Members of State Legislatures John T. Bragg, Deputy Speaker, Tennessee House of Representatives Ross 0.Doyen, Kansas Senate David E. Nething, North Dakota Senate

Elected County Officials Philip B. Elfstrom, Kane County, Illinois, County Commission Harvey Ruvin, Metropolitan Dade County, Florida, County Commission Sandra Smoley, Sacramento County, California, Board of Supervisors

- , A Commission Report

Metropolitan Organization:

I. I.I Louis Case

ADVISORY COMMISSION ON INTERGOVERNMENTAL RELATIONS

Washington, DC 20575 September 1988 M- 158

Preface

This report marks the first publication in a series izational "overlays" can knit jurisdictions together at of case studies being undertaken by ACIR in an ef- key points. Economies of large scale can be cap- fort to learn more about how complex metropolitan tured without sacrificing the economies of small areas are organized and governed in our federal sys- scale. None of these good things, of course, come tem. This study focuses on the metropolitan area of about without paying a price-in this case, the costs St. Louis, Missouri, and, therein on St. Louis consist of the time and effort needed to bring about County. This area and county were selected for multijurisdictional coordination and development. study precisely because of a governmental structure The best evidence that these activities are worth the that is among the most complex in metropolitan price, however, is that citizens and officials voluntar- America. St. Louis County, organized separately ily make the effort. The costs, however great, are from the City of St. Louis, now contains 91 munici- not prohibitive. A system that keeps the cost of im- palities, 23 school districts, and 25 fire protection proving the human condition low enough that people districts, and continues to grow and develop. choose to make the effort is a system that is dynamic The research presented here follows a different and workable over the long haul-a system that will approach from typical metropolitan area studies. In- tend to make things better, not make things worse, stead of adopting a reform perspective, the ap- over time. proach in this study explicitly seeks to describe and This report is likely to be controversial because understand the organizational dynamics of a com- it challenges many of the traditional approaches to plex set of local jurisdictions in a metropolitan metropolitan reform, which have been championed county. No effort is made to measure St. Louis for many decades by those who sought to improve County against the standard of a single metropolitan local and metropolitan governance. ACIR, of government. Instead, a multiplicity of local govern- course, shares in this enduring quest to improve lo- ments is viewed as a potentially productive "local cal governance. One must, however, seek continually to learn and be willing to rethink long accepted public economy," in the terms of ACIR's recent re- points of view if necessary. ACIR offers this report port The organization of Local Public Economies in that spirit-seeking to learn, through discussion (A-109), and is judged according to functional cri- and debate, as well as through systematic analysis teria. No such system is perfect, and this report has and inquiry, how to improve the ways in which we some criticisms to offer and improvements to sug- choose to govern ourselves. gest. The focus of the report, however, is on the les- The American tradition of self-governance is, as sons to be learned from the creative energies of a Alexander Hamilton put it in Federalist 1, a process productive metropolitan community of communities. that joins "reflection and choice." The American Among these lessons are some of the positive people remain dedicated to a tradition of self-gov- aspects of a jurisdictional pattern long characterized ernance and local liberty that is much older even as "fragmented. " These positive aspects include than the U.S. Constitution. No less dedication is ex- both the potential for and realization of more acces- pected from all of us who work in intergovernmental sible representation of local citizens, more economi- relations to continue linking reflection with choice in cal patterns of accountability, and greater local re- the practice of American local democracy. sponsibility for community problems. The report also notes that jurisdictional fragmentation need not Robert B. Hawkins, Jr. lead to functional fragmentation. A variety of organ- Chairman Acknowledgments

This report is the product of collaborative re- drafts of this report: Donald E. Clark, St. Louis search by Roger B. Parks, Director of the Center for County Director of Planning; Professor James Bras- Policy and Public Management, School of Public field, Webster University; Professor Donald Phares, and Environmental Affairs, Indiana University, University of Missouri, St. Louis; Richard Patton, Bloomington, who was principal investigator, and Center for Metropolitan Studies, University of Mis- Ronald J. Oakerson, ACIR Senior Analyst, who was souri, St. Louis; and Professor Donald Elliott, project director. Other assistance was provided by Southern Illinois University at Edwardsville. Their John Shannon, former Executive Director, and valuable comments and criticisms improved this re- Lawrence A. Hunter, former Director of Research, port in many ways. during the early phases of this project. The report Many other public officials, private citizens, and was edited by Joan Casey. Secretarial assistance was scholars in the St. Louis area and elsewhere were supplied by Anita McPhaul and Lori O'Bier. kind enough to share their insights and, frequently, Authorship of the report was shared principally data and analyses bearing upon St. Louis, in the between Parks and Oakerson. Parks had primary recourse of the research. Their contributions came in sponsibility for Chapters Four, Five, and Eight, and the form of interviews, data and analysis reports, Oakerson, for Chapters Three and Six. Chapter and detailed critiques of working drafts of this re- Seven was the joint work of Oakerson and Henry A. port. These contributors are listed on the following Bell at the Workshop in Political Theory and Policy pages with their affiliation. To each, we offer a pro- Analysis, Indiana University, Bloomington. In addi- found "thank you!" We are also very grateful to two tion to Parks, Oakerson, and Bell, the field research groups of persons not listed: individuals in 45 mu- team included Jan Leighley and Nanette Sharp. nicipalities who returned a mailed questionnaire on Preliminary plans for this study were reviewed street services and trustees in 53 residential subdivi- by an Academic Advisory Board consisting of the sions who responded either by telephone or by mail following persons: Louis de Alessi, Robert Bish, to a questionnaire on private streets. James Ferris, Bryan Jones, Ted Kolderie, Daniel Acknowledgments in a study this large always Mandelker, Stephen Mehay, Elinor Ostrom, Paul entail the risk of overlooking a significant contribu- Peterson, Derek Shearer, and Richard Wagner. tor. To anyone we have neglected to mention, Initial findings were reviewed at a critics' session on please accept our apology and thanks. Being ac- April 22, 1987, attended by Ann Cole, William Col- knowledged as a contributor does not imply agree- man, Edwin Connerly, Lawrence A. Hunter, Susan ment with the data, analysis, and interpretations Lauffer, Michael Libonati, Dolores Martin, James presented in this report. Nor are the contributors L. Martin, Richard Moore, Steve Moore, William and critical reviewers in any way responsible for re- Niskanen, Doug Peterson, Mark Schneider, John maining errors. That responsibility belongs to the Thomas, and Louise White. ACIR and its staff. ACIR is very appreciative of the time, effort, and expertise contributed by five individuals in the John Kincaid St. Louis area who critically reviewed one or more Executive Director Contributors to the Research

Stephen Ables, St. Louis County Municipal League Charles Dooley, Mayor, Northwoods E. W. Abram, Mayor, Bridgeton Leatrice Dowd, Mayor, Pagedale Gene Anthony, Trustee Chairman, Marlborough George Appel, Mayor, Bella Villa Donald Elliott, Professor, Southern Illinois University at Edwardsville Robert Elsea, Executive Director, Cooperating Donald Bauer, Mayor, Manchester School Districts of the St. Louis Suburban Area William Bell, Superintendent, Steve Elsner, Mayor, Oakland Jennings School District Harold Evangelists, Mayor, Black Jack Edwin Benton, Superintendent, Riverview Gardens School District Jerome Feldman, City Manager, Olivette Clyde Blakey. Mayor, Velda Village James Ferrel, Administrative Assistant to the Norman Brust, Superintendent, County Executive, St. Louis County Affton School District Tim Fischesser, Executive Director, Louis Bryant, East-West Gateway Normandy Municipal Council Coordinating Council Larry Fitzgerald, Mayor, Maryland Heights Jeffrey Buck, Mayor, Pasadena Hills James Brasfield, Professor, Webster University Charles Grimm, Mayor, Ferguson Leonard Burns, Superintendent, Roger Grow, St. Louis County Parkway School District Department of Planning Pat Hambrough, Mayor, Normandy Donald E. Clark, Director of Planning, Frank Hamshire, Legal Assistant to the Mayor, St. Louis County City of St. Louis Roger Clough , Superintendent, Edsel C. Hankins, Assistant to the Mayor, Ladue Pattonville School District Douglas Harms, City Administrator, Des Peres George Coates, Mayor, Beverly Hills Suzanne Hart, Executive Director, St. Louis Terril Cook, Association of Neighborhood Councils, Commission on Crime and Law Enforcement Chesterfield Area James Hennessey, Director of Administration, St. Louis County Police James Damos, Chief of Police, University City Michael Herring, City Administrator, Ballwin Richard Daykin, Director, St. Louis County Colleen Hilbert, Trustee Chairman, Pasadena Park Department of Highways and Traffic Francis Huss, Assistant Superintendent, John De Arman, Superintendent, Hazelwood School District Ritenour School District Thomas Jenkins, Mayor, Wellston Julie DeRienzo, private citizen, St. Louis County Herbert Jones, Mayor, Kirkwood Robert Dick, Fire Chief. Dee A. Joyner, Executive Director, Black Jack Fire Protection District St. Louis County Economic Council Allan B. Dieckgraeffe, City Engineer, University City Daniel Keck. Superintendent, Harold Dodge, Superintendent, Ferguson-Florissant School District University City School District Patricia Killoren, Mayor, Crestwood Charles Kindelberger, Director of Planning, Douglas Palmer, Mayor, Hazelwood St. Louis Community Development Agency Richard Patton, Center for Metropolitan Studies, Carol Kohfeld, Professor, University of Missouri, University of Missouri, St. Louis St. Louis Donald Phares, Professor, University of Missouri William J. Kramer, Fire Chief, Crestwood St. Louis Bill Kuehling, Assistant to the Mayor, John Pickins, Chief of Police, Pinelawn City of St. Louis Thomas Reagan, St. Louis County Ronald Lober, Mayor, Moline Acres Department of Revenue Dan Lowery, Mayor, Shrewsbury Natalie Rullkoetter, Executive Director, St. Louis County Municipal League Eileen McCartney, Mayor, Cool Valley Gene McNary, St. Louis County Executive Samuel Scarnato, Superintendent, John Mehan, Assistant City Manager, Special School District of St. Louis County Webster Groves Harold Schuchmann, Mayor, Town and Country Dan Meyer, Assistant Executive Director, Glenn Sheffield, Mayor, Webster Groves Confluence St. Louis Ronald Stodghill, Superintendent, Don R. Mitchell, Director, Audiovisual Education Wellston School District Department, Cooperating School Districts of the St. Louis Suburban Area William Tharp, Mayor, Jennings Linda Moore, Assistant to the Mayor, Wellston Gerald Troester, Executive Director, Joseph Morgan, Mayor, Fenton Cooperating School Districts of the Frank Munsch, Mayor, Overland St. Louis Suburban Area Dottie Niederkorn, Planning Supervisor, Susan Uchitelle, Director, St. Louis Public Schools Voluntary Interdistrict Coordinating Council Tom Utterback, Attorney, Elinor Ostrom, Professor, Workshop in Chesterfield Incorporators Political Theory and Policy Analysis, Indiana University, Bloomington Peggy Vickroy, Mayor, Creve Coeur Vincent Ostrom, Professor, Workshop in Political Theory and Policy Analysis, Thomas Ward, Trustee Chairman, Peerless Park Indiana University, Bloomington Thomas Wehrle, St. Louis County Counselor James Owen, Professor, Indiana University, Gordon P. Whitaker, Professor, Fort Wayne University of North Carolina at Chapel Hill Contents

Chapter One Metropolitan Organization: The Challenge of Understanding ...... Introduction ...... The Critique of Fragmentation ...... TraditionalTheory ...... Difficulties with the Traditional View ...... Jurisdictional Fragmentation. Functional Arrangements ...... A Theory of How Fragmentation Works ...... Research Questions ...... Evaluative Criteria ...... The St . Louis Case Study ...... Methodology: Distinguishing Provision and Production ...... Metropolitan Governance ...... Conclusion and Prospectus ...... Chapter Two St. Louis: A Variety of Jurisdictions ...... Introduction: City-County History and the Growth of Suburbs ...... Basic Political Geography ...... Overlapping Jurisdictions and Special Purpose Governments ...... Multiorganizational Arrangements ...... Joint Administrative Arrangements ...... Local Government Associations ...... Patterns of Political Representation ...... Fiscal Patterns ...... Revenue Sources ...... Patterns of Revenue Variation ...... Governmental Reform Movements ...... Conclusion ...... Chapter Three Metropolitan Governance in St. Louis: State Rules. Local Arrangements ...... Introduction ...... Functional Governance Arrangements ...... A Local Government Constitution ...... "Local Laws" and the Local Delegation to the State Legislature ...... Multijurisdictional Forums for Metropolitan Governance ...... A Typology of Rules ...... Rules of Association: Municipal Incorporation. County Home Rule. and District Formation ...... 33 City-County Separation ...... 34 Municipal Incorporation ...... 34 Special Purpose Districts ...... 35 County Government ...... 36 Contractual Rules: Interjurisdictional and Public-private Agreements ...... 37 Boundary Change Rules ...... 38 Consolidation ...... 38 Annexation ...... 38 Fiscal Rules ...... 39 The Functions of Fiscal Rules ...... 40 Tax Rules ...... 41 The Sales Tax Controversy ...... 43 Debt Rules ...... 45 Recent Annexation and Incorporation Activity ...... 45 Annexation Activity ...... 46 Incorporation Activity ...... 47 Conclusion ...... 48 Chapter Four Police Services ...... 55 Introduction ...... 55 Arranging for the Production of Police Services ...... 55 Police Contracting ...... 56 Patterns of Police Production ...... 57 Department Size and Costs Per Capita ...... 58 Service Components ...... 60 Production Efficiency: A Comparative View ...... 61 Cooperative Production Arrangements ...... 61 Computerized Call Routing ...... 62 Regional Justice Information System ...... 62 Mutual Aid among Police Agencies ...... 62 The Major Case Squad ...... 63 Police Performance ...... 63 Conclusion ...... 65 Chapter Five Fire Services ...... 69 Introduction ...... 69 Fire Service Provision and Production ...... 69 Economies of Scale? ...... 71 Coordinated Fire Protection ...... 72 Fire and Emergency Service Dispatch ...... 73 Fire Service Conflict ...... 74 Fire Service Performance ...... 75 Conclusion ...... 75 Chapter Six Streets and Street Services ...... 81 Introduction ...... 81 Thinking about Streets ...... 81 Development ...... 83 Arterial Streets ...... 83 Residential Streets ...... 84 Maintenance ...... Arterial Streets ...... Residential Streets ...... Reconstruction ...... Regulation of Use ...... Arterial Streets ...... Residential Streets ...... Basic Patterns of Street Provision ...... Arranging for Production ...... Conclusion ...... Chapter Seven Elementary and Secondary Education ...... Introduction ...... Provision Arrangements ...... School District Development in St . Louis County ...... The Special School District of St . Louis County ...... The St . Louis City School District ...... Private Alternatives to Public Provision ...... School District Organization ...... Patterns of Provision ...... Variation in Local Provision ...... The State Role in Provision ...... Fiscal Disparity-How Big a Problem? ...... Patterns of Production ...... Variation in Production ...... Variation in Student Achievement ...... Joint Production Arrangements ...... Special Education and Vocational-Technical Training ...... Indirect Services ...... Municipal and Community Relationships ...... Bilateral Arrangements between School Districts ...... Arrangements for Public School Desegregation ...... The Intracity Plan ...... The Interdistrict Plan ...... Conclusion ...... Fiscal Relationships ...... Patterns of Performance ...... Indirect Services and Special Education ...... Chapter Eight The Political Economy of Metropolitan St. Louis ...... Introduction ...... Economies of Scale ...... Fiscal Variation and Fiscal Equity ...... Patterned Inequalities? ...... Residential Tax Burden ...... The Case of School Districts ...... The Sales Tax Controversy Revisited ...... Reducing Fiscal Disparities ...... Economic Development ...... A Comparative Perspective on the St . Louis Development Experience ...... Economic Development and Fragmentation ...... 136 Conclusion ...... 137 Chapter Nine Functional Dimensions of Metropolitan Organization ...... Introduction ...... Functional Dimensions: How Fragmentation Works ...... A Local Government Constitution ...... ~ultijurisdicti&nalForums for Discussion and Negotiation ...... A Large Investment in Representation ...... Citizen Mayors ...... The Use of Overlapping Jurisdictions and Special Districts ...... Separation of Provision and Production ...... Coordinated and Joint Service Production ...... Public Entrepreneurship ...... Functional Evaluation ...... Self-Determination and Citizen Choice ...... Representation and Accountability ...... Coordination ...... Competition ...... Metropolitan Problem Solving ...... Fiscal Evaluation ...... Economies of Scale ...... Equity ...... Others' Evaluations ...... Citizens' Evaluations ...... Current Reform Activity ...... Conclusion ......

Figures and Tables Figure 2.1 Municipalities in St . Louis County ...... 16 Figure 2.2 Distribution of Average Household Income in County Municipalities ...... 19 Table 2.1 Distribution of Race. Poverty. and Income in St . Louis ...... 18 Table 2.2 General Purpose Local Governments in St . Louis ...... 18 Table 2.3 Municipalities in St . Louis County ...... 19 Table 2.4 Organization of Fire Protection in St . Louis ...... 20 Table 2.5 Revenues of General Purpose Governments and Fire Districts-1985 ...... 24 Table 2.6 Revenues of Municipal Governments and Fire Districts in St . Louis County-1985 . . 24 Table 2.7 Revenue Sources in County Municipalities- 198 5 ...... 25 Table 2.8 Revenue Variations in County Municipalities-1985 ...... 27 Table 4.1 Provision of Policing in St . Louis ...... 56 Table 4.2 Police Service Producers in St . Louis ...... 57 Table 4.3 Number of Full-Time Police Officers and Police Expenditures per 1. 000 Population ...... 58 Table 4.4 Regression Coefficients for Police Expenditures in County Municipalities-1985 ..... 59 Table 4.5 Assessments of and Experiences with Local Police ...... 64 Figure 5.1 Fire Districts and Fire Departments in St . Louis County ...... 70 Table 5.1 Fire Services in St . Louis ...... 71 Table 5.2 Fire Service Ratios in St . Louis ...... 71 Table 5.3 Size Effects in Fire Services: County Municipalities and Fire Protection Districts- 1985 ...... Appendix Table 5.1 Fire Protection Data for St; Louis ...... Table 6.1 Percentage of Subdivisions that Provide Common Street Services ...... Table 6.2 Subdivisions' Evaluation of Street Conditions ...... Table 6.3 Relation between Municipal Population Size and Production Arrangement for Minor-Routine Street Repair and Maintenance ...... Table 6.4 How Street Provision Responsibilities Are Shared and Distributed among Provision Units ...... Figure 7.1 School Districts in St . Louis County ...... Figure 7.2 Relation between Assessed Valuation Per Student and Levy Rate. by District-1984-85 ...... Figure 7.3 Variation in Per Student Spending. St . Louis County Schools-1984-85 ...... Figure 7.4 Relation between Current Expenditures Per Student and Foundation Funding Per Student. by District. St . Louis County-1984-85 ...... Figure 7.5 Relation between Current Expenditures Per Student and Student-Educator Ratio for Elementary Schools. by District. St . Louis County-1984-85 ...... Figure 7.6 Relation between Current Expenditures Per Student and Student-Educator Ratio for High Schools. by District. St . Louis County-1984-85 ...... Table 7.1 Percent of School Age Children in Private Schools. by District. St . Louis County-1980 ...... Table 7.2 County School Districts Ranked by Median Household Income. with Fiscal and Demographic Characteristics ...... Table 7.3 County School Districts Ranked by Student Achievement on Standardized Tests of Basic Skills-1982 ...... Table 8.1 Regression Cpefficients for Local Service Expenditures in County Municipalities-19 8 5 ...... Table 8.2 Correlation and Regression Analysis for Total Per Capita Revenues and Community Characteristics in County Municipalities-1985 ...... Table 8.3 Tax Burdens on Residential Property-County Municipalities ...... Table 8.4 Regression Coefficients for Residential Tax Burdens in County Municipalities-1985 Table 8.5 School Revenue Variations in St . Louis ...... Table 8.6 Correlation Analysis for Total and District Revenues Per Student in Average Daily Attendance- 1984-85 ...... Table 8.7 Variables Used for Computing Cost Index of County Municipalities-1985 ...... Table 8.8 Total Local Cost Model for County Municipalities-1985 ...... Table 8.9 Actual and Predicted Service Costs in County Municipalities by Sales Tax Status-1985 ...... Table 8.10 Employment Growth in St . Louis and Comparison Metropolitan City-Counties- 1970-85 ...... Table 8.1 1 Patterns of Employment in St . Louis City and County-1970-85 ...... Appendix Table 8.1 Revenues and Expenditures in County Municipalities-1985 ...... Appendix Table 8.2 Service-Cost Burden on an Average Household-1985 ......

Chapter One Metropolitan Organization: The Challenge of Understanding

INTRODUCTION research, tending to support this theory, is briefly reviewed. Further research is clearly needed to ex- Metropolitan organization has for much of this plore the validity of competing theories. Propositions century been viewed as a major challenge for to guide such an inquiry are developed, as are a se- American local government institutions. The focus ries of research questions and criteria for evaluating of concern has been largely on the "fragmentation" the results. The chapter concludes with a discussion of metropolitan areas by a host of governmental ju- of the St. Louis case study and its methodology, to- risdictions-municipalities, counties, special districts, gether with a preview of the remaining chapters in and other governmental units-that often overlap, this report. producing a pattern of organization frequently characterized as a "maze," a "jungle," or a "crazy- THE CRITIQUE OF FRAGMENTATION quilt." Many students of local government have argued that there is a need to solve "the metropolitan Traditional Theory problem" by means of areawide reorganizations in- In the traditional view of metropolitan organiza- tended to simplify the structure of local governments tion, a number of specific pathologies are thought to in metropolitan areas. This approach to metropoli- affect metropolitan areas, including (1) fragmenta- tan reform is predicated on an assumption that the tion of authority, (2) overlapping jurisdictions, (3) people and economic enterprises located within any duplication of effort, and (4) fiscal disparity and in- metropolitan area constitute a single, integrated equity. Metropolitan communities are viewed as community in need of a uniform pattern of organiza- fragmented among scores of separately organized tion. From this perspective, a metropolitan area municipalities and special purpose districts. As long ought to be organized with reference to a single, as these various units of government have authority metropolitanwide unit of government and, if neces- to act independently, coordinated action on the ba- sary, a uniform arrangement of subunits.' sis of metropolitanwide concerns is thought to be Alternative perspectives that view the fragmen- difficult-often impossible. Overlapping jurisdic- tation of local governments quite differently have tions-such as counties, townships, and special dis- also received considerable attention in recent years.* tricts that often overlap municipalities-are believed In this chapter, the traditional theory is presented to sow conflict, confusion, and discord. Redundant and some difficulties with it are discussed. Jurisdic- service capabilities on the part of overlapping juris- tional fragmentation is distinguished from functional dictions are thought to create wasteful duplication of fragmentation, and the implications of this distinc- effort on the part of local officials, managers, and tion are explored. Following this discussion, an al- service producers. In addition, a large number of ternative theory is outlined, and a body of empirical relatively small, independent jurisdictions is identified with lopsided patterns of fiscal disparity, a mis- forms with the consent of citizens. To the contrary, match between tax resources and service needs. jurisdictional fragmentation not only persists but also The lack of uniform, fully integrated patterns of continues to develop in most metropolitan areas. metropolitan organization, it is argued, leads to inef- Institutional reform has been a basic part of the fectiveness in service delivery, administrative ineffi- American political experience-a method by which ciency, lack of political accountability, and interjur- institutions are adapted to changing conditions in a isdictional inequities. Service delivery is considered self-governing society, based on a diagnosis of pa- deficient and often ineffective because some urban thologies associated with some set of problems expe- and metropolitan problems do not coincide with mu- rienced by ~itizens.~Effective reform depends, how- nicipal boundaries. Spillover effects-ranging from ever, on a discriminating analysis of the factors that fleeing criminals to stray dogs that elude animal con- lead to problems. Fragmentation, instead of being trol-may impede effective service delivery. At the subjected to discriminating analysis and thorough same time, an inability on the part of a large number study, has often been treated as a general malady to of small jurisdictions to take advantage of economies be eradicated, not as a configuration of variables of scale may lead to inefficiency. Lack of political that may or may not be related to a range of spe- accountability stems from the confusion thought to cific-desirable or undesirable-outcomes. be associated with overlapping jurisdictions, that is, The term fragmentation itself is pejorative. citizens cannot pin responsibility on a single official While usually defined as the number of governments or small set of officials; instead, officials are able to per 10,000 or 100,000 persons in a metropolitan "pass the buck" and avoid taking responsibility for area, fragmentation is hardly a neutral descriptor. It their actions or inactions. Metropolitan fragmenta- evokes an image of something that is broken down tion is also believed to be a means of segregating the and cannot work effectively. The term itself invites rich from the poor. Fiscal disparities among local ju- the treatment of empirical propositions as self- risdictions are identified with serious inequities evident. Thus, frequently, the only evidence offered based on social class and on race. of fragmentation amounts to a definition-a count of Metropolitan reform proposals, which initially governmental units, or the ratio of governments to focused on the need for a single metropolitan gov- population. Enumerating the governmental units ernment, have more recently been modified to in- found in a metropolitan area, and then characteriz- clude a two-tier arrangement-two levels of local ing the result as fragmented, does no more than government with a metropolitan level predominant.3 name and measure a variable; it does not establish This modification was developed in response to even a single proposition that links one variable to growing criticism of big-city governments as unre- another. The relevant empirical question is the ex- sponsive to neighborhood concerns and to a call for tent to which jurisdictional fragmentation leads to neighborhood governments to achieve community functional fragmentation, that is, a lack of necessary control of basic local services.4 The two-tier ap- and useful coordination among jurisdictions with re- proach still insists on a metropolitan unit of govern- spect to specific functions. This proposition has been ment able to govern comprehensively. Although widely accepted as self-evident. there would be overlap, the larger unit would be The most prominent metropolitan reform pro- dominant, and the crazy-quilt pattern of fragmenta- posals have taken the form of city-county consolidation and overlap would be replaced by a uniform set tions. Proposed consolidations, with a few notable of jurisdictions with functions neatly divided be- exceptions,' have been put before a popular refer- tween the two tiers of a metropolitan government. endum. The acceptance rate, averaging about 20 The underlying theory was basically unchanged, and percent, has steadily declined since the 1940s.8 comprehensive reform remained the fundamental Many of the consolidations approved have fallen metropolitan challenge. considerably short of achieving a unified metropolitan government, either by leaving independent mu- Difficulties with the nicipalities within the county or by allowing special Traditional View districts to continue in operation.9 One two-tier sys- Critics have pointed to a number of difficulties tem has been established, in Miami-Dade County, with the traditional theory of metropolitan organiza- Florida. In the meantime, the number of special dis- tion. One difficulty derives from a primary emphasis tricts in the United States has increased by more on reform and action instead of inquiry and analysis. than 50 percent over a 25-year period, from 18,323 Little systematic evidence has been collected that in 1962 to approximately 29,487 in 1987. This num- supports the reform view.5 Another is the relative ber includes some 900 districts created on an lack of success enjoyed by reform advocates despite areawide basis.10 Special districts are often created repeated opportunities to adopt metropolitan re- with the consent of local voters." In many parts of the country, municipal incorporation has also con- sponse to emergencies.lS A related study of the ef- tinued as metropolitan areas grow. Nationwide, the fect of fragmentation and overlap on the efficiency number of municipal governments in the United of police agencies in metropolitan areas found that States has increased from about 18,000 in 1962 to areas exhibiting greater fragmentation with overlap about 19,205 in 1987. Perhaps the major consolida- out-perform both (1) areas with fragmentation but tion success, on the other hand, has been the drastic less overlap and (2) areas with less fragmentation.18 reduction in the number of independent school dis- In terms of fiscal effects, studies have shown that tricts, reduced from 34,678 in 1962 to about 14,741 highly fragmented areas tend to have lower expendi- in 1987. (Most of the reduction occurred in the tures per capita, even when controlling for the level 1960s. with 18,897 school districts eliminated be- of demand for services." One study found that, in tween 1962 and 1972; only about 1,040 school dis- terms of equity effects, the least advantaged jurisdic- tricts have been eliminated since 1972.) Often, how- tions in metropolitan areas tend to be brought into ever, the elimination of local districts may have been parity with the most advantaged jurisdictions by in- accomplished by weakening the authority of local tergovernmental revenue transfers.18 The same communities to decide issues of school reorganiza- study reports that, contrary to the prevailing image, tion. fragmented suburban areas are not sharply divided The repeated failure of metropolitan reorganiza- along income lines. Instead, most suburban munici- tion efforts over a period of several decades is a seri- palities are heterogeneous with respect to income, ous anomaly in a reform-centered theory. If metro- and a growing number are heterogeneous with re- politan reform addresses salient problems spect to race. Other studies have found mixed ef- experienced by citizens, then why do citizens tend to fects,lQbut, on balance, the hypothesis that decreas- reject reform proposals at the polls? This question ing fragmentation will lead consistently to greater has attracted the attention of numerous analysts.12 efficiency, responsiveness, or equity has found little In a careful review of the issue, Willis D. Hawley empirical support. concluded that the way in which voters calculate the This line of research cannot yet be regarded as costs and benefits of "metropolitan integration" is a conclusive. Much of the research on fiscal effects crucial intervening variable in determining the suc- (taxes and expenditures) employs measures of frag- cess of integration proposals, including not only met- mentation that tend to perpetuate a shortcoming of ropolitan consolidation but also the establishment of traditional theory-assuming fragmentation to be a metropolitanwide special districts and authorities.13 simple characteristic that can be represented by a The creation of special districts and, to a some- single indicator. To study fragmentation in a more what lesser extent, the continuing process of munici- discriminating way, researchers must be able to dis- pal incorporation, indicate that local voters are not tinguish metropolitan areas on more than a single indiscriminately opposed to change. Rather, they se- indicator. Metropolitan areas with similar fragmen- lectively adopt those changes that often maintain, tation scores can exhibit quite different patterns of and even increase, the degree of both fragmentation organization.20 Research must get "insiden a frag- and overlap within metropolitan areas. mented system and learn to understand it on its own In the traditional approach to metropolitan or- terms. The metropolitan "bumblebee" has to be ex- ganization, metropolitan areas in the United States plained before it can be redesigned to make it fly are a lot like a bumblebee. According to the laws of better. thermodynamics, it is said, a bumblebee can't fly. For these reasons, the Advisory Commission on According to the precepts of traditional theory, a Intergovernmental Relations has initiated a research fragmented metropolis can't work. Yet fragmenta- program to investigate the operational characteristics tion has not fallen, as it were, from its own weight. of jurisdictionally fragmented metropolitan areas. Why not? The focus of inquiry is on functional arrangements- the arrangements through which specific functions, JURISDICTIONAL FRAGMENTATION, such as taxation, service delivery, and governance, FUNCTIONAL ARRANGEMENTS are actually performed. Functional arrangements A decade and a half of research has begun to can be distinguished from jurisdictional arrange- suggest some answers as to why fragmentation per- ments. Jurisdictional arrangements refer to authority sists in metropolitan areas.14 An extensive series of patterns, while functional arrangements refer to ac- studies comparing the performance of police agen- tion patterns. The problem of metropolitan organi- cies in small jurisdictions with those in larger juris- zation is concerned with how authority patterns af- dictions has consistently found that small agencies fect action patterns. Jurisdictional arrangements, by tend to be more responsive in the delivery of neigh- allocating authority, establish the legal capacity to borhood-type police services, such as patrol and re- create a range of different functional arrangements. For this reason, jurisdictional arrangements do not The basic methodological problems in a case determine functional arrangements, but instead cre- study are knowing what to look for and how to inter- ate possibilities within limits. While some possibilities pret what is found. These research ingredients must are foreclosed by jurisdictional fragmentation, be supplied a priori, through the use of theory. The among those left open are a wide variety of possible explicit use of an a priori theory does not imply that arrangements for interjurisdictional coordination. a study lacks objectivity. Knowing what to look for Functional arrangements are crucial intermedi- does not mean finding it. The use of theory does ate or intervening variables in an understanding of imply self-conscious attention to the assumptions metropolitan organization-variables that intervene used to direct inquiry. A study that is well grounded between (1) jurisdictional arrangements (some con- in theory can be critically appraised by others, who figuration of fragmentation and overlap) and (2) may propose different assumptions and thus make outcomes (efficiency, responsiveness, and equity). different interpretations of findings. In what follows, The range of possible functional arrangements in- a theory of metropolitan organization is described cludes both independent action by each jurisdiction that shows how jurisdictional fragmentation and (with potential duplication) and various forms of co- overlap can provide an institutional foundation for a ordinated action by two or more jurisdictions. The productive system of functional relationships among decisionmaking processes related to the emergence jurisdictions. of such functional arrangements may include cooperation, contracting, collusion, competition, conflict, A THEORY OF HOW negotiation and conflict resolution, rulemaking and FRAGMENTATION WORKS enforcement, problem solving, and buck passing. One of the most prominent nontraditional theo- This approach to metropolitan research takes a ries of metropolitan organization and governance is step beyond the traditional study of metropolitan ar- presented in a paper published in 1961 by Vincent eas. Rather than assuming that jurisdictional frag- Ostrom, Charles M. Tiebout, and Robert Warren.22 mentation inexorably leads to functional fragmenta- Their approach is rooted in a theory of goods in the tion, the relevant processes and arrangements economic sense, distinguishing public goods from become objects of inquiry. In this way, it should be private goods as well as recognizing a range of goods possible to learn the basic organizational dynamics and services with mixed public and private charac- of a jurisdictionally fragmented metropolitan area- teristics. Local public goods are basically of the in a phrase, to learn how fragmentation works or mixed variety. A private good is one that individuals operates. can be excluded from consuming and that is con- Such a research program is best conducted by sumed separately by individuals or households, with- means of a series of case studies-where each case is out affecting others. A public good is one that indi- a metropolitan area-designed to study the linkages viduals cannot be excluded from consuming and between jurisdictional patterns and functional pat- that is consumed jointly by individuals. Private goods terns. The limitation of this approach is that it does are sometimes characterized as packageable, public not shed light directly on the comparative perform- goods as not packageable. Ostrom, Tiebout, and ance of more and less fragmented areas. The latter Warren argue that local public goods can be con- requires a focus on the final outcomes of service strued as packageable, but not to individuals. In- provision-how responsive are the police? how effi- stead, local public goods are provided in community cient is the trash collection?-based on performance packages: individuals outside a community are rela- measurements. Research that focuses on outcomes is tively unaffected by provision, while those within the better pursued through comparative analysis that community enjoy a benefit in common.23 draws on aggregate data from a large number of These community packages come in different metropolitan areas.21 sizes. Some local public goods affect only a very While not final outcomes, the functional ar- small community of interest, such as a residential rangements that lead directly to outcomes are inter- neighborhood. Others affect larger communities of mediate effects of practical, as well as theoretical, interest-including subregions within a metropolitan interest. One practical interest lies in being able, po- area, the metropolitan area itself, and beyond. This tentially, to improve the functioning of metropolitan diversity creates various problems of scale in local areas in ways not usually considered in the tradi- public organization. Ostrom, Tiebout, and Warren tional metropolitan reform approach. If, as aggre- suggest four criteria to consider in choosing an ap- gate research increasingly indicates, metropolitan ar- propriate scale of organization: control, efficiency, eas that exhibit a high degree of jurisdictional representation, and self-determination. fragmentation can perform comparatively well, it is The criterion of control is used to establish the important to learn how and why. boundaries of a jurisdiction so as to be able to exercise an appropriate degree of control over some set government would be quite simple-single, general of circumstances. Metropolitan reform assumes that purpose local governments with mutually exclusive a metropolitanwide jurisdiction is required to exer- (nonoverlapping) jurisdictions. Perhaps these would cise control over circumstances that encompass a be metropolitan jurisdictions, perhaps not. Fre- metropolitan community. Examples include major quently, however, the criteria suggest quite different highway developments, traffic control on thorough- scales of organization for different circumstances, fares, and problems associated with smog. The diffi- involving different types of local public goods. This culty is that other circumstances that need to be suggests the desirability of organizing different com- controlled in order to produce local public goods munities of interest in metropolitan areas with over- (e.g., the control of street crime) involve neighbor- lapping jurisdictions, some communities nested in- hood conditions that, considered on a metropolitan side of others. The two-tier approach favored by scale, exhibit great variety. A central problem of some advocates of metropolitan reform recognizes a metropolitan organization, therefore, is to permit potential need for two different scales of organiza- the control of some circumstances on the basis of tion. There is no a priori reason, however, to limit uniform regulations across a metropolitan area, the number to two, nor to assume that the number while permitting variation in the control of other cir- should be uniform throughout a metropolitan area. cumstances that are experienced on a much smaller The larger and more complex a metropolitan area, scale. Still other circumstances requiring control the greater the amount of overlap or nesting of juris- may embrace an intermediate community, not fully dictions to be expected, and the more complex the metropolitan, but bigger than a neighborhood (e.g., overall pattern may be. the regulation of a ground water supply). Some complexity could still be avoided if the The criterion of efficiency looks at scale in terms first three criteria were to coincide for broad catego- of production economies and seeks the most favor- ries of local public goods, even though a degree of able economy of scale. Presumed economies of scale fragmentation and overlap would appear to be opti- also underlie much of the metropolitan reform ef- mal. Where the criteria diverge for the same good or fort. Nevertheless, economies of scale differ dra- service, however, local organization becomes even matically among various types of goods and services. more complex. Suppose, for example, that the crite- Labor-intensive goods (e.g., police or education) rion of efficiency suggests one scale of organization are understood to exhibit considerably lower econo- while the criterion of political representation suggests mies of scale than capital-intensive goods (e.g., a smaller scale or a larger scale. In this situation, the water or sewer supply). The most favorable econ- provision of public goods can be separated from the omy of scale in production can therefore be ex- production of goods. Provision refers to decisions pected to vary among different local public goods about whether to provide a good, how much to pro- and services. vide, what the quality standards ought to be, and The criterion of political representation requires how to arrange for production and delivery. Produc- that those whose interests are substantially affected tion refers to the actual transformation of resources by some set of circumstances be included within the into products or services. Provision can be consti- boundaries of a political jurisdiction able to act. This tuted to reflect the criterion of political representa- criterion also requires that those who are substan- tion while production is organized to obtain the most tially unaffected be excluded from such a jurisdic- favorable economy of scale. The nexus between pro- tion." vision and production may consist of a contractual Finally, the criterion of local self-determination or cooperative relationship.25 presumes that the persons directly affected by a par- Overlapping jurisdictions can also be used to ticular set of circumstances are the appropriate de- bring different communities of interest to bear on cisionmakers to apply the first three criteria noted the same set of circumstances. In this case, the crite- above. The choices of local citizens, rather than the rion of representation recommends that two or more decisions of external authorities, will control the ba- jurisdictions, one nested inside the other, be used to sic structure of organization in a metropolitan area. make collective decisions. Larger, overlying jurisdic- This approach involves more than allowing citizens tions can affect the decisions of smaller jurisdictions to make choices through popular referenda at a met- in two ways: (1) offering financial incentives- ropolitan level. Instead, it may allow various com- grants-in-aid-to modify the allocation of resources munities of interest (including the metropolitan by recipient jurisdictions and (2) regulating the pro- community) to constitute themselves as separate ju- vision and/or production of goods and services by risdictions. making and enforcing rules that take into account a Ideally, if the first three criteria coincided for all larger community of interest. In these ways, metro- local public goods, the preferred pattern of local politan governance can allow those directly affected by a set of circumstances to make some decisions patterns are constrained by rules, which may not al- autonomously, while at the same time allowing more low private production of a public service (e.g., inclusive jurisdictions to take account of indirect or often the case with police), and by economies of spillover effects in their fiscal and regulatory deci- scale and/or the criterion of control, which may limit sions. the number of service producers that can feasibly The foregoing theory can be used to derive a set operate in an area. Competition is not always desir- of general propositions, as follows: able, but competition among the producers of goods Proposition 1: The variabk nature of local and services is generally understood as economically public goods, when combined with a reliance on beneficial to consumers, in this case the collective local self-determination, will be associated with consumers who comprise local government jurisdictions. Thus: a complex pattern of jurisdictional fragmentation and overlap. Proposition 5: Depending on the rules and the nature of specific goods and services, sepa- Proposition 2: A complex pattern of juris- ration of provision and production will tend to dictional fragmentation and overlap plus a base engender patterns of competition among pro- rule of willing consent for constituting relation- ducers and, in turn, choices for collective con- ships among jurisdictions will be associated with sumers. functional arrangements that link multiple jurisdictions, including (to some degree) the separa- Competition can also occur to some extent tion of provision and production for specific among separately organized communities when citi- services. zen consumers search out their preferred tax-benefit packages.26 Proposition 3: A complex pattern of juris- Conflict is also expected to arise. The sources of dictional fragmentation and overlap that in- conflict include spillover effects from one jurisdic- cludes access to rulemaking authority on the tion to another, in those situations where community part of an overlying jurisdiction will tend to gen- packages are not "airtight," and disputes that arise erate areawide rules that constrain more local- from tax-service mismatch, where some taxpayers ized provision and production activities in order perceive themselves as unfairly subsidizing others or to control potential spillover effects. not receiving services commensurate with their tax bills. More generally, the circumstances that need to These propositions presume that the particular be controlled in order to produce public goods are pattern of organization in a metropolitan area, in- not unchanging. Jurisdictional boundaries and the cluding both jurisdictional and functional arrange- distribution of public authority have to be adapted to ments, can be explained by (1) the nature of those changing conditions. The critical question in such a specific goods provided and (2) the specific rules of system of government is not the presence or absence local self-determination. of conflict, or the ability to suppress it, but the main- Ostrom, Tiebout, and Warren anticipate that tenance of facilities to resolve conflicts. State and such a pattern of organization will engender both co- federal courts and state legislatures to some extent operation and competition among local jurisdictions. provide such facilities in the governance of metro- Cooperation among jurisdictions organized to reflect politan areas in the United States. The use of exter- the criterion of political representation, and the for- nal decisionmakers, however, potentially threatens mation of new overlapping jurisdictions, can be used local self-governance. This possibility creates incen- to address problems that arise in relation to both the tives for local citizens and officials to develop local criterion of control and the criterion of efficiency. forums for conflict resolution within a metropolitan Thus : area. External decisionmaking facilities (such as a Proposition 4: Depending on the rules of state legislature) are used primarily to constrain self-determination and the nature of specific holdouts, bring the relevant parties into negotiations, goods and services, fragmented jurisdictions will and ratify agreements negotiated locally. Traditional tend to develop overlapping jurisdictions (to the metropolitan reform theory frequently refers to a extent the latter are not already present) and Balkanization of local jurisdictions, but this charac- both will tend to develop patterns of cooperation terization presupposes an absence of conflict resolu- on the basis of mutual advantage. tion arrangements. Thus: The ability to separate provision and production Proposition 6: Jurisdictional fragmentation can also lead, however, to patterns of competition will tend to engender conflict and, depending on among producers, both public and private. These the rules of self-determination and the availability of overlapping jurisdictions, also engender service providers take advantage of potential functional arrangements for conflict resolution competition? locally. To what extent do metropolitanwide jurisdictions exist? Have multijurisdictional arrangements been developed to attend to RESEARCH QUESTIONS metropolitanwide problems and circum- The theory sketched above can be used to de- stances? sign a research program to search for characteristic Are consistent patterns of fiscal advantage functional arrangements in jurisdictionally frag- or fiscal disadvantage observed among local mented metropolitan areas. Two basic sets of vari- jurisdictions? ables are involved: (1) the variable nature of local public goods and the circumstances surrounding What arrangements have been developed to their provision and (2) the rules of local self- resolve conflicts and facilitate agreement determination or self-governance. Within these con- among local jurisdictions? straints, local citizens and officials are expected both What unresolved problems are present? Do to form a variety of jurisdictions and to work out these threaten the basic processes of metro- various functional arrangements among jurisdictions politan governance? for (1) solving problems of scale and (2) resolving conflicts. The precise arrangements depend on the EVALUATIVE CRITERIA independent and joint effects of the two sets of vari- Having produced a description of the functional ables noted above. arrangements associated with jurisdictional fragmen- In addition to functional arrangements, a case tation, there remains the task of evaluating those ar- study can examine fiscal patterns among jurisdic- rangements. Ultimately, this task belongs to citizens tions-relationships that can also be considered im- and their elected officials. Research can contribute portant intermediate effects of jurisdictional struc- to evaluation only by carefully specifying criteria and ture. Fiscal disparities among jurisdictions are a indicating to what extent existing patterns and out- frequent source of conflict in metropolitan areas. An comes satisfy those stipulated criteria. The simple ability to sustain agreeable and equitable fiscal rela- existence of jurisdictional fragmentation and overlap tionships may, therefore, be a critical factor in main- cannot be used as a criterion of evaluation in taining a framework of local self-governance. this study. Instead, the study must specify either A number of research questions can be sug- functional-process criteria, to evaluate functional gested for investigating the operational characteris- arrangements and associated decisionmaking proc- tics of a jurisdictionally fragmented metropolitan esses directly, or performance criteria, to evaluate area: the products and outcomes of functional arrangements. Performance measurement, however, as dis- 8 What are the rules of local self- cussed above, lies for the most part beyond the determination? To what extent do the rules scope of this study. allow citizens to organize themselves on the Process criteria are used to evaluate processes basis of various communities of interest, ac- directly rather than (or in addition to) evaluating the cording to the criterion of political represen- outcomes of those processes. Democratic govern- tation? ment, for example, is usually evaluated according to 8 To what extent have local voters created certain process criteria, as well as according to out- overlapping jurisdictions? Is the overlap come criteria. The functional arrangements associ- congruent with the nature of goods pro- ated with metropolitan organization can also be vided? evaluated according to process criteria. The tradi- rn To what extent is production separated from tional and alternative theories of metropolitan or- provision? Is this separation explained by ganization approach these criteria somewhat differ- economies of scale? ently. Some of the following functional-process criteria are acceptable to both schools of thought, rn Do adjacent jurisdictions and overlapping while others are the subject of disagreement: jurisdictions regularly cooperate to obtain rn Self-Determination and Citizen Choice. mutual advantage? Is there significant and This is a criterion on which a highly frag- needless duplication of effort among these mented area is expected to perform well. To jurisdictions? what extent are citizens as voters able to es- 8 Are there significant patterns of competition tablish and control a variety of local govern- among differentiated service producers? Do ments, adjust their boundaries, and transfer authority among jurisdictions? To what ex- necessary and appropriate for obtaining ac- tent can citizens as individuals choose tion on metropolitan problems. among alternative jurisdictions to locate a residence or business? The question here is These criteria do not directly answer perform- one of process, not just rules: to what extent ance questions, such as how good garbage pickup or do the rules of self-determination actually police protection or street maintenance may be in enable self-determination and citizen choice any of the numerous jurisdictions of a metropolitan to occur? Disagreement among analysts is area. However, the criteria do identify processes or likely to be normative, with critics of frag- arrangements that are necessary to effective and ef- mentation arguing that too much self- ficient performance, even if they do not guarantee determination and citizen choice leads to it. By focusing on the functional arrangements adverse consequences in terms of other cri- through which performance occurs, the criteria used teria. in this study can illuminate metropolitan organization in a way that measuring the outcomes of per- Representation and Accountability. Can formance alone cannot. This approach provides a various communities of interest-of differing basis on which to evaluate the working parts of an sizes-gain effective representation of their organizational mechanism rather than treating the views? Are the costs to citizens of making inner "works" of public organization as a "black their views known to public officials kept box" that cannot be known and understood.27 Such reasonably low? Are elected public officials a focus on process is especially important when an and administrators effectively accountable inquiry is expected to help decisionmakers "repair" to relevant communities of interest? Here those inner works by reforming the institutions of again a fragmented system might be ex- metropolitan governance. pected to perform well, except that the tra- In addition to process criteria, the study gives ditional approach to metropolitan organiza- attention to intermediate effects measured in terms tion would anticipate a deficiency of of revenues and expenditures. The fiscal relation- representation and accountability for ships among local government units, like functional areawide interests. arrangements, do not fully determine the quality of performance, but clearly do affect performance ca- Coordination. Are closely related services pacity. Fiscal relationships can therefore provide and service components produced in a coor- clues to performance, using the following criteria: dinated manner? Traditional theory antici- Economies of Scale. Are services and serv- pates serious deficiencies in coordination- ice components produced on a scale that the price, perhaps, of too much captures significant scale economies? Are self-determination-while the alternative there significant uncaptured economies of theory of fragmentation expects significant scale that remain? functional coordination to emerge from arrangements that cross jurisdictional bounda- ' rn Equity. Do fiscal differences among local ries. jurisdictions reflect patterns of racial or income-class difference~?To what extent are Competition among Service Producers. some types of communities relatively advan- Do local jurisdictions "shop around" for the taged or disadvantaged in fiscal capacity? most economical mode of service production? Traditional and alternative theories Although fiscal data alone, without measures of differ on whether competition among pro- performance, cannot establish a definitive assess- ducers is beneficial or detrimental in a sys- ment of either efficiency or equity, the relationships tem of local government. Shopping around among fiscal variables can provide indicators in each implies occasionally shifting from one pro- case. ducer to another, which also may be costly. The orientation taken in this study presumes that problems will arise among a multiplicity of local Metropolitan Problem Solving. Are genu- jurisdictions. Efforts to solve these problems may inely metropolitanwide problems effectively not always be successful. The key difference be- addressed on a metropolitan basis? Tradi- tween the approach taken here and traditional met- tional and alternative theories agree that this ropolitan research is that jurisdictional fragmenta- is an appropriate criterion, differing only in tion per se is not presumed to be problematic. the institutional arrangements thought to be Instead, the inquiry searches for processes of problem solving that may operate among fragmented ju- pal limits reside in cities or villages of 5,000 to risdictions. 25,000 people. Still, there are 21 municipalities with a population under 1,000, many of which are clus- THE ST. LOUIS CASE STUDY tered closely together in an area of the county The selection of the St. Louis area as a research known as Normandy. All those who live outside mu- site was prompted by its status as one of the most nicipal boundaries, and most of those who live in the jurisdictionally fragmented metropolitan areas in the smaller municipalities, receive their fire protection nation, an area often cited as a prime example of services from one of 25 independent fire protection perverse organizational patterns.28 St. Louis County districts. There is also a large, undetermined num- (which is separate from St. Louis City) was listed by ber of organized subdivisions, both inside and out- the U.S. Bureau of the Census in 1982 as having side municipalities, that provide their residents with 151 governmental units. Measuring fragmentation as an array of local street services. The ACIR study the number of governmental units per 10,000 peo- counted at least 427 street-providing subdivisions in ple, the county has a fragmentation score of 1.55. the incorporated area of the county alone. Public For purpose of comparison, Cook County in Illinois education is provided by 23 independent school dis- (which includes its central city, Chicago) has 516 tricts, varied in size, and by a countywide special governments, but is somewhat less fragmented by school district. this measure, with a score of 0.98. Allegheny The study is strongly focused on St. Louis County, Pennsylvania (which includes Pittsburgh), County, since the degree of fragmentation in the on the other hand, is more fragmented, with 323 county is much greater than in St. Louis City and governments and a score of 2.8 1. County combined. St. Louis County is a diverse The selection of St. Louis represents, therefore, metropolitan county that would qualify as a metro- the deliberate choice of a "hard case."29 The proc- politan area entirely on its own, containing almost esses and arrangements that characterize the organi- half of the population of the St. Louis Metropolitan zation and governance of fragmented metropolitan Statistical Area (MSA). The MSA had a total 1980 areas ought to be writ large in St. Louis County. If population of 2.35 6,460 and included three addi- jurisdictionally fragmented areas are marked by con- tional counties in Missouri and four in Illinois, plus fusion, discord, and a lack of cooperation among the City of St. Louis. The MSA as a whole was actu- independent units of government, St. Louis County ally more fragmented than St. Louis County, con- and the greater St. Louis area should display these taining 663 governments for a fragmentation score characteristics in abundance. The traditional view of of 2.8 1 governments per 10,000 people. fragmentation would suggest that the greater the Although St. Louis County provides the princi- fragmentation, the greater the discord and confu- pal frame of reference in this study, much of the sion. analysis also includes reference to St. Louis City. Moreover, since St. Louis has frequently been a The choice depends on what is being analyzed. In focus of study in the past, there exists a substantial some cases, the frame of reference is St. Louis City base on which to build new research. In the 1950s, and County, and, in a few cases, the MSA. The re- St. Louis was the focus of an extensive series of lationships among units of government within St. studies inspired by the challenge of metropolitan re- Louis County are quite different from relationships form.30 In the mid-7Os, the St. Louis area was one among all governments in the MSA. Somewhat dif- of a number of metropolitan areas selected for an ferent conclusions can be drawn about functional intensive study of police services.31 Later, Oscar metropolitan arrangements depending on the frame Newman studied "private street associations" in St. of reference. Louis in a search for architectural and organiza- St. Louis City and County have been the focus tional arrangements conducive to the maintenance of myriad recommendations for reform-principally of residential communities.32 for total or partial consolidation of governments. Ef- The County of St. Louis, which has been institu- forts were made to reunite St. Louis City and tionally separate from the City of St. Louis since County in 1926, and again in 1959 and 1962, with- 1876, is a metropolitan community of nearly a mil- out success. The 1959 and 1962 efforts were aimed lion people-973,896 according to the 1980 census. at establishing different variants of a two-tier system. Sixty percent of the county's residents live in 90 mu- City voters approved the 1926 attempt, but it failed ni~ipalities;~3the other 40 percent live in unincor- in the county. In 1959 and 1962. voters in both ju- porated areas (about the same population as St. risdictions disapproved, although county voters dis- Louis City). The largest municipality in the county- approved by a much larger margin than did city vot- Florissant-had a 1980 population of only 5 5,372. ers. At the same time, however, St. Louis voters The great majority of those who live within munici- approved the creation of countywide and joint city- county special districts. Like voters in metropolitan 2. What private activities to regulate, and the areas throughout most of the United States, resi- type and degree of regulation to use; dents of St. Louis were willing to address metropolitan problems, but only in ways that maintained and 3. The amount of revenue to raise, and how to extended the type of system already in place. Most raise it (whether by various forms of taxa- recently, the St. Louis County Executive has tion or by user pricing); launched an effort to incorporate the entire county while reducing the number of municipalities by more 4. The quantities and quality standards of than half. A new period of discussion centered on goods and services to be provided; and reorganization and consolidation is now under way. 5. What arrangements to make for the production of goods and services. METHODOLOGY: The basic choice on the provision side is DISTINGUISHING PROVISION AND PRODUCTION whether to make any sort of provision at all. If no provision is made, that activity remains private (ex- The ACIR field study collected data from both cluding those activities tended to by other units of personal interviews34 and published data sources, government). If provision is to be undertaken, how and included both legal and economic materials. In- to provide is another basic choice. The two major terviews were conducted in more than a third of the alternatives are (1) taxing and spending to provide a county's municipalities and school districts, usually public good or service and (2) regulating private ac- with elected mayors and school superintendents, tivity in order to shape private decisions to a public and with county government officials. Data were also purpose. Provision activities thus include, on the one collected from a sample of 53 organized subdivisions hand, setting both tax rates and user charges and that provide street services. Some police and fire of- choosing how to spend public money, and, on the ficials were also interviewed, in addition to a small other hand, enacting and enforcing laws or rules that number of private citizens active in municipal incor- constrain private behavior according to public crite- poration efforts. Most of the field work and data ria. Finally, if goods and services are to be provided, collection was completed in the summer of 1986. then choosing to organize a production unit and/or Extensive correspondence has also taken place with selecting and hiring producers (both public and pri- a number of local officials and academics. vate), monitoring the quality and quantity of goods The field research focused on three main areas and services supplied, and representing the interests of inquiry: of consumers to producers all become significant provision activities. Legal rules and functional arrangements for Production, on the other hand, refers to the the governance of St. Louis County and the more technical processes of transforming inputs into city-county region, as reflected in the basic outputs-making a product or, in many cases, ren- "rules of the game" (found, for the most dering a service. Although production is often part, in state statutes) and accompanying viewed as entirely the work of agents (public or pri- rulemaking practices; vate), it is frequently better viewed as "coproduc- 2. Fiscal relationships among local government tion," a procek whereby a specialized producer units in St. Louis City and County; and (such as a teacher) interacts with a citizen-consumer (a student) to produce a good (education) 3. Functional arrangements for the provision The distinction between provision and produc- and production of local public goods and tion lays the conceptual foundation for a new under- services in St. Louis County, with a detailed standing of the organization of local public econo- examination of arrangements in four service mies.36 Different considerations apply in the choice areas: police, fire, streets, and education. of an organizational unit to provide a service from The inquiry into functional service arrangements those involved in the choice of an organizational unit to produce a service. The work of local govern- was guided by a distinction between the provision of local public goods and services and their production. ment is increasingly viewed in terms of provisioning, Provision and production are functionally quite dif- and not necessarily in terms of producing. Although ferent. Provision refers to collective choices that de- the organization of production can be, and often is, termine: governmental, frequently it can become a private responsibility. Patterns of organization on the provi- 1. What goods and services to provide (and sion side can differ from those on the production what are to remain private) ; side. Various functional arrangements can also be tele.37 Even if two or more provision units make used to link provision units to production units: some arrangement for service production, duplica- In-House Production. A provision unit or- tion can be avoided through a variety of functional ganizes its own production unit. This is the arrangements: (1) joint service delivery by a single traditional model of local organization. Mu- production unit (jointly established by two or more nicipalities organize municipal departments provision units) ; (2) coordinated service delivery on for police, fire, public works, and so forth. the part of two (or more) production units; and (3) alternated service delivery in which two (or more) Coordinated Production. Two or more production units alternate in delivering service on production units (organized by their respec- the basis of time, place, or clientele. Instead of as- tive provision units) coordinate their pro- suming that overlapping jurisdictions necessarily im- duction activities, in whole or in part. ply duplication, the degree of duplication that exists Joint Production. Two or more provision can be subject to empirical investigation. units jointly organize a single production The argument that overlapping jurisdictions lead unit. to duplication is loosely related to a companion argument that fragmentation among a large number of Intergovernmental Contracting. A provi- relatively small jurisdictions leads to inefficiency sion unit contracts for production with an- from a failure to capture potential economies of other provision unit, which assumes respon- scale. This concept refers to a potential for decreas- sibility for organizing a production unit. ing average costs of production as the scale of pro- Private Contracting, A provision unit con- duction increases. The relationship is usually as- tracts with a private vendor, who is responsi- sumed to be curvilinear; that is, after some point in ble for organizing a production unit. increasing the size of a production unit, disecono- Franchising. A provision unit sets pro- mies of scale result. One cannot assume, for exam- duction standards and selects a private ple, that adjacent police or fire departments that en- producer, but allows individual citizen- gage in very similar administrative activities are consumers to choose whether to purchase duplicating one another in the sense of failing to the service. take advantage of economies of scale. Vouchering. A provision unit sets produc- METROPOLITAN tion standards and decides on the level of GOVERNANCE provision (through its taxing and spending The governance of a metropolitan area is a powers), but allows individuals (or groups) function that need not be directly involved with to engage different producers, public or pri- either provision or production. Functionally, govern- vate, at their discretion. ance depends on a capacity to make and enforce The potential variety in organizing both the pro- rules, not to provide or produce services. Govern- vision side and the production side, and in relating ance operates through arrangements that focus on provision to production, is much greater than is ordi- (1) a choice of rules, within which patterns of provi- narily suggested by the traditional view of local gov- sion and production emerge, and (2) the resolution ernment. Public administration theory has long as- of conflict among participants, including the mainte- sumed that the best sort of local government was a nance of agreeable and equitable relationships. full-service government-able both to make provi- The basic function of governance in a frag- sion for and to produce a full range of local public mented metropolitan area is to put together, and pe- goods and services. Small local governments, how- riodically adjust, a configuration of rules that, in ever, may regard their primary role as one of provi- turn, enables local officials and citizens to seek out sioning rather than producing, choosing to produce and create mutually beneficial provision and produc- only when in-house production, through a govern- tion arrangements, among both individuals and comment bureau or department, is the better alternative. munities. The relevant rules pertain to (1) organiza- Frequently, more than one provision unit has tion on the provision side (which depends on authority to provide service to a particular commu- collective choice capabilities, tax instruments, the nity. When jurisdictions overlap, redundant author- use of police powers in regulation, elections, and ity is usually found in some service areas. Redun- referenda, and the public accountability of officials); dancy in authority does not. however, imply (2) the organization of production, whether by pub- duplication of service. Duplication should instead be lic bureaus organized by provision units individually understood as delivery of the same service, at the or jointly or by a wide variety of other alternatives; same time, in the same place, to the same clien- and (3) relationships among both provision units and production units, including attention to joint Louis County against the background of city-county and contract arrangements, the use of overlapping separation. The concept of a "local government jurisdictions to handle benefit or cost spill-ins and constitution" is introduced and applied to four sets spill-outs, and mechanisms for conflict resolution. of rules. Included are current controversies over When conflict occurs (as for example over mu- municipal incorporation, special district formation, nicipal boundaries or tax incidence), governance ar- annexation, and sales taxation. rangements must exist to apply general rules to spe- Chapters Four through Seven analyze functional cific cases and to constrain participants to reach arrangements in four senrice areas: police, fire, settlements. If settlements cannot be reached on the streets, and education. Each analysis uses the dis- basis of existing rules, new rules may be needed. tinction between provision and production to map Fiscal disparities among provision units are a poten- the service arrangements in St. Louis County and tial source of conflict in most highly fragmented met- assess the implications of the patterns found. Chap- ropolitan areas. So, too, are benefit or cost spill-ins ter Eight is an analysis of fiscal relationships among and spill-outs, and externalities in general. Adjust- the local government units in both city and county, ments in the fiscal rules governing revenue capabili- including discussions of economies of scale, fiscal ties-in particular, the availability (and possible shar- disparity and equity, and economic development. ing) of various tax bases-are common responses to Chapter Nine concludes the volume, summariz- fiscal conflict. Appeals to the more inclusive collec- ing the -functional dimensions of metropolitan or- tive decisionmaking capabilities inherent in overlap- ganization in the St. Louis area. The research ques- ping jurisdictions are equally common responses to tions and evaluative criteria discussed above are externality problems. reviewed in order to ascertain what has been learned Metropolitan areas do seem to require some from the study. Evaluations made by others are also form of metropolitan governance, but it is a form of reviewed, alongside current reform efforts, and governance that can be exercised apart from both some conjectures are offered concerning the future provision and production of public goods and serv- of metropolitan organization in St. Louis. ices. "Metropolitan governance," in other words, is not identical to "metropolitan government."38 ENDNOTES ' See Committee for Economic Development, Modernizing CONCLUSION AND Local Government (New York: 1966). For an extended discussion of metropolitan problems drawing on traditional PROSPECTUS theory, see Advisory Commission on Intergovernmental Re- The problem of metropolitan organization in lations, Improving Urban America: A Challenge to Federal- 20th-century America has been viewed by many ob- ism (Washington, DC: ACIR, M-107, September, 1976). servers as presenting, first and foremost, a challenge *See Robert L. Bish and Vincent Ostrom, Understanding Urban Government (Washington, D.C.: American Enter- of reform. Today, propositions linking the fragmen- prise Institute, 1973). tation of jurisdictions with disorganization and inef- =See the recommendations of the Committee for Economic fectiveness can no longer be accepted as self- Development, Reshaping Urban Government (New York: evident. The basic problems of metropolitan 1970); also, Joseph F. Zimmerman, The Federated City: Community Control in Large Cities (New York: St. Mar- organization have come to pose, first and foremost, tin's Press, 1972). a challenge of understanding. For this purpose, the 'See Alan Altshuler, Community Control: The Black De- ACIR has undertaken a research program consisting mand for Participation in Large American Cities (Indian- of a series of case studies, beginning in the St. Louis apolis: Pegasus, 1970); Milton Kotler, Neighborhood: The area.30 From a better understanding of metropolitan Local Foundations of Political Life (Indianapolis: Bobbs- Merrill, 1969) ; and Howard W. Hallman, Neighborhood organization and governance, more discriminating Government in a Metropolitan Setting (Beverly Hills: Sage and, in the end, more effective efforts to adjust the Publications, 1974). structure of metropolitan areas can be developed. In OReform proposals have often been based on presupposed re- the absence of understanding, reform efforts, to the lationships between fragmentation and urban pathologies, relationships that were believed to be so self-evident as to extent they are successful, may yield a harvest of make empirical inquiry unnecessary. See, for an example in unintended consequences. the St. Louis context, the discussion in Scott Greer, "Di- In what follows, Chapter Two describes the ju- lemmas of Action Research on the Metropolitan Problem," in Morris Janowitz, ed., Community Political Systems risdictional configuration of St. Louis City and (Glencoe, Illinois: The Free Press, l96l), p. 193. See also County, with a strong focus on the county. Patterns Elinor Ostrom. "Metropolitan Reform: Propositions De- of growth, political geography, representation, and rived From Two Traditions," Social Science Quarterly 53 (December 1972), pp. 474-493, and " Righteousness, Evi- basic fiscal relationships are described, as well as a dence, Reform: The Police Story, " Urban Affairs Quarterly history of reform efforts. Chapter Three analyzes (June l97S), pp. 461-486. the legal rules and functional arrangements of met- 'See the discussion by Vincent Ostrom, The Political Theory ropolitan governance, again focusing strongly on St. of a Compound Republic: Designing the American Experi- ment (Lincoln: University of Nebraska Press, 1971, 1987), ment for goods and service such as police and education, pp. 178-190. the best one can do is to compare the performance of one 7For example, the partial consolidation of Indianapolis- institution with another. Examples of research taking this Marion County. Indiana-Unigov-which was accom- comparative performance perspective are cited in notes IS plished by act of the Indiana legislature in 1969. through 19 above. of 85 such referenda conducted between 1921 and *Vincent Ostrom, Charles M. Tiebout, and Robert Warren, "The Organization of Government in Metropolitan Areas: 1979, 80% were rejected. See ACIR, State and Local Roles in the Federal System (Washington, DC: ACIR, A-88, A Theoretical Inquiry, " American Political Science Review 55 (December 1961), pp. 831-842. See also Robert L. April l982), p. 396. Bish, The Public Economy of Metropolitan Areas (Chi- eACIR, State and Local Roles, p. 400. cago: Markham, 1971), and Bish and Ostrom, Under- loJoseph F. Zimmerman, State-Local Relations: A Partner- standing Urban Government. ship Approach (New York: Praeger, l983), p. 92. 230strom, Tiebout, and Warren, "The Organization of Gov- llFor an account of special district formation and governance ernment in Metropolitan Areas, " p. 833. in California, which, alongside Illinois, leads the nation in P4To insist on a principle of inclusion-No Taxation without number of special districts, see Robert B. Hawkins, Jr., Representation-logically implies a principle of exclusion- Self-Government by District: Myth and Reality (Stanford: No Representation without Taxation. "Regulation" may be Hoover Institution Press, 1976). substituted for "taxation" in these statements of principle. 12ACIR, Factors Affecting Voter Reactions to Governmental =The distinction between provision and production is dis- Reorganization in Metropolitan Areas, (Washington, DC: cussed more fully in ACIR, The organization of Local Pub- ACIR, M-15, May 1962). lic Economies. 13Willis D. Hawley, "On Understanding Metropolitan Politi- =This is the sort of competition usually associated with the cal Integration," in Hawley, et al, Theoretical Perspectives work of Charles M. Tiebout, "A Pure Theory of Local Ex- on Urban Politics (Englewood Cliffs, N.J.: Prentice-Hall, penditure," Journal of Political Economy 44 (October 1976), pp. 100-145. 1956), pp. 416-424. 14This body of research is summarized and reviewed in ACIR, "Paul E. Peterson, Barry G. Rabe, and Kenneth K. Wong The Organization of Local Public Economies (Washington, explain their use of a similar methodology in When Federal- DC: ACIR, A-109, December 1987). ism Works (Washington, D.C.: The Brookings Institution, 16For a summary of this research program, see Elinor 1986), p. 30. Ostrom, "Size and Performance in a Federal System," =See, for example, ACIR, Improving Urban America: A Publius 6 (Spring 1976), pp. 33-73. Challenge to Federalism. laRoger B. Parks, "Metropolitan Structure and Systemic Per- =Reynolds characterizes St. Louis as "the archtypical 'frac- formance: The Case of Police Service Delivery," in K. tionated' metropolitan area of the 'reformer', " in "Progress Hanf and T.A.J. Toonen, eds., Policy Implementation in toward Achieving Efficient and Responsive Spatial-Political Federal and Unitary States (Dordrecht, The Netherlands: Systems in Urban America," p. 466. Martinus Nijhoff Publishers, 1985), pp. 161-191. %ee John C. Bollens, ed., Exploring the Metropolitan Com- 17See Thomas J. Dilorenzo, "Economic Competition and Po- munity (Berkeley: University of California Press, 1961). litical Competition: An Empirical Note," Public Choice 40 31See Elinor Ostrom, "Size and Performance in a Federal Sys- (l976), pp. 203-209 and Mark Schneider, "Fragmentation tem. " and the Growth of Government," Public Choice 48 (1986), pp. 255-263. 320scar Newman, Community of Interest (Garden City, NY: Anchor PresslDoubleday, 1980), pp. 124-156. leMark Schneider and J.R. Logan, "Fiscal Implications of Class Segregation: Inequalities in the Distribution of Public 33A new city of Chesterfield (estimated population 33,000) Goods and Services in Suburban Municipalities. " Urban Af- began operation June 1, 1988, raising the number of mu- fairs Quarterly 17 (September 1981), pp. 23-37. nicipalities in St. Louis County to 91. laDavid L. Chicoine and Norman Walzer, Governmental 34A list of those persons who were interviewed and/or who Structure and Local Public Finance (Boston: Oel- contributed data to the study can be found at the beginning geschlager, Gunn & Hain, 1985). of this report, pp. v-vi. ?Osee the work of David R. Reynolds comparing the organiza- %See Gordon P. Whitaker, "Coproduction: Citizen Participa- tion of St. Louis, Detroit, and Los Angeles in his "Progress tion in Service Delivery," Public Administration Review 40, toward Achieving Efficient and Responsive Spatial-Political no. 3 (May-June 1980), pp. 240-246, and Roger B. Parks, Systems in Urban America, " in John S. Adams, ed., Ur- et al, "Coproduction of Public Services," in Richard C. ban Policymaking and Metropolitan Dynamics: A Com- Rich, ed., Analyzing Urban Service Distributions (Lexi- parative Geographical Analysis (Cambridge, Mass. : Ballin- ngton, Massachusetts: Lexington Books, 1982), pp. ger, 1976), pp. 463-53. 185-199. 21This is so because no system of government is perfect. It is =For a fuller discussion, see ACIR, The Organization of Lo- always possible to point to problems and shortcomings. cal Public Economies. Metropolitan reform advocates use the manifestation of 37This discussion is based on a typology of interlocal arrange- problems and shortcomings to argue for reform, but the ments developed in Elinor Ostrom, Roger B. Parks, and evaluation of institutional performance requires a compara- Gordon Whitaker, Patterns of Metropolitan Policing tive assessment of institutions. Only on this basis can we (Cambridge, Mass. : Ballinger Publishing Co., 1978), pp. know whether proposed reforms are more likely to solve 30-31. problems or make them worse. The simple existence of problems and shortcomings does not establish that a metro- *This idea is developed in ACIR, The Organization of Local politan area is not comparatively well governed. Moreover, Public Economies. there are difficulties with performance measurement in the 30A similar case study of the Pittsburgh-Allegheny County, public sector. In the absence of a cardinal scale of measure- Pennsylvania, area is underway.

Chapter Two St. Louis: A Variety of Jurisdictions

INTRODUCTION: jected to dual taxation-city taxes levied to support CITY-COUNTY HISTORY AND THE the city government and county taxes overlaid on GROWTH OF SUBURBS these to support the county government. Ninety per- The building of what was to become St. Louis cent of county residents lived within the city in the began in 1764, with the construction of a fur trading 1870s. They objected to taxes which they viewed as settlement by French pioneers.' Its construction was primarily benefiting the much smaller proportion of ordered by Pierre Laclede Liguest, a principal in the the population residing outside the city limits. In Louisiana Fur Company, and initiated by Auguste 1875, Missouri adopted a new constitution that in- Choteau, his stepson. Laclede named the settlement cluded provisions authorizing the city to: (1) extend St. Louis in honor of Louis IX, the patron saint of its boundaries westward, increasing its land area the reigning French monarch, Louis XV. Laclede's nearly three-fold, (2) frame a home rule charter, plans for St. Louis went far beyond the establish- and (3) join with the county in selecting a board of ment of a trading post for the fur company. In his freeholders whose task it would be to propose a plan words, "I intend establishing a settlement which, in to adjust relationships between the city and the the future, shall become one of the most beautiful county. The city's home rule charter and the cities in the world."2 By 1804, when the Louisiana freeholder's plan for separation of the city from the Territory became a pan of the United States, St. county were presented to voters in the county in Louis had grown to a city of some 1,000 inhabitants, August 1876. After substantial wrangling involving and was chosen as the seat of the District of St. charges of vote fraud, St. Louis City became a sepa- Louis. St. Louis was incorporated as a city under the rate entity with both city and county powers and of- rules of the Louisiana Territory in 1809 and again in fices in December of that year. It is ironic that non- 1822 within the recently admitted state of Missouri. city voters in the county were opposed to separation St. Louis County was first proclaimed as an ad- by a 3-to-1 majority. In later years, when proposals ministrative unit of the Louisiana Territory in 18 12. to reunite the city and county were presented to vot- It assumed most of its present boundaries in 18 18, ers, substantial majorities of county residents voted when Jefferson and Franklin Counties became inde- in opposition. pendent units. Until 1876, the City of St. Louis was At the time of separation, the St. Louis area had an integral part of St. Louis County, and was gov- seven units of local government-the City of St. erned along with the rest of the county by a county Louis, St. Louis County, and five incorporated mu- court, later to become a St. Louis County Board of nicipalities in the county. By 19 10, four more mu- Commissioners. nicipalities had been incorporated. At that time, five As early as 1840, citizens of the City of St. of the six municipalities with populations presently Louis proposed the separation of the city from greater than 20,000 were in place. In the next two St. Louis County. As is common in America, the decades, ending in 1930, ten more municipalities dispute arose over taxation. Residents of the city ob- were incorporated in the county, yielding a total of 73 Sunset Hills 74 Sycamore Hills 75 Town & Country 76 Twin Oaks 77 University Ci 78 Uplands Park 79 Valley Park 80 Velda Village 81 Velda Village Hills 82 Vinita Park 83 Vinita Terrace 84 Warson Woods 85 Webster Groves 86 Wellston 87 Westwood 88 Wilbur Park 89 Winchester 90 Woodson Terrace

SOURCE: St. Louis County, Missouri, Fact Book- 1986. St. Louis County Department of Planning.

- 16 - 19. A wave of municipal incorporation, however, Louis County reached a high of 98 in 1959. Since was about to begin. then, nine mergers of small municipalities and one In the two decades spanning 1931 to 1950, disincorporation (Times Beach), together with two more than two-thirds of the present set of munici- incorporations, resulted in the 90 municipalities palities in St. Louis County were incorporated, with existing when this study was done. Fifteen years half of these formed in the five years after the end elapsed between the incorporation of Black Jack in of World War 11. Virtually all of these municipalities 1970 and Maryland Heights in 1985. As the re- began as subdivisions located in unincorporated search for this study was under way, incorporation parts of the county. The incorporation of these sub- efforts were progressing in the Chesterfield area, cul- divisions into villages and small cities ensured their minating in a new incorporation approved on April citizens a right to choose their own officials and to 5, 1988, raising the number of county municipalities participate in collective decisions regarding local to 91. All of the analysis in this report, however, is services, taxes, planning, zoning, building codes, based on a count of 90 municipalities, or as other- and other aspects of community life. Faced with al- wise indicated. ternatives of leaving these decisions to county officials or of being annexed by other, adjacent municipalities,3 citizens of much of St. Louis County chose BASIC POLITICAL GEOGRAPHY local self-government. "St. Louis" is composed of 92 general purpose This pattern of subdivision incorporation was local jurisdictions. The City of St. Louis and the un- most prominent in the Normandy area, extending incorporated portion of St. Louis County are both out along Natural Bridge and St. Charles Rock roads quite large. Together they include 58 percent of the in the northwest corridor of the county, and in areas total St. Louis population-30 percent in the city adjacent to Normandy. Of the 62 present-day mu- and 28 percent in the unincorporated county.4 nicipalities that incorporated during this 20-year pe- Eleven percent of the St. Louis land area lies in the riod, more than 60 percent lie in this corridor or city and 52 percent in unincorporated parts of the immediately adjacent to it. Many small municipali- county. The remaining 90 general purpose units of ties in this area maintain to this day the subdivision local government are, for the most part, small. Only organizations that predated incorporation, while oth- 6 have populations exceeding 20,000 persons, and ers have included preexisting subdivision rules re- 54 have fewer than 5,000 residents. Figure 2.1 de- garding housing styles and other zoning provisions in picts the municipal geography of St. Louis, illustrat- their municipal ordinances. ing the small areal extent of most of these municipal The pattern of suburbanization during this pe- governments. Table 2.1 describes the distribution of riod was not exclusively that of municipal incorpora- local governments in population and land area. tion and/or annexation, however. At the same time Table 2.2 presents data on the distributions of that the northwest corridor and some other parts of race, poverty, and income in St. Louis City and the county were incorporating as separate munici- County. For the municipalities in the county, aver- palities, much of South County was also growing rap- age values for each population quintile are com- idly, but without significant incorporation. Five very puted to illustrate municipal variations.5 small municipalities were created in this area be- The 1980 Census reported 22 percent of the St. tween 1940 and 1950, but most of the population Louis City and County population to be minority, chose to remain a part of the unincorporated with blacks comprising virtually all of the minority county. Two parts of South County, Affton and population. Approximately two-thirds of the area's Lemay, achieved sufficient identity to become rec- black population resides in the City of St. Louis. Of ognized as urban places by the U.S. Bureau of the the blacks who live in St. Louis County, 85 percent Census, yet they remain unincorporated today. are residents of municipalities. About two-thirds of Other suburbanized parts of South County, Oakville the county's black population is located in the Nor- and Mehlville, for example, also have quite clear mandy area and in a few municipalities immediately identities in socio-spatial terms, but have chosen, adjacent to that area.6 nevertheless, to remain unincorporated. The resi- Housing patterns by race in St. Louis demon- dents of these areas did choose, however, to retain strate continuing segregation of blacks from whites. small local school districts and to create small fire St. Louis City's population is approximately one-half protection districts. black, but racial mixing in housing is limited. Most Ten more municipalities incorporated from of the city's black population resides in predomi- 195 1 to 1986, when the field work for this study was nantly black neighborhoods in the northern half of conducted. Eight of these incorporations occurred the city, while the southern half is predominantly prior to 1960. The number of municipalities in St. white. Racial integration of residential areas is Table 2.1 General Purpose Local Governments in St. Louis Population Percent of Land Area Percent of Number (1 984 estimatel) Total (square miles) Total Municipalities in St. Louis County 902 589,996 42.0% 208 37.0% With Population of: 0-1,000 232 10,838 0.8 5 0.9 1,001-2,000 14 22,944 1.6 4 0.7 2,001-5,000 17 56,718 4.0 42 7.5 5,001-10,000 17 112,720 8.0 37 6.6 10,001-20,000 13 186,409 13.0 62 11 .O 20,001-56,000 6 200,366 14.0 58 10.0 Unincorporated St. Louis County 1 397,213 28.0 29 1 52.0 Total St. Louis County 912 987,209 70.0 499 89.0 City of St. Louis 1 429,300 30.0 61 11 .O Total City and County 922 1,416,508 560 '1984 population estimates for the County are from MissouriPopulation and Census Newsletter, December 1985, cited in 1985 Fact Sheet: PoliceSetvices in St. Louis County. Bureau of Management Services. St. Louis County Deparbnent of Police. A few adjustments were made to reflect recent annexations. using data from the 1986 St. Louis Counfy Fact Book, St. Louis County Department of Plan- ning. The estimate of population for the City of St. Louis is taken from U.S. Department of Commerce. Bureau of the Census, Local Government Finances in Major County Areas, 1984-85, GF 85-No. 6. 21ncludes the City of Pacific which lies primarily in Franklin County. Pacific has 12 residents in St. Louis County. somewhat more common in St. Louis County. If one county, and about 30 percent of the county's black accepts a population mix of 30 to 70 percent black population. Seventy-three percent of black residents or white as an integrated community, nine county in the county live in communities that are less than municipalities are integrated. The blacks living in 50 percent black.' these integrated communities represent about 10 Separation by poverty status and income classes percent of the black population of the city and is also found in St. Louis, although the patterns are somewhat less pronounced than those for race. Ap- Table 2.2 proximately 10 percent of the area's population had Distribution of Race, Poverty, and incomes below the poverty level in 1979. Two-thirds Income in St. Louis of these persons resided in St. Louis City, which has Percent a poverty rate of 22 percent. In St. Louis County, 72 Munlclpalities Below Average percent of the poverty population resided in munici- in St. Louis Percent Poverty Household palities, with the remainder in the unincorporated County Minority - 1979 Income county. Approximately half of the municipal poverty Quintile 11 1.5% 2.3% $1 6,972 population was found in the Normandy area and im- Quintile 2 2.8 3.0 21,039 mediately adjoining jurisdictions. Quintile 3 6.5 4.5 23,684 Average household incomes in St. Louis County Quintile 4 17.4 6.6 26,178 were 75 percent higher, on average, than those in Quintile 5 56.2 14.0 40,989 St. Louis City. Average income in the unincor- Municipal Average 16.7 6.0 25,218 porated county was slightly higher than the average Unincorporated in municipalities. The range of average incomes St. Louis County 4.1 3.1 26,969 across municipalities is relatively flat (see Figure Total 2.2), with 60 percent of the municipal population St. Louis County 11.3 4.8 25,899 residing in communities with average incomes be- St. Louis City 46.4 21.8 14,723 tween $20,000and $30,000.About 4 percent of the Total City and municipal population lives in communities where the County 22.4 10.2 22,038 average income is below $15,000,3 percent in com- Quintites display average values for each 20 percent of the to- munities where it exceeds $50,000. tal municipal population in the county, ordered from lowest to highest on each data series. One aspect of the population distribution warrants particular attention. While nearly 60 percent Figure 2.2 Distribution of Average Household lncome in County Municipalities

90000 80000 Average 70000 Household 60()00 Income 50000 (thousands) 40000 30000 __C loo00

0 20 40 60 80 100 Percent of Municipal Population of the local governments in St. Louis City and The 90 St. Louis County municipalities are or- County have populations below 5,000, these com- ganized variously as villages, third and fourth class munities in total include less than 7 percent of the cities, and home rule cities. The 26 villages are all total St. Louis population, including both city and quite small-only one exceeds 5,000 population (Ta- county. In addition to the 58 percent of the St. ble 2.3). In total, villages contain approximately 6 Louis population found in the local jurisdictions of percent of the population of county municipalities. the two largest units, another 14 percent reside in The 45 fourth class cities, with 36 percent of the the six municipalities with populations exceeding county's municipal population, have populations 20,000. Thus, while small local governments are generally below 10,000. Third class cities tend to be common in St. Louis, most people in St. Louis City somewhat larger and the 13 home rule municipalities and County reside and work in the larger govern- are the largest in the county. Third class cities con- mental jurisdictions.8 Roughly 40 percent of the tain about 12 percent and home rule municipalities county population, however, resides in municipali- about 46 percent of the population of county mu- ties with fewer than 20,000 people. nicipalities.

Table 2.3 Municipalities in St. Louis County Number Population Percent of Land Area Percent of (1 984 estimate) Total (square miles) Total Villages 26 33,419 6% 11 5% (31-5,696)' Fourth Class Cities 45 21 1,314 36 87 42 (360-1 9,049) Third Class Cities 6 69,786 12 34 16 (4,447-26,413) Home Rule Cities 13 275,476 46 76 37 (7,875-55,949)

Point-of-Sale Municipalities 38 308,038 52 (67-27,990) Sales Tax Pool 52 281,957 48 (31 -55,949) Range of municipal populations. Table 2.4 Organization of Fire Protection in St. Louis St. Louis County Fire Departments and Population Percent of Land Area Percent of Fire Protection Districts (1 984 estimate) Total (square miles) Total Municipal Fire Departments (19) 276,655 20% 76 13% (4,1 12-49,929)' Fire Protection Districts (24) 71 0,553 49 423 76 (4,000-96,500)1 Total St. Louis County (42) 987,208 69 499 89 City of St. Louis (1) 453,085 3 1 61 11 'Range of populations served, including contracts.

OVERLAPPING JURISDICTIONS AND county's population, however, receive fire protec- SPECIAL PURPOSE GOVERNMENTS tion services from 24 independent fire protection In addition to the general purpose units of local districts (Table 2.4). Rather than making provision government in St. Louis, a variety of special districts arrangements for fire protection through their gen- add an array of overlapping jurisdictions to the di- eral purpose local governments, residents of the un- verse pattern of municipalization. Most numerous incorporated county and 62 incorporated munici- are school districts and fire protection districts. Two palities provide for fire services through districts related to education are countywide and two independently organized special districts governed others-a sewer district and a museum-zoo district- by elected trustees.9 embrace both St. Louis City and County. These separately organized school districts and Elementary and secondary education is organ- fire protection districts add another 49 collective ized by 23 school districts in St. Louis County and service provision units to the 92 general purpose one in St. Louis City. The county districts range in governments in St. Louis City and County-mostly in size from a few hundred to some 20,000 students. the county. There are, of course, numerous other The larger districts are products of reorganization governmental and quasi-governmental units in St. and consolidation. St. Louis City has a separately Louis. Areawide special districts provide for sewer organized school district that, with an average daily treatment and cultural facilities. A recently com- attendance in excess of 50,000 students, is more pleted agreement between St. Louis City and County than twice the size of the largest county district, the makes public health services a cooperative venture. Parkway Consolidated School District. The St. Louis There are also a few small districts for purposes such City district and four of the county districts have as roads and street lighting. The number is much boundaries that are coterminous or nearly so with greater, however, if more than 400 subdivision or the boundaries of municipalities. The remaining 19 homeowners' associations that provide for residen- districts in the county cut across municipal bounda- tial street services are included (see Chapter Six). ries and portions of the unincorporated county. All When these units are added to the large number school districts, governed separately by elected of municipalities, the characterization of St. Louis as boards, are autonomous governmental jurisdictions. fragmented-in the sense of having a very large Overlaid on the county districts are the county- number of organized public jurisdictions-appears to wide Special School District, which was established be quite accurate. Fragmentation is accompanied by to provide education for handicapped students and, an extensive use of overlapping jurisdictions. In later, vocational-technical education, and the Com- some portions of incorporated St. Louis County, a munity College district. St. Louis City obtains special citizen may be served by the following governmental education through its regular school district. or quasi-governmental units: Fire services in much of St. Louis County are 1) Subdivision to provide streets and other also organized in part by special districts that are in- neighborhood amenities; dependent of general purpose units of local government. St. Louis City has its own fire department, as 2) Municipality (village or city); do 19 municipalities in the county. Eight county mu- 3) School district; nicipalities arrange for fire services through contracts 4) Fire protection district; and with municipal or special district fire departments. More than 70 percent (710,000 residents) of the 5) County government. This count does not include countywide and district, are members of the Cooperating School Dis- city-county special districts. In other portions of the tricts of the St. Louis Suburban Area (CSD), an or- county, the functions of a subdivision and fire dis- ganization that supplies its members with support trict may be performed by a municipality. The pat- services in the areas of audiovisual equipment and tern is complex and far from uniform. As will be media and joint purchasing, and that articulates seen in the discussions of service delivery in Chap- members' concerns to the state legislature. Most of ters Four through Seven, however, this large number the districts are also members of the Regional Con- and variety of units do not necessarily fragment the sortium for Education and Technology (RCET) . The area in the pejorative sense of that term. Jurisdic- St. Louis City school district and 16 of the county tional fragmentation does not necessarily lead to school districts are involved in an additional coop- functional fragmentation. erative effort, the Voluntary Interdistrict Coordinat- ing Council (VICC). The council, created in re- MULTIORGANIZATIONAL ARRANGEMENTS sponse to litigation charging racial discrimination in Cooperative endeavors among the extensive ar- the city schools and between the city and county ray of local governments in St. Louis County have school districts, manages an extensive program of created a multijurisdictional overlay of functional ar- cross-district movement of students. Black students rangements that span the boundaries of separate from the city are able to choose among the county municipalities, school districts, and fire districts. The districts that are a party to the agreement, while organization of St. Louis cannot be fully and accu- white students from the county districts are able to rately described without reference to these multi- attend any of the various "magnet" schools in the organizational undertakings. If the large number of city district. (These cooperative agreements are dis- separate jurisdictions described earlier tends to frag- cussed more fully in Chapter Seven.) ment the metropolitan area, the multiorganizational overlay created by officials and citizens of these Local Government Associations same jurisdictions tends to integrate the area, tying Local governments and local officials in St. separate units together in numerous common ef- Louis County have also formed associations that forts. These include administrative and contractual serve as connecting tissue among jurisdictions- arrangements for producing services, as well as po- forums for sharing information, resolving differ- litical forums for working out differences and seek- ences, and developing solutions to common prob- ing changes in state law affecting the region. lems (obtaining state legislation when necessary). The St. Louis County Municipal League, founded in Joint Administrative Arrangements 1918,lO includes among its members all but a very Municipalities in the county cooperate exten- few of the smallest municipalities in the county (see sively with one another for administrative purposes, discussion in Chapter Three). The CSD (discussed especially in the production of police services. Many above) represents all county school districts. In of the smaller municipalities have created joint dis- North County, a cluster of small municipalities have patch centers, for example. There is also extensive organized a forum called the Normandy Municipal cooperation between municipal and county police. Council with a small staff that provides its 16 mu- (Police services are discussed in detail in Chapter nicipal members with representation to county and Four.) Extensive intergovernmental cooperation also state governments, as well as with planning and characterizes the delivery of fire protection services. other support services. Virtually every municipal and district fire depart- Both elected officials and department heads in ment has signed formal mutual aid agreements with county municipalities have created additional forums every other department. These agreements, for the through which they are able to meet and discuss most part, include first-response provisions, whereby common problems. The Mayors of Large Cities is the nearest fire company responds to a fire, regard- one such group originally formed to meet with and less of jurisdictional boundaries. Most of the fire advise the county executive. Another group, the protection districts and several of the municipal fire Mayors of Small Cities, recently organized to pro- departments have joined together to create fire vide a collective voice for officials not regularly con- alarm districts as well. These districts provide fire sulted by county government. Administrators and and emergency medical dispatch services for their bureau chiefs of the municipalities have also created members, thus facilitating the provision of mutual forums for discussion and initiative on issues extend- aid among them. (The provision of fire protection ing beyond the boundaries of their individual juris- services is discussed in Chapter Five.) dictions. Active groups include the Board of Gover- All of the county's school districts, together with nors of the Law Enforcement Officials of the districts in nearby counties and the St. Louis City Greater St. Louis Area (see Chapter Four), the [22Greater St. Louis Fire Chiefs Associaiion (see per elected official. Overall, the residents of county Chapter Five), and the local chapter of the Ameri- municipalities have relatively high levels of per cap- can Public Works Association (see Chapter Six). ita representation compared to the area's two largest City administrators and managers of a number of the jurisdictions. The ratios are much larger in both St. municipalities have joined together to create an in- Louis City and St. Louis County, with more than surance pool covering workmen's compensation and 16,000 citizens per alderman in the city and ap- municipal liability for their communities. proximately 140,000 citizens per council member in At the city-county level, multijurisdictional fo- the county. rums are more limited. Two quasi-governmental In interviews, many local officials pointed with agencies have areawide purview, but are limited in pride to these low citizen-elected representative ra- their abilities to take action. The East-West Gateway tios, usually stating that local officials have to be re- Coordinating Council, formed in 1965 as a "Metro- sponsive to individual citizens in the community be- politan Planning Organization" required by federal cause each citizen's vote weighs heavily in local law to qualify the area for continued federal highway elections. The contrast with county government is funds, conducts an extensive array of statistical dramatic. Seven elected council members and one analyses on an areawide basis." The council pro- elected executive represent 987,209 county resi- vides a forum in which local government officials dents. County council members, with constituencies from throughout the bistate (Missouri and Illinois) averaging in excess of 140,000 persons, are, munici- area can discuss regional issues, but has no inde- pal officials argued, unable to respond individually pendent authority or revenues to act with respect to to their constituents and can, in fact, pay less atten- areawide projects. A second organization, the Bi- tion to many citizen requests for council action be- State Development Agency, was chartered in 1949 cause individual citizens are less consequential for with broad authority to act on issues of concern to their electoral success.12 the area as a whole, but lacks independent taxing Mayors and trustee chairmen in county munici- authority. Functionally, Bi-State has become the op- palities argued that local bureau heads (such as po- erator of an areawide transit system, and is depend- lice and fire chiefs), too, have to be more responsive ent for this and any other functions it might wish to undertake on voluntary contributions from local, to individual citizens. These bureau chiefs are, they general purpose units of government. In addition to said, exposed to council supervision and to individ- these quasi-governmental bodies, local citizen or- ual citizen requests to a much greater extent than ganizations, such as Confluence St. Louis, provide their counterparts in the two largest political jurisdic- forums for discussion of metropolitan issues, but are tions. One reason that citizens of St. Louis County limited to the issuance of advisory reports. cling to their relatively small local governments, even in the face of repeated criticism from reform advo- PATTERNS OF cates and periodic proposals for consolidation, may POLITICAL REPRESENTATION well be the extraordinary opportunities afforded them for low-cost participation in collective choices The legal classification of a municipality determines the pattern of representation afforded its citi- regarding taxation, zoning and building codes, pub- zens. Villages elect five or, in the largest, nine trus- lic service levels and costs, and other aspects of local tees. In all but two villages, the ratio of citizens to community life. Citizens who are active in municipal elected officials is less than 500. Fourth class cities- incorporation efforts in presently unincorporated generally the next largest in size to villages-elect a parts of the county often state their case in these mayor and four, six, eight, or, in one case, ten al- same terms, focusing on one or another aspect of dermen. Seventy percent of the fourth class cities community life that citizens of a specific community have citizen-elected official ratios of less than 1,000, want to be able to shape for themselves. never exceeding 3,000. The six third class cities, The existence of a number of very small munici- with populations somewhat larger than the average palities, especially in the Normandy area in North city of the fourth class, elect a mayor and four, six, County, highlights another aspect of representation or eight councilmen. Citizen-to-elected-official ratios for residents in fragmented areas. The Normandy in these six cities are also small, with only one city's area is the home of most of the county's nonwhite, ratio exceeding 3,000. The 13 home rule charter cit- mostly black, population. Many of the elected offi- ies in the county, generally the largest of the munici- cials, bureau chiefs, and public employees in these palities, elect a mayor and four, six, eight, or nine communities are black. These small communities councilmen. Their citizen-elected official ratios are give black citizens access to political participation also small, with only two cities-Florissant and Uni- and representation that they have yet to achieve in versity City-having ratios that exceed 5,000 citizens county government. Many of these communities, which include a variety of income levels and life municipal population) have city administrators.j4 styles, are racially integrated. These administrators, together with others desig- In addition to participation in collective choices, nated administrative assistant, assistant to the citizens in many municipalities are active in the pro- mayor, city clerk, and village clerk have, over the duction of local services for themselves. This pro- years, developed significant professional skills in duction ranges from active participation on local managing the day-to-day affairs of their communi- boards and commissions to physical production of ties. Working in tandem with elected officials, they beautification and maintenance services and police are able to provide continuity in local government management for their municipalities. services through such mechanisms as neighborhood watches and citizen patrols. One can hypothesize FISCAL PAlTERNS that the relatively high level of citizen involvement in General purpose units of local government in St. local service delivery results, at least in part, from Louis City and County raised $527 million in reve- the sense of community and individual efficacy af- nues in 1985 (Table 2.5). In addition, $34 million in forded them by small local jurisdictions. revenue was raised by the fire protection districts in The men and women who serve as mayors the county. Of these revenues, 44.5 percent accrued (chairmen of trustees in villages), aldermen, coun- to St. Louis City, with the remainder distributed cilmen, or trustees of the county municipalities are, among St. Louis County government, county mu- for the most part, employed in other full-time jobs nicipalities, and fire protection districts.15 or are retired. Only one municipality, Florissant, has In St. Louis County, revenues raised in the mu- a full-time mayor. Legislative officials come together nicipal jurisdictions vary with the class of municipal- once or twice a month in most of the municipalities ity and status with respect to sales tax distribution to conduct local business. Mayors and village trustee (Table 2.5). Villages and the fire protection districts chairmen devote somewhat more time to their du- that serve them accounted for 3.7 percent of total ties, usually attending to the duties of their offices at local government revenues in the county.le More regular times during each week.13 Many of the than 40 percent of total local government revenues smaller municipalities have no city hall or other in the county was raised within the borders of fourth building devoted exclusively to city business. Rather, class cities, 6.5 percent in third class cities,'7 and they rent office space in existing structures or use almost 50 percent in home rule cities. About two- space in recreation centers, police or fire stations, or thirds of municipal and fire district revenues was schools. The part-time status of most elected offi- raised in municipalities that receive sales tax reve- cials has led some critics to characterize them as nues according to point-of-sale (which included 55 "amateur" politicians. percent of the municipal population), while the re- To refer to these officials as amateurs simply be- maining third was raised in municipalities that pool cause of their part-time status, however, is an over- their sales tax revenues with the unincorporated simplification. For the most part, the elected offi- county. cials who were interviewed in the course of this research demonstrated that they are quite knowl- Revenue Sources edgeable about the affairs of their communities. The largest single source of revenues for St. Many have served for a number of years in their Louis County government and for the municipalities positions and have developed high levels of skill at in the county is a countywide sales tax. A sales tax is managing their respective municipality's business. also an important source of revenues for St. Louis Given the small size of many of the municipalities in City, but is outweighed there by an earnings tax col- the county, it is not surprising that part-time elected lected from city residents (wherever they are em- officials are able to function as effective managers. ployed) and from nonresidents who are employed These officials have the advantage of knowing many within the city limits. Sales tax revenues in excess of of their constituents personally, an advantage not as $27 million represented nearly 20 percent of total readily available to officials in the largest county mu- county revenues (including county government and nicipalities, or in St. Louis City and County. fire protection districts serving the unincorporated In addition to their elected officials, almost all county), while the $61 million accruing to munici- of the municipalities have at least one full-time ad- palities (and the fire protection districts serving some ministrator, ranging from a village clerk in the of them) represented nearly one-third of their reve- smaller jurisdictions through city administrators and nues. St. Louis City's $30 million raised from the managers in the larger. Nine municipalities, encom- sales tax was about 13 percent of its 1985 reve- passing about one-fourth of the municipal popula- nues.'8 tion, employ city managers, and another 14 munici- The local revenue source most directly subject palities (encompassing another one-fourth of the to control by citizens is the property tax on real, per- Table 2.5 Revenues of General Purpose Governments and Fire Districts- 1985 (in-thousands of dollars) General Fire Combined Population Purpose Districts 1 Revenues (1 984 estimate) All Units2 3 $527,231 $34,284 $561,515 1,392,466 (per capita) (378) (25) (403) City of St. Louis 234,059 - 234,059 429,300 (per capita) (545) - (5451 Percent of total 44.4% - 41.7% 30.8% Total St. Louis County2 293,172 34,284 327,456 963,166 (per capita) (304) (36) (344 Percent of total 55.6% 100.00% 58.3% 69.2% St. Louis County Government3 118,453/76,2634 22,005 140,458/98,2684 397,21P (per capita)4 (192) (55) (247) Percent of total 22.4% 64.2% 25.0°h 28.5% Municipalities in County2 174,719 12,280 186,999 565,953 (per capita) (309) (22) (330) Percent of total 33.2% 35.8% 33.3% 40.6% 'Fire district revenues are those from the property taxonly. In the municipalitiestheserevenues are the product of the firedistricttax rate and the total assessed valuation of the municipality. Fire district revenues in the unincorporated county are countywidefire district tax revenues minus those raised in the municipalities. *Does not include Maryland Heights or Velda Village Hills. 3For revenues excluded from the city and county totals, see endnote 15. 4Unincorporated portion only, revenues estimated from St. Louis County, Missouri, 1985 Annual Budget, using methodology employed in 'St. Louis County Revenue History. 1974-1983'"a briefing paper prepared for the County Annexation Study Commission in 1985. - - Table 2.6 Revenues of Municipal Governments and Fire Districts in St. Louis County- 1985 (in thousands of dollars) General Fire Combined Population Purpose District1 Revenues (1 984 estimate) All Units2 $174,719 $12,280 $186,999 (per capita) ($309) ($22) ($330) Villages3 $5,602 $1,286 $6,888 (per capita) ($175) ($40) ($215) Percent of total 3.2 10.5 3.7 Fourth Class $69,621 $5,875 $75,496 (per capita) ($324) ($27) ($351 Percent of total 39.8 47.8 40.4 Third Class4 $1 1,682 $485 $12,167 (per capita) ($269) ($11) ($280) Percent of total 6.7 3.9 6.5 Home Rule $87,814 $4,633 $92,448 (per capita) ($319) ($17) ($33'3) Percent of total 50.3 37.7 49.4 Point-of-Sale Cities $1 19,552 $7,408 $126,960 (Per capita) ($383) ($24) ($407) Percent of total 68.4 60.3 67.9 Pool Cities2 $55,167 $4,872 $60,039 (per capita) ($217) ($19) ($236) Percent of total 31.6 39.7 32.1 Firedistrictrevenues inthe municipalities are the product of the fire districttax rate and the total assessed valuation ofthe municipality. 2Does not include Maryland Heights or Velda Village Hills. 3Does not include Velda Village Hills. 4Does not include Maryland Heights.

- 24 - Table 2.7 Revenue Sources in County Municipalltiesl (in thousands) Sales Tax Municipal Fire District Ucenses, Permits, Other Property Tax Property Tax and Utilities Sources All Units2 $60,616 $18,671 $1 2,280 $49,547 Percent of total revenues 32.4% 10.0% 6.6% 26.5% Villages 2,558 442 1,286 2,171 Percent of total revenues 37.1% 6.4% 18.7% 31 5% Fourth Class 28,511 5,776 5,875 18,388 Percent of total revenues 37.8% 7.6% 7.8% 24.4% Third Class 3,751 1,708 485 3,076 Percent of total revenues 30.8% 14.0% 4.0% 25.3% Home Rule 25,796 10,745 4,633 25,914 Percent of total revenues 27.9% 11.6% 5.0% 28.0% Point of Sale 43,885 12,052 7,408 31,682 (per capita) (141) (39) (24) (1 02) Percent of total revenues 34.6% 9.5% 5.8% 25.0% Pool 16,731 6,619 4,872 17,865 (per capita) (66) (26) (1 9) (70) Percent of total revenues 27.9% 11 .O% 8.1 % 29.8% Including revenues collected in municipalities by fire protection districts, where relevant. 2Does not include Maryland Heights or Velda Village Hills. sonal, and other property. In years past, this was a Countywide, this cluster provided approximately 29 primary source of revenue for municipalities in St. percent of municipal revenues and 25 percent of to- Louis and elsewhere. In recent years, taxpayer "re- tal revenues including fire protection districts. Vil- volts," forcing reliance on alternative revenue lages in St. Louis County are permitted to tax only sources, have reduced the property tax to a much electric utility gross receipts at a 2 percent rate, how- smaller component of local revenues.lQ Within the ever, and this tax and the license and permit fees municipalities of St. Louis County, local property that they can impose supplied only about 6 percent taxes (imposed by municipalities and fire protection of total revenues raised in their jurisdictions. St. districts) provided an average of 16.6 percent of to- Louis City derived 20 percent of its revenues from tal municipal and fire district revenues in 1985 (Ta- this combined source, while the county government ble 2.7). The range of variation was broad, however. obtained 22 percent. Eight municipalities had no municipal property tax Other sources of revenue for general purpose lo- on real property. Two-thirds of the municipalities cal governments in St. Louis include sales taxes on raised less than 10 percent of their municipal reve- cigarettes and motor fuel, a property tax surcharge nues from the property tax, and only four munici- on commercial and industrial property implemented palities raised more than 30 percent from this in 1985, the road and bridge tax (essentially an add- source. These four municipalities are upper-income on to the property tax that is distributed among local communities composed almost exclusively of high- jurisdictions in proportion to assessed valuation), value residential property. Home rule and third class and intergovernmental revenues, principally general cities were, on average, somewhat more reliant on revenue sharing (until it was terminated in 1986) the local property tax for municipal revenues than and community development block grants.2' Federal villages and fourth class cities, but this is primarily revenues have become increasingly less significant an artifact of the way that fire protection services are sources for most municipalities in recent years.22 financed in the county.20 St. Louis City was well be- Among the municipalities in St. Louis County, low the average in its reliance on the local property the reliance on various sources of revenues varies tax, while the county government was somewhat with class of municipality and status with respect to above average because it collects this tax county- sales tax distribution (Table 2.7). Villages and wide. fourth class cities, generally the smaller units, rely Another cluster of locally generated revenues more heavily on the sales tax than do their larger are derived from license, franchise, and permit fees neighbors. Villages and fourth class cities are distin- and a tax on the gross receipts of public utilities. guished from each other, however, by their relative reliance on property taxes and revenues from li- county government, municipalities, and fire districts censes, permits, and the utilities gross receipts tax. in the county were 58 percent of the combined total, Villages are limited with respect to the gross receipts while the county contained 69 percent of the com- tax, and so this source provides much less of their bined population. On a countywide average, total revenue than it does for municipalities of the other per capita revenues of the county government, mu- classes. Villages do not have their own fire depart- nicipalities, and fire protection districts were $340, ments and, with a few exceptions, do not contract approximately 62 percent of those found in the for fire protection services. Fire services for most vil- city." On a per capita basis, the variation in reve- lages are supported by the property taxes of the fire nues raised by municipal governments in the county protection districts that serve them. The result is is from slightly below $100 to nearly $1,700. With that, while property taxes are a relatively small per- fire district revenues allocated to the territory of centage of the revenues raised directly by villages, these governmental units where relevant, the range when fire district property taxes are included, vil- is from $121 to $1,901, a 1-to-17 range. lages have the highest percentage of total revenues Stating total revenues on a per capita basis for raised from the property tax. The larger, third class each municipality, with residential population as the and home rule cities rely less on the sales tax than base, exaggerates the actual revenue variation to do their smaller neighbors, and somewhat more on some extent. First, the combined population of mu- license, permit, and gross receipts taxes. nicipalities at the very low and the very high ends of County municipalities differ in how they obtain the distribution is a small proportion of the total sales tax revenues. Some municipalities are point-of- population residing in municipalities. Five small sale, thereby obtaining all the sales tax revenue from communities containing 2.4 percent of the total mu- sales within their borders, while others are part of a nicipal population lie at the low end of the range, sales tax pool that includes the unincorporated por- while three communities, also with about 2.4 per- tion of the county, with sales tax revenues distrib- cent of the total municipal population, are found at uted according to population. Point-of-sale cities are the high end. If the revenues of these eight commu- more reliant on the sales tax and less reliant on the nities are excluded, the range of variation in per property tax, when measured as a percent of total capita revenues, including fire district revenues, for revenues, than pool cities. In point-of-sale cities, on 95 percent of the municipal population is reduced average, 35 percent of total revenues are derived from 17 to 1 to 4.5 to 1, still a significant range, but from the sales tax, compared to 28 percent in pool much less striking. Second, using residential popula- municipalities. In point-of-sale cities, about 15 per- tion as a base does not account for differences be- cent of total municipal and fire district revenues tween daytime and night-time populations. In county come from property taxes. The comparable figure in municipalities where large commercial or industrial pool cities is 19 percent. These percentages conceal enterprises are located, this difference (like that an interesting additional difference, however. Per found in St. Louis City) may help to explain their capita municipal and fire district property tax reve- significantly higher per capita revenues.24 nues raised in point-of-sale cities were approxi- Data presented in Table 2.8 provide some mately $63 in 1985, but only about $45 in pool cit- insight on these variations. Municipalities were or- ies. Although point-of-sale cities appear to be using dered by total revenues (municipal and fire service) sales tax revenues to offset property taxes when per capita, and then grouped into rough quartiles of compared to pool cities on the percentage of total the population in the incorporated county. Thirty- revenues raised this way, per capita property tax col- two municipalities with total revenues per capita be- lections in point-of-sale cities are larger than those low $224 contain the least advantaged 25 percent of found in pool cities. Indeed, point-of-sale cities raise county residents in per capita revenue terms. In to- more revenues from all sources than do pool cities. tal, the revenues of these municipalities and the fire Excluding revenues from the sales tax, total per cap- districts that serve them constituted 15 percent of ita revenues raised in point-of-sale cities in 1985 total municipal and fire district revenues. At the were $266, compared to $170 in pool cities. other end of this scale are 18 municipalities with per capita revenues in excess of $423. These 18 munici- Patterns of Revenue Variation palities, also containing about 25 percent of the mu- Per capita revenues in St. Louis City are sub- nicipal population, account for 39 percent of total stantially higher than those in most of St. Louis municipal and fire district revenues, and are clearly County. The city, with 31 percent of the estimated most advantaged in this sense. city-county population, raised 42 percent of the total One significant factor separating more and less revenues in 1985, for a per capita value of $545 (see advantaged municipalities in the county, when ad- Table 2.5 above). The total revenues raised by vantage is measured in terms of per capita revenues, is their status with respect to the sales tax distribu- City residents favored the amendment, but county tion formula. Among the 32 least advantaged mu- residents opposed it, and the amendment failed to nicipalities, only one is a point-of-sale city., On the receive a majority vote in a statewide referendum. other hand, all but one of the 18 most advantaged In 1954 the first successful reform initiative created cities have point-of-sale status. As the last 2 lines of the Metropolitan St. Louis Sewer District, thus inte- Table 2.8 demonstrate, point-of-sale cities tend to grating the sewer systems of the city and of urban- have substantially higher revenues per capita than ized portions of St. Louis County. City and county do pool cities, on average about 72 percent greater. residents supported the creation of this special district by a margin of 3 to 1. In the next year, how- GOVERNMENTAL ever, city and county residents rejected the creation REFORM MOVEMENTS of a similar special district for the provision of transit Throughout its history, St. Louis has been the services, and in 1959 rejected the creation of a object of numerous studies and proposals aimed at multipurpose special district for the provision of changing the structure of local government in the sanitation, roads, transit, economic development, area. The first reform was the aforementioned sepa- police training and communications, and civil de- ration of St. Louis City from St. Louis County in fense. 1876. Since that time, reform proposals have fre- Proposals for governmental reform in St. Louis, quently been advanced to undo this separation. Ad- as elsewhere in America, have been similar to those ditional proposals have focused on reducing the discussed in Chapter One. The number of local gov- number of governmental units in St. Louis County. ernments in the St. Louis area has been taken as Some reform proposals have been successful, nota- obvious, direct evidence of organizational pathology. bly in reducing the number of special districts for the In the late 1950s and early 1960s, for example, a provision of sewer services and education. St. consortium of scholars from local universities con- Louisans have also agreed to areawide special dis- ducted a study of governments in the St. Louis area, tricts to fund the St. Louis zoo and museums25 and including their capacities to provide urban services. the local junior colleges, and are currently imple- The researchers approached the study believing that menting areawide provision of public hospital serv- the "congery of heterogeneous and overlapping gov- ices. On the other hand, proposals to reverse the ernmental units" in the St. Louis area must have separation of city and county have not been success- adverse effects. The adverse effects they presup- ful; neither have proposals to reduce the number of posed, to quote Scott Greer (a member of the municipalities in the county through general reor- group), included: ganization. 1) Great variation in output, or service levels, The first formal attempt to reverse the separa- among the different units, tion of city and county took place in 1926.26 Resi- dents of the city supported, by a margin of 7 to 1, a 2) Great variations in the efficiency, or cost proposition that the city absorb the county and all its benefit ratio among the units, municipalities. County residents opposed the propo- 3) A generally low level of some services sition by a margin of 2 to 1 and it was rejected. In throughout the area, due to the deleterious 1930, an amendment to the Missouri Constitution effects of poor services in one governmental was proposed that would have allowed the establish- unit on the services in other, interdependent ment of federations of local governments. St. Louis units. . . .27

Table 2.8 Revenue Variations In County Municlpaiities- 1985 Percent of Percent of Average Range of Number of Municipal Total Revenues Revenues Municipalities1 Population Revenues* Per Capita Per Capita 32 25% 15% $196 $121-224 20 26 21 264 224-289 17 24 26 348 293-422 18 25 39 518 423-1,901 Point-of-Sale Municipalities1 (37) 55 68 407 216-1,901 Pool Municipalities1 (50) 45 32 236 121-423 lDoes not include Maryland Heights or Velda Village Hills. 2~otalrevenues include those of general purpose governments and, where relevant, fire protection districts. They also believed that there was no relationship and by a bare majority in the county. Voters in the between the vitality of local political processes and city and county approved the addition of an histori- the size of governmental units. Nor did they believe cal district to the zoo-museum district in 1987. these presuppositions to be subject to debate. As Most recently, the St. Louis County Executive Scott Greer put it, the propositions "were not in- and the citizens organization Confluence St. Louis itially stated as hypotheses; their validity was as- have proposed incorporating all of the unincor- sumed, for they were part of the overall ideology of porated county and simultaneously reducing the the movement to save the cities." Given this point of number of municipalities by 50 percent or more.28 departure, it is not surprising that they recom- The proposal followed a number of recent municipal mended governmental consolidation for the area annexations, a new incorporation, and additional in- even in the face of evidence that would have chal- corporation efforts-all having the effect of reducing lenged their basic presuppositions. the unincorporated territory taxed and serviced by In partial response to their recommendations, a county government. Fire districts would also be con- city-county consolidation plan was advanced, known solidated, and fire services removed entirely from locally as the Borough Plan. Under this plan, the municipal jurisdiction, leaving a total of four fire dis- city and county were to merge, existing local govern- tricts together covering the whole county. Counter ments were to be disbanded, and the resulting gov- proposals have also been advanced-one by the St. ernmental entity was to be divided into several bor- Louis County Municipal League-that would leave existing municipalities intact.29 A new period of re- oughs, most of which would have included portions form discussion has begun. (The various reform pro- of both the city and the county within their borders. posals are discussed further in Chapter Nine.) This plan elicited widespread opposition from local government officials in both the city and the county, and from organized interest groups and black citizens of the city. Black residents of the city, for ex- CONCLUSION ample, argued that the Borough Plan would greatly reduce their growing influence in city government. St. Louis is a diverse area in terms of its sub- The plan failed to receive a majority statewide-fail- populations, and this diversity is reflected in its gov- ing in every Missouri county-and was rejected by ernmental jurisdictions, especially in St. Louis margins of 5 to 4 in the city and 4 to 1 in the county. County. Many analysts would characterize the juris- It was the last proposal to reverse the 1876 separa- dictional pattern in the county as a patchwork or tion of city and county that has been presented to maze that defies any rational order. If rationality im- the voters. plies uniformity, these characterizations are accu- After the defeat of the Borough Plan, reform rate. Yet the fragmentation represented by the large proposals began to focus on governmental arrange- number of formal jurisdictions in the county has to be balanced against the integrating effects of numer- ments in the county. In 1968, county residents ous multiorganizational arrangements-political fo- adopted a new home rule charter that greatly re- rums and joint administrative units created as an duced the number of elected county government of- overlay on top of the jurisdictional patchwork. Ef- ficials and, thereby, consolidated the government of forts at comprehensive governmental reform, all the unincorporated county under a strong county su- aimed at simplifying the political geography, have pervisor with broad executive powers. In 1970, a failed to date. The patchwork not only persists, but county home rule amendment to the Missouri Con- also continues to develop, as does the elaborate stitution was adopted with majorities statewide and overlay of multiorganizational arrangements. Fiscal in St. Louis City and County. This amendment al- disparities, especially between point-of-sale and pool lows citizens of charter counties to decide what serv- cities, have been the greatest continuing source of ices will be provided by county governments in in- conflict. corporated and unincorporated portions of those A possible new wave of annexations and incor- counties. Citizens of St. Louis County now have the porations, however, currently threatens to over- capability of transferring functions from municipal to shadow all other issues and to alter the political ge- county government responsibility, but have so far ography by taking county government largely out of not used this power. its municipal-type service provision role. Once again, In 1971, the last of the current areawide special citizens are being asked to consider sweeping pro- districts was created. A proposal to create a cultural posals to restructure the county. To examine these district to provide for the St. Louis Zoo and muse- issues, however, first requires a closer look at the ums was accepted by a margin of 3 to 1 in the city basic rules and functional arrangements for metropolitan governance in the St. Louis area-the subject ers, it also made local government more accessible to residents, most of whom also hold full-time jobs. Rather than of Chapter Three. being forced to take time off from work to meet with local officials, citizens could attend city meetings at times that ENDNOTES were convenient to them. 'This discussion of the early development of St. Louis and 140nly one municipality, however, has a council-manager St. Louis County is derived in large part from the excellent form of government in which there is no separately elected monograph, History and Growth of St. Louis County, Sixth mayor. Edition, written by Robert A. Cohn and published by the St. Louis County Office of Public Information in 1973. Iscertain revenues of St. Louis City and County are not included in these totals to maintain comparability with mu- ZEugene M. Violette, History of Missouri, first published in nicipal and fire district revenues. For the city, revenues of 1918 and reprinted in 1953, pp. 14-15. the debt service, capital projects. expendable trusts, con- 3Missouri law in this period gave no voice to residents who vention center, miscellaneous special, and hospital division were the target of a proposed annexation, requiring only a special funds were excluded, as were revenues which were simple majority vote among residents of the municipality expended for health and welfare purposes. The county reve- proposing an annexation. nue total excludes revenues of the debt service and "other" 4Based on 1984 population estimates. See Missouri Popula- funds, together with revenues associated with community tion and Census Newsletter, December 1985, cited in the health and medical accounts. 1985 Fact Sheet: Police Services in St. Louis County, pub- l61n St. Louis County, some municipalities maintain fire de- lished by the Bureau of Management Services, St. Louis partments with municipal revenues. Citizens in other mu- County Department of Police. The estimate of population nicipalities and in the unincorporated county receive fire for the City of St. Louis is taken from U.S. Department of services from fire protection districts, and are taxed sepa- Commerce, Bureau of the Census, Local Government Fi- rately for these services. To place service-related revenues nances in Major County Areas, 1984-85, GF 85, No. 6. and costs on a comparable basis, fire protection district to- SQuintile averages are a means of illustrating variance in a tals are included with municipal data for some analyses in distribution. To compute these averages, data on the series this chapter. Where this is done, revenues andlor costs are of interest, percent minority for example, and municipal referred to as total municipal and fire protection revenues- population are sorted in ascending order, from the lowest costs or total local revenues-costs. value to the highest. Then, an average value is computed 17~henew third class City of Maryland Heights is excluded for each portion of the resulting list that contains roughly 20 from these data as it had revenues for only a portion of percent of the total municipal population. Table 2.1 shows, 1985. for example, that 20 percent of the County municipal population resides in municipalities having 1.5 percent minority 18~otalcounty and city revenues with the exclusions listed in population, 20 percent in municipalities with an average of Note 15 above. 2.8 percent minority, and so on. The highest quintile, rep- 191n 1971-72, property taxes comprised 27 percent of the gen- resenting those municipalities with the highest percentages eral revenues of municipal governments in St. Louis of minority population, has a minority population which av- County. By 1976-77 this percentage had dropped to 16, and erages 56.2 percent. The quintile averages in the Table are in 1984-85 it was approximately 11 percent. Property tax computed separately for each data series. revenues as a percentage of total municipal revenues na- 'In addition to the Normandy area municipalities, Berkeley, tionwide decreased over this same period, from 31 percent Ferguson, Jennings, Kinloch, and University City. in 1971-72 to 21 percent in 1981-82, although Missouri's 'Including the unincorporated portion of the county. rate of decrease was much greater. See Significant Features BThis observation raises a question for those who argue that of Fiscal Federalism, published annually by the Advisory too many governments is the problem of the St. Louis area. Commission on Intergovernmental Relations. It is difficult to see how the many small governmental units 20Third class and home rule cities are more likely to produce in the county, given their very limited populations and land fire services with their own departments; villages and fourth areas, could constitute the metropolitan problem. class cities are more likely to receive such services from fire @Pacificand the Pacific Fire Protection District are not in- protection districts. cluded in these counts. 21 1n St. Louis County, some municipalities obtain significant 1°For a history of the League, see Charles W. Sisler, History revenues from the sale of services, principally police serv- of the St. Louis County Municipal League (Clayton: St. ices, to other municipalities under contract. One municipal- Louis County Municipal League, May 1986). For a discus- ity raised in excess of 30 percent of its revenues in this fash- sion of recent kague activities, see James M. Brasfield, ion. Several other municipalities, benefitting from their "The Great Land Rush: Annexation and the Sales Tax in geographic locations astride heavily travelled traffic arter- St. Louis County," a paper presented at the Annual Meet- ies, raised significant revenues from fines and fees-princi- ing of the American Political Science Association, New Or- pally traffic related-imposed by their municipal courts. leans, Louisiana, August 1985. Three small municipalities obtained more than 30 percent of their revenues from this source, and an additional eight "See the East-West Gateway Coordinating Council Publica- municipalities raised at least 10 percent of their 1985 reve- tions List, March 1985. nues from court fines and fees. The general subject of local 12Citizens who were concerned about local problems in the fees and charges is explored in ACIR, Local Revenue Diver- Chesterfield area addressed those concerns to their repre- sification: User Charges (Washington, DC: ACIR, Report sentatives in the Missouri legislature, each of whom serves SR-6, October 1987). a significantly smaller and more geographically focused 221ntergovernmental revenues are not insignificant for the City constituency than do county council members. of St. Louis, however, which received nearly $70 million in 13~benefit of having officeholders with full-time jobs in addi- grants in 1985, principally from the federal government. tion to their municipal responsibilities was evident across 23The higher revenues accruing to the City of St. Louis might many jurisdictions. Village and city meetings were sched- be looked on as an example of inequity in the local political uled on weekday evenings or on weekends. Not only did this economy. Only six municipalities in St. Louis County, with accommodate the full-time working schedules of officehold- a combined resident population of 28, 502 persons, raised total per capita revenues (municipal and fire district) in ex- grams, St. Louis University, published in Fostering Devel- cess of the city's. Forty-five municipalities had total per opment in Metropolitan St. Louis, a report submitted to the capita revenues that were less than 50 percent of city reve- City-County Task Force of Civic Progress in 1982. nues. But discussion of fiscal disparities in St. Louis has not focused on this comparison. The city is viewed as the locus 27S~ottGreer, "Dilemmas of Action Research on the Metro- of extraordinary service demands, requiring (and, in the politan Problem," in Morris Janowitz, ed., Community Po- sense of equity, deserving) extraordinary revenues. Cer- litical Systems (Glencoe, Illinois: The Free Press, 1961), tainly the city's status as an employment center for the area p. 193. For a broader discussion of the approach, see Scott engenders additional service costs (and revenues) beyond Greer, Metropolitics: A Study of Political Culture (New those of primarily residential communities. York: John Wiley and Sons, 1963). ?%deed, as shown in Chapter Eight, revenues in St. Louis 2eSee St. Louis County Department of Planning, A Compre- County are closely related to the presence of such commer- hensive Proposal for Local Government Reorganization in cial and industrial enterprises, and presumably with the St. Louis County, November 1987, and Confluence St. population influx which they produce. Louis, Too Many Governments? A Report on Governmental =A referendum has just added an historic district to the zoo- Structure in St. Louis City and County with Recommenda- museum special district. tions for Change, February 18, 1987. =This review of reform proposals in St. Louis draws, in part, =See St. Louis County Municipal League, Policy: Govern- from information compiled by the Center for Urban Pro- mental Organization, June 4, 1987. Chapter Three Metropolitan Governance in St. Louis: State Rules, Local Arrangements

INTRODUCTION areawide governance is nonexistent, or chaotic, in To understand the organization of metropolitan fragmented metropolitan areas. Governance ar- St. Louis, especially St. Louis County, the first sub- rangements, however, depend not on a single center ject to examine is not how local governments oper- of power, but on a capacity to make and enforce ate, but how they are created. No single organizer general rules of law. The creation of a local govern- designed the present configuration of local government unit is, in functional terms, a "constitutional" ments in St. Louis County. Rather, each local juris- choice-a lawmaking decision on the pan of citizens diction is a product of community choice-an arti- that governs the processes of local government. Un- fact that was locally created, albeit with legal tools derlying this local constitutional capability is an supplied from the state capital. Local citizens, acting elaborate structure of enabling rules embodied in through processes of petition and referendum, make state law. The choice of enabling rules-one step re- a series of incremental choices that, in the aggre- moved from the actual creation of local government gate, produce and maintain a diverse pattern of ju- units-can also be viewed as a constitutional func- risdictional fragmentation and overlap. tion, a process that governs the establishment of lo- The study of local government law is often cal governments. These processes jointly provide a predicated on the idea that local governments are functional basis for metropolitan governance. "creatures of the state." The functional arrange- The basic theme that emerges from the constitu- ments for creating local governments, however, are tional history of St. Louis County is one of local radically different from the prevailing legal concep- freedom-not so much freedom of individual choice, tion. Although local governments in every state are but freedom of collective constitutional choice in a subject to state law and the state constitution,l state variety of local contexts. The freedom to organize governments seldom create discrete local govern- local governments is sufficiently unconstrained that ments, except for basic legal subdivisions, such as one student of metropolitan areas nominates the St. counties and townships.* Instead, state constitutions Louis County area as "one of the best examples of and statutes provide a set of legal tools that citizens what suburban America would probably look like if use locally to create and modify local governments. city formation were left to the voluntary efforts of Functionally, local governments tend to be the local resident^."^ These efforts are not in the strict "creatures" of local communities.3 sense, however, voluntary. The choices made are collective choices, binding on the individuals who are part of the relevant community of interest. Indi- FUNCTIONAL GOVERNANCE ARRANGEMENTS vidual freedom, apart from participation in collec- The absence of a single determining center of tive choice, depends on relocation to a different power may explain the frequent presumption that community. The multiplicity of locally governed communities increases, to some extent, the opportu- into a statewide jurisdiction that could be highly in- nities for individual choice as well. flexible in relation to the specific problems of governance that arise in any particular metropolitan A Local Government Constitution area. The fact that this consequence does not follow It is useful to think of St. Louis County as hav- is due to functional arrangements that involve the ing a local government constitution, a body of rules use of "special legislation." for creating local government units and defining their powers.5 Even though the various constitutional "Local Laws" and the Local Delegation to the rules are not derived from a single legal source, they State Legislature can be understood functionally as a set. By provid- Special local laws are prohibited by the Missouri ing legal instruments for the creation of new govern- Constitution of 1945. The state legislature, there- ments, a local government constitution is even more fore, may not legislate for particular local communi- open ended than a single unit constitution.6 At least ties by name. The device routinely employed by the two levels of constitutional choice are implied: state legislature to comply with this requirement is to 1. At the first level, state legislators and citi- use general language to create a class for which only zens who participate in state constitution making one area or jurisdiction qualifies at the time. Thus, choose a set of enabling rules designed to organ- special legislation for St. Louis County refers to a ize a process by which local communities can "first class county operating under a charter form of create local governments and to empower those government and not containing a city or a part of a local governments. These rules are found mostly city of over 400,000 inhabitants." No other county in state constitutions and state statutes. qualifies as a member of this class. The Missouri courts allow this form of legislation on the rationale 2. At the second level, local citizens choose that other counties may qualify in the future. Simi- to create (or otherwise modify or even termi- larly, St. Louis City is characterized as "any city not nate) local governments in their communities, within a county." Even though this class has only a adopting charters, or other organizing acts, that single member, the state courts consider it to be a actually establish local government jurisdictions. reasonable classification for legislative purposes. This process generates a second set of rules The routine use of special legislation to maintain found in county and municipal charters, as well a unique local government constitution to govern as in the organizing acts of special purpose dis- metropolitan St. Louis casts the local delegation to tricts and residential subdivisions. the state legislature in a specific functional role, Both levels of choice are constitutional in a given the customary deference accorded local bills functional sense, adopting rules that govern future by other legislator^.^ The St. Louis County delega- collective choices. Instead of making those collective tion is composed of 7 senators and 31 representa- choices directly, constitutional decisionmakers pro- tives, all elected from separate districts. The City of vide a set of rules that enable others to make those St. Louis adds 3 senators and 15 representatives. choices and, at the same time, constrain the choices Where state law is concerned, the members of the they can make. Constitutions do not operate govern- local delegation become, functionally, the keepers ments, but constitutions do establish the constraints of the rules by which the diverse jurisdictions and opportunities within which governments oper- of St. Louis County are governed. Though not politi- ate. The greater the number of relatively autono- cally neutral, the state legislators do not-at least not mous, yet interdependent, governmental units within directly-represent any particular unit of local gov- a metropolitan area, the more salient such constitu- ernment. Together they provide an institutionally tional rules become. separate locus of local decisionmaking, accountable The bi-level process of constitutional choice in by election directly to local citizens. metropolitan St. Louis depends on an institutional Legislative deference or courtesy tends to work separation of the two levels into distinct decision- only when the local delegation is in substantial making processes involving different decision- agreement. Consensual decisionmaking does not im- makers. The use of statewide instrumentalities-the ply the absence of conflict; on the contrary, achiev- state constitution and state law-to develop the ena- ing a consensus can be more conflictual, and involve bling rules that comprise the local government con- more hard bargaining, than putting together a mini- stitution achieves such a separation. Yet the use of mum-winning coalition. Nevertheless, the practical statewide decisionmaking processes does not in and requirement of consensus may focus attention on of itself give the St. Louis area, and particularly St. widely shared common interests-those common to Louis County, a unique local government constitu- the many diverse communities of interest found in a tion. Rather, it would appear to absorb St. Louis metropolitan area. Multljurlsdlctlonal Forums for fect one another. When ordinary politics leaves par- Metropolitan Governance ticipants dissatisfied, however, there is another- constitutional-level of decisionmaking available to Even though state legislators can be viewed as them. Local community leaders can shift levels and keepers of the rules, functional arrangements for attempt to work out an agreement for modifying the metropolitan governance extend well beyond the "rules of the game" under which they operate. At legislature. State legislators are not the only, or even some point in this process, locally elected state legis- the primary, participants in local decisionmaking lators enter as necessary participants. that pertains to metropolitan governance. As problems and opportunities develop locally that cannot A Typology of Rules be addressed satisfactorily within existing rules, local The enabling rules that comprise the local gov- officials begin to give attention to possible modifica- ernment constitution of St. Louis can be grouped as tions. Various countywide forums are available for follows: the officials of the various local government units to 1) Rules of association-or how local govern- work out their differences. The local delegation to ment units may be formed by local citizens; the state legislature tends to reflect the consensus that develops among local leaders. In this manner, 2) Fiscal rules-the ways in which local gov- annexation rules have been changed, permissive tax ernment units are authorized to raise reve- laws developed, and authority to create special dis- nue; tricts established in ways that are unique to St. Louis 3) Boundary change rules-mostly pertaining County. to annexation; One important forum for considering issues that 4) Contractual rules-the relative freedom of cut across municipal boundaries in the county is the local government units to enter into agree- St. Louis County Municipal League, with member- ments with one another. ship including all but a very few of the smallest mu- For each type of enabling rule, there exists a corre- nicipalities. The league has been quite active over sponding set of constitutive and collective choices- the years in developing, lobbying for, and obtaining those which form a local government, adopt a tax, legislation pertaining to the county that modifies alter a boundary, or enter into contracts. These statutory rules affecting local government organiza- rules, as well as their development and use in St. tion, in ways that accommodate changing local con- Louis County, are explored in the remainder of this ditions. One of its most recent legislative successes chapter. was the development of a compromise plan for the distribution of sales tax revenues.8 Currently, the RULES OF ASSOCIATION: league is active in the debate over governmental re- MUNICIPAL INCORPORATION, organization in the county. The league's successes COUNTY HOME RULE, AND can be attributed to its ability to articulate a com- DISTRICT FORMATION mon interest that transcends the individual interests The fundamental set of rules in a local govern- of its members, and on the basis of which the county ment constitution are rules of association, those ena- delegation to the state legislature can often obtain bling rules that allow citizens to organize their com- legislative approval for modifications in the rules of munities as governmental units. Rules of association local government bearing on St. Louis County. contain several distinct elements: Other interlocal forums are also involved in develop- 1. Provisions for creating different types of ing state legislation that pertains to particular service governmental units for different purposes-mu- areas. Legislation related to police, fire, or educa- nicipalities, school districts, fire protection dis- tion is likely to be developed by associations that tricts, and other special purpose districts. Mu- represent police chiefs, fire chiefs, and school nicipalities are usually further distinguished by boards and superintendents (see Chapter Two). population class and by various forms of govern- Through such associations, the local self- ment-mayor-council, council-manager, and governance made possible by jurisdictional fragmen- commission. The broad purpose of this subset of tation can be extended to include metropolitan self- rules is to enable local communities to select, governance-the capacity to make rules that apply to within limits, the type of association they want to multiple jurisdictions. Intergovernmental politics in create, choosing from a somewhat open menu of St. Louis County is carried on, therefore, at two dif- possibilities. ferent levels. One level might be viewed as "ordinary politics," as multiple units of government both inde- 2. The choice of collective decision rules by pendently and often jointly make decisions that af- which local residents can act to constitute a new local unit. These rules generally consist of (a) a ernments. The flexibility available to communities petition requirement, by which some number of within St. Louis County contrasts with the relatively local citizens is able to initiate consideration of fixed governmental boundaries established for St. the creation of a new unit, and (b) an election Louis City as a single unit. requirement, together with a voting rule, by St. Louis City and St. Louis County retain the which the relevant local community is able to capacity to reunite or otherwise adjust their relation- decide whether to create the new unit. Some lo- ship in whatever manner they might agree on. Under cal government constitutions also contain veto the Missouri Constitution, a board of freeholders provisions that require a more inclusive unit of may be appointed jointly by county government, city local government, such as a county, or a bound- government, and the governor of Missouri. This ary review commission to approve a citizen peti- board is authorized to propose changes in the exist- tion or referendum decision to create a local ing city county relationship, including: government .9 Consolidation of St. Louis City and County 3. Provisions for reconstituting an associa- as a single metropolitan government consisting of the City of St. Louis; tion and for disassociation. Local government units are created to be perpetual. They persist as 8 Reentry of St. Louis City into St. Louis initially created unless affirmative action is taken County; to modify or abolish them. A set of rules similar Annexation of county territory by St. Louis to those used to constitute a unit (i.e., petition City; and election) can enable local citizens to recon- Establishment of special service districts; stitute a unit (e.g., changing the class or form of and government of a municipality), to consolidate with other units, or to abolish the unit. Any other plan. To be adopted, a plan proposed by the board of In the discussion that follows, the rules of asso- freeholders must be approved by concurrent majori- ciation are described for St. Louis County and ana- ties in popular referenda held in both the city and lyzed initially according to the type of unit created the county. or empowered. A distinction will be maintained between St. Louis County as a basic legal subdivision Municipal incorporationlo of the state and St. Louis county government as a The minimum requirement for municipal incor- local service provision unit created by local choice. poration in Missouri is a contiguous residential The types of local units include municipalities, population of an urban character. Initially, the class school and fire protection districts, various special of an incorporation is determined by population purpose districts, and county government. size. Communities of no more than 500 people may incorporate as a village. Communities of more than City-County Separation 500 but no more than 3,000 residents may incorpo- The development of local government in the St. rate as a fourth class city;" those with more than Louis metropolitan area received one of its basic le- 3,000 residents, as a third class city. Once incorpo- gal parameters in 1876, when the citizens of the City rated, however, a village or city remains in its initial of St. Louis and those in the remainder of St. Louis classification regardless of population growth or loss, County voted to separate, as authorized by the Mis- unless it elects by popular vote to change its classifi- souri Constitution of 1875. Before separation, the cation. Over time, therefore, the class of incorpora- land area of the city was increased more than three- tion can become to some extent a matter of local fold through annexation, sufficient, or so it was choice. thought, to allow for future growth. In effect, the Communities of more than 5,000 residents have City of St. Louis became a city-county. The tradi- still a third option: to incorporate as a home rule city tional panoply of separately elected county adminis- under the Missouri Constitution.Q Villages and third trative offices, including for example a sheriff and or fourth class cities incorporate under provisions of clerk, continue to exist in the city today, alongside state law, while home rule cities derive their legal the mayor, council, and mayoral directed city de- authority directly from the state constitution of partments. As its own county, the boundaries of St. 1945. Although still subject to state law, as any mu- Louis City were fixed, unable to expand further nicipality is, home rule cities do not require specific through annexation. St. Louis County, on the other authorization in state law in order to act. In con- hand, became a rural, soon-to-be-suburban jurisdic- trast, villages and third or fourth class municipalities tion, setting the stage for the significance of the may only do that which they are authorized to do by county as an arena for the creation of new local gov- state statute. Home rule cities remain subject to state prohibitions, thus permitting unlimited state pree- by two-thirds of the legal voters in a fourth class city mption of local decisionmaking. and by three-fourths in a village, and submitted to The initial organization of a city as a municipal the county council, which decrees the disincorpora- corporation has been and remains a relatively easy tion. No election is required.19 There is no estab- process in Missouri. A three-step procedure is fol- lished procedure for disincorporation of a third class lowed, as prescribed by state law. First, 15 percent city-an omission that apparently has not created a of the voters in the area to be incorporated must problem. Disincorporation of a home rule charter sign a petition. (The cost of this effort obviously de- city can be accomplished by charter amendment or pends on the population of the area to be incorpo- as otherwise directed by the charter. One disincor- rated.) Second, county government organizes an poration effort in the predominantly black City of election within the area. The role of county officials Kinloch was under way as field research for this in this process is generally held to be nondiscretion- study was being conducted.20 ary. Finally, if a simple majority of those who vote in the election approve, the city is created. Special Purpose Districts2l Villages follow a different procedure.13 Two- In addition to municipalities, state law enables thirds of the taxable inhabitants of an unincor- citizens to create a variety of special purpose govern- porated town petition county government (in St. ments, including districts that are subcounty, Louis County, the county council would hear the pe- countywide, multicounty, and (in the case of a dis- tition), which may declare the incorporation when trict including both St. Louis City and St. Louis satisfied that the petition is valid and the incorpora- County) city-county in scope. Special districts have tion is reasonable. No election is held, and greater the distinct advantage of being able to overlap mu- discretion is vested in county government to deter- nicipalities, thus serving to supplement municipal mine the final result. Presumably, the smaller size service provision while at the same time extending involved in the incorporation of a village makes an the service provision boundary. Multicounty (or election unnecessary, since the costs of gathering city-county) districts perform in a similar way by signatures on petitions is determined by the absolute overlapping more than one county jurisdiction. number required, not the percentage of the popula- Countywide districts, on the other hand, serve a dif- tion involved. 14 ferent purpose-that of bypassing county govern- The structural choices available to incorporating ment in order to create a separate collective provi- communities vary by classification. Villages must sion unit to serve the same countywide community. elect a board of trustees who, in turn, elect a chair- Countywide districts are thus redundant (which is man who acts as chief executive officer-there is no not necessarily to imply wasteful) in a sense that other option.15 Fourth class cities may choose be- subcounty and multicounty districts are not. tween a mayor-council form and mayor-council with Subcounty districts in St. Louis County include administrator.16 Third class cities have as additional school districts, fire districts, street lighting districts, options the council-manager form and commission and road districts. All of these can be created (and form.17 Thus, to incorporate as a village is to acquire modified) by citizens voting in a special election in a a particular form of government; in becoming a city, proposed or existing district. All are governed by a community chooses its form of government from elected boards. The school districts in St. Louis among the legally permitted options, depending on County originated as common school districts, basi- population. On the other hand, incorporation as a cally matching a district with each school. Over time, home rule city allows a community to create and population growth has enlarged some districts, while adopt any republican form of government. Several others have been reorganized or consolidated, under home rule cities in St. Louis County have used this rules supplied by state law. These rules and the re- capability to create a hybrid form of government sults of their application are discussed in greater de- that combines a separately elected mayor with an tail in Chapter Seven. appointed city manager. Any community of 5,000 Fire protection districts are of more recent ori- people or more potentially has this sort of flexibility. gin, first authorized by state law in 1947.22 The for- To alter the legal status of a home rule city mation procedure was simple: petition by 100 voters would apparently require an amendment to the state (originally taxpaying electors) followed by the ap- constitution. Ordinary state legislation is not suffi- proval of a majority of those voting in a referendum cient. Home rule cities are therefore afforded ex- in the proposed district. The powers of the three- traordinary legal protection not extended to munici- member board of directors include eminent domain; palities that incorporate under statutory provisions. the adoption of fire protection and fire prevention Disincorporation is considerably more difficult ordinances, punishable as misdemeanors; and the than incorporation.18 A petition is required, signed imposition of property tax levies up to a ceiling provided by state law. In 1969, fire protection districts In 1950, St. Louis County was organized as a were authorized to provide emergency ambulance first class home rule county under the Missouri Con- service, subject to approval by district voters. Fire stitution of 1945.Z4 A charter was framed and districts may overlap municipalities, even though adopted by the voters providing for a single county both villages and cities are also authorized to pro- supervisor (now county executive), elected at large, vide fire protection. Dissolution of a fire district may and a separate county council to replace the tradi- be accomplished on the petition of 100 voters fol- tional county court. The county council is composed lowed by approval from two-thirds of those voting in of seven members elected from districts. As a home an election.23 Consolidation of fire districts can be rule charter county, the St. Louis County govern- initiated by the fire boards involved and approved ment has gradually developed as a single-executive by a majority of the voters in each district, and vot- organization, replacing the sheriff as a law enforce- ers may use an initiative and referendum procedure ment officer with a county police department and, to do the same. with the adoption of a new charter in 1968, elimi- The enabling rules for subcounty districts have nating all other separately elected executive officers an open-ended character that is lacking in the rules except the prosecuting attorney.25 permitting creation of countywide districts, as well as The home rule charter provided St. Louis multicounty and city-county districts unique to the County, in effect, with a dual system of government. St. Louis area. Subcounty districts are created in the In the incorporated area, municipalities continued same manner as municipalities and represent a con- as the predominant providers of local services. The tinuing set of options available to local citizens, while county government was not permitted to preempt the other special districts reflect unique authority to their role because a charter provision required that create a specific district. As constitutional choices, any transfer of service responsibility to county gov- the two types are therefore quite different. ernment be approved by the voters in each munici- The two countywide districts are related to edu- pality. This gave any municipality a veto over cation. One is the Special School District authorized countywide assumption of existing municipal func- to provide special education for handicapped stu- tions. In the unincorporated area, however, county dents and vocational-technical training for all stu- government was given municipal-type powers with dents. The other is a community college district, respect to public health, police and traffic, building authorized to provide adult and junior college edu- construction, planning, and zoning. Functionally, cation. Each was created by citizens voting in a ref- but not formally, county government became a mu- erendum, and each is governed by an elected board. nicipality for the unincorporated area. Two districts have been organized that include The county reorganization effort, begun in 1950 both St. Louis City and County-the Metropolitan and completed in 1968, established for the county Zoological Park and Museum District and the Met- as a whole a form of government characterized by ropolitan St. Louis Sewer District. While created by vigorous political leadership in the office of the popular vote, both are administered by appointed county executive. It has also supplied the county as boards closely tied to city and county governments. a whole, but especially the unincorporated areas, with a large professional civil service. This develop- County Government ment set the stage for a period of rivalry between two fundamentally different patterns of organization: A county is different from most municipal units (1) a noncentralized system based on the highly lo- in that county boundaries are generally determined calized constitutive choices of citizens in various by state law rather than by local citizen action. One communities, which has resulted in a collection of primary function of a county is to serve as a legal relatively small municipalities and special districts, subdivision of the state. Locally, a county becomes a loosely tied together via a range of cooperative decisionmaking arena within which citizens can cre- agreements and (2) a centralized system based on ate other local units. To some extent-in Missouri to the energy of a single executive, areawide represen- a large extent-local government constitutions can tation in a small local legislature, and a professional- vary among counties or classes of counties. The de- ized bureaucracy. The contest might also be viewed gree to which counties undertake local service provi- as one between two different conceptions of metro- sion responsibilities, in addition to delivering certain politan order, personified on the one hand by a statewide services locally, varies widely throughout highly visible public official-the county executive- the nation. This aspect of county government may and, on the other hand, by hundreds of part-time also be subject to local citizen choice-as in the case mayors and other public officials.26 of St. Louis County-by virtue of county home rule For a time, the contest took the form of a side- provisions in the state constitution. by-side rivalry. St. Louis County was, in effect, divided into two sectors: incorporated and unincor- and the private sector are also subject to regulation porated. In the incorporated area, pluralism reigned by the local government constitution. As with the supreme. There, county government was limited rules of association, substantial freedom also charac- mainly to the provision of support services and arte- terizes the rules of contract with respect to all local rial streets, in addition to the delivery of some mu- government units. Freedom of contract has been nicipal services under contract arrangements with crucial in the development of St. Louis County. villages and cities, chiefly tax collection and inspec- Contracting and other cooperative arrangements tions-less frequently, police patrol. In the unincor- among all municipalities and other political subdivi- porated area, a unified county government supplied sions of the state were authorized by state law as arterial and residential streets, police patrol and po- early as 1939. This general authorization was incor- lice support services, as well as professional planning porated into the new state constitution adopted in and land use control. 1945. A second provision in the state constitution In 1970, the Missouri Constitution was amended specifically authorized any home rule county to per- to allow the citizens of a home rule charter county to form any of the services and functions of any mu- determine what services shall be supplied to their in- nicipality or other political subdivision of the county, corporated and unincorporated areas by local and except school districts. County contracting was con- county governments. County government for the ditioned, however, on approval by a majority of vot- first time acquired general preemptive powers, sub- ers in the municipality or other unit receiving the ject only to countywide voter approval. However, ef- contracted service. Contracting authorization was forts to transfer service responsibility from municipal extended by state law in 1957 to include "any pri- to county jurisdiction, pursuant to this grant of vate person, firm, association, or corporation."28 authority, were not successful. In 197 1, three pro- The principal limitation on intergovernmental posed charter amendments would have established contracting in Missouri is a requirement that each of minimum qualifications for police officers county- the governmental parties to a contract have author- wide and performance standards both county and ity to perform the function or activity that is the sub- municipal police departments, a uniform countywide building and construction code, and a minimum ject of the contract.29 This is sometimes referred to housing code. All were defeated by a countywide as the requirement of equal or common powers. Thus a fire district, for example, cannot contract to vote .27 County home rule remains a limited grant of produce nonfire services-or any service it is not power, one that applies mostly to county govern- authorized to provide-for a municipality. County ment, and only marginally to the other units of gov- government cannot enter into public education via ernment located within the county, some of which the contracting process. Contracting, in other words, also enjoy home rule powers. Full county home rule requires overlapping jurisdiction in terms of substan- would make the county charter functionally equiva- tive authority. It does not, however, require overlap- lent to the local government constitution. The 1970 ping jurisdiction in territorial terms. While the latter amendment to the Missouri Constitution took a would be a serious impediment to interlocal con- small step in that direction, but for the most part the tracting, virtually prohibiting contracting among mu- local government constitution remains, at least for- nicipalities, the requirement of common substantive mally, in the hands of state legislators. authority does not pose a serious obstacle to most Two basic issues have divided county and mu- intergovernmental contracting. nicipal governments in St. Louis County. One is State law specifically provides for the establish- concerned with partially competitive sources of reve- ment by contract of joint units of public administra- nue, especially sales taxation, and the competition tion (Section 70.260, Missouri Revised Statutes). between municipal and county governments, as well Such units may be managed by boards or officers as among municipal governments, for this attractive established under the contractual arrangement. revenue source. The other has to with appropriate Interlocal contracts thus are not limited simply to procedures for municipal annexation and, espe- one unit performing a service for another unit, but cially, the role of county government in that process. extend to the use of joint service agreements that These issues are explored below. establish new administrative units to serve two or more jurisdictions at once. CONTRACTUAL RULES: Fire districts are limited when entering into con- INTERJURISDICTIONAL AND tractual arrangements to supply fire service for a mu- PUBLIC-PRIVATE AGREEMENTS nicipality to a charge equal to the tax rate of the The voluntary relationships between local gov- district for fire (but not ambulance) service multi- ernment jurisdictions and between local government plied by the annual assessed valuation of all property taxed by the municipality (Section 32 1.21, Missouri have been the source of significant, and continuing, Revised Statutes). Ambulance services supplied out- controversy. side the district must be supported by user charges. This provision effectively prohibits bargaining over Annexation price by fixing the compensation at a determinate One of the potential sources of substantial con- amount. The fire district must in effect charge its flict among multiple jurisdictions in a metropolitan municipal customer the same as it charges its own area is boundary adjustment through annexation. residents. Municipal annexation always occurs at the expense Other units of local government are free, how- of county government responsibilities and, poten- ever, to employ whatever pricing policy they please. tially, at the expense of neighboring municipalities Frequently, the marginal costs of extending service that might wish to annex the same area. The rules to some additional population are considerably less governing annexation are therefore among the most than the average cost of service provision for every- sensitive in the local government constitution. The one. This may give a municipality an incentive to history of Missouri annexation law as it affects St. agree to a price for service delivered outside the mu- Louis County reflect periodic efforts to adjust the nicipality below the average price per person served. rules. Despite the appearance of inequity, all persons may An expansion of the boundaries of a municipal- be left better off as a result of this bargain.30 ity is equivalent to an incorporation decision for those subject to annexation, except that the munici- BOUNDARY CHANGE RULES pality already exists and others are simply joining a previously constituted association. For much of the period of the development of St. Louis County, an- Under present state law, the consolidation of nexation was nonetheless involuntary, not just on municipalities, school districts and fire districts can the part of individuals, but on the part of the com- be accomplished in Missouri by concurrent majori- munity being annexed. Later, a judicial procedure ties of those voting in an election held in each mu- was established that allowed county government to nicipality, subject only to the requirement that the represent the interests of those outside a municipal- jurisdictions be contiyous.32 This procedure pre- ity seeking to annex territory. This evolved into a serves self determination for those communities al- virtual veto power wielded by the county council. As ready incorporated. School districts, however, can county government further developed its own service also be merged under reorganization authority that provision capabilities, it became a major competitor requires only a simple majority across all districts in- with municipalities for the service franchise of grow- cluded in the proposal (see Chapter Seven). Pre- ing areas. This created a potential conflict of inter- sumably the state legislature could directly consoli- est: county government was given the authority to date municipalities without their consent, except represent the interests of citizens in an area pro- (most likely) for those that have incorporated as posed for annexation, when its own interests (mainly home rule cities under the Missouri Constitution. A because of the potential loss of tax revenue) lay fre- city-county board of freeholders, discussed above, quently with a denial of the annexation petition. In may also have authority to propose consolidations, the meantime, state law was modified to require dual and other changes to existing relationships among election, approval by a simple majority of voters in municipalities and county government, in St. Louis both the annexing municipality and the area to be County alone. Debate over this question has recently annexed. Then, most recently, the state Supreme surfaced. A board of freeholders, recently ap- Court concluded that county government, given the pointed, was engaged in legal research on this requirement of dual election, no longer had legiti- point.33 If legal, such a proposal could bring about mate interests to represent in an annexation pro- municipal consolidations if accepted by concurrent ceeding. majorities in both the City and County of St. Louis, The history of annexation rules in St. Louis voting in a referendum (see Chapter Nine). County can be divided into four periods:35 Municipal consolidations have occurred, infre- 1. Prior to 1953, state law provided that cit- quently, in St. Louis County, primarily in the form ies of all classes could annex adjacent territory of the merger of a very small municipality with a at their own discretion. All that was required larger neighbor.34 (Eight of the current 23 school was a resolution by the mayor and council, fol- districts, however, were created through mergers, lowed by approval by a simple majority of legal seven by means of reorganization authority.) Mu- voters in the annexing city. No vote was re- nicipal consolidations have generally been noncon- quired in the area to be annexed. This mode of troversial, but boundary changes through annexation annexation, unilateral action by the annexing city, was historically the legal basis for the ex- selves the interest previously represented by pansion of America's major cities, closely fol- county government in their acceptance or rejec- lowing patterns of population growth. It was tion of the annexation in a separate election. abandonment of this rule by many states early in When residents in the area proposed for an- this century that precipitated the growth of inde- nexation elected to become part of Town and Coun- pendently incorporated s~burbs.~e try, a new era in municipal-county relations was in- 2. For a decade, between 1953-63, .St. augurated. Municipalities began to prepare Louis County was governed by the terms of Mis- annexation plans for submission to the voters. Time souri's Sawyers Act, which required that annex- is of the essence in these proceedings because state ing cities file for a declaratory judgment in cir- law gives priority to the municipality that is first in cuit court, seeking authorization to annex. The time with its proposal for annexation of any given ability of the annexing city to annex unilaterally parcel. was left intact, subject only to a judicial finding Fire protection districts, because the services that the reasonableness of the proposed annexa- they provide can also be provided by municipalities, tion be fairly debatable. Once authorization was may also find their territorial domain reduced by obtained, the procedure remained the same as municipal annexation. Whether this occurs is deter- before. The basic rule during this period might mined by voters in the annexation election. When be termed unilateral annexation with judicial re- an area to be annexed is served by one or more fire view. districts-almost always the case-voters in each service area must elect whether to continue to be 3. In the early 1960s, the St. Louis County served by the district or to receive fire protection government successfully intervened in an an- from the municipality. If the decision is in favor of nexation proceeding initiated by the City of the fire district, the municipality must rebate a por- Olivette. The Missouri Supreme Court in City of tion of the property tax collected in that service area v. held that the circuit court Olivette GraeleP7 to the district in an amount equal to the district levy. ought to weigh the interests of a county against Voters thus make separate decisions on the question those of the annexing municipality in determin- of municipal annexation and the question of fire ing the reasonableness of a proposed annexa- service pro~ision.~g tion. Over the next 23 years, county government used this rule to intervene and block municipal FISCAL RULES annexations in St. Louis County.38 In the mean- The local government constitution is also a fiscal time, in 1963, state law for the first time intro- constitution.40 Its fiscal rules, together with eco- duced a rule of concurrent majorities in first nomic conditions that vary from community to com- class charter counties only (St. Louis County is munity, constrain both the choices of revenue one of only two such counties in the state). In sources by local officials and their communities and addition to the previous procedural require- the rates that they impose. In spite of these con- ments, a separate vote must be taken in the area straints, a significant range of revenue options re- to be annexed and approved there by a simple mains. Local officials and communities also exercise majority of those voting. This new rule intro- a significant range of choice with respect to levels duced the principle of self-determination, long and patterns of expenditure. the governing principle of incorporation, into With respect to property taxes, local fiscal rules annexation decisions. in Missouri are based primarily on limits found in 4. Beginning in 1977, the City of Town and the state constitution,41 supplemented by state law. Country sought to annex a substantial area that The state legislature may tighten, but not relax, the would have the effect of doubling its land area. limitations required by the state constitution. The County government opposed the annexation be- basic rule for authorizing local governments to raise cause it would deprive the county treasury of tax revenue from other, nonproperty sources is a re- revenues, and described it as a land grab in- quirement of action by the state legislature. Whether tended to enrich a city at the expense of the home rule cities have legal authority to impose a tax county. The Missouri Supreme Court in City of without prior authorization by the state legislature is Town and Country v. St. Louis County, et a1 a matter of dispute,42 but the usual practice is to (1983) ruled that county government was no seek legislative permission before acting. State legis- longer entitled to have its interests weighed in lation may prohibit any local tax, before or after the the circuit court determination of reasonable- fact, in any event. The practical effect is to place a ness. The reason given was that the residents of primary political burden on those who seek addi- the area to be annexed could express for them- tional taxing authority. Within the rules established by the constitution and laws of the state, municipali- service provision. Some members will find that they ties and other local jurisdictions make collective are paying for more services than they would prefer; choices with respect to their particular taxation and others may prefer higher service levels than they are spending levels. The process of decisionmaking that currently receiving and be willing to pay for them. emerges is one of building and maintaining a local Both deviations from a preferred tax benefit pack- consensus around a certain configuration of fiscal age are inefficient.48 rules, while accommodating the revenue-authority Citizens who prefer higher service levels may be preferences of a variety of local jurisdictions.43 able to express their preferences by augmenting public service provision with services purchased in the The Functions of Fiscal Rules private sector (e.g., through the hiring of private se- Fiscal rules serve at least three functions: (1) to curity personnel to augment local police protection, encourage fiscal equivalence, matching the benefits purchasing additional street services from private received from local government activity to the reve- contractors, or, in the case of educational services, nue raised to support that activity; (2) to distribute opting to send their children to private or parochial and, if fiscal equivalence is relaxed, redistribute schools). One option available to those who are revenues among jurisdictions; and (3) to restrict the forced to consume (and pay for) more public serv- taxing and spending powers of local officials, while ices than they would prefer is that of moving to a often empowering local citizens. These functions are different community with a lower benefit and cost explored below. package. Another option for citizens who are dissatisfied with the current benefit cost package in their Fiscal Equivalence community is the exercise of voice, participating in One economic principle applied to the choice of the political realm and attempting to convince their fiscal rules is fiscal equivalence.44 Public finance is fellow citizens that service levels and costs should be understood to rest on both benefit-based taxes and increased or reduced. The cost of exercising voice in ability-to-pay taxes. Local government finance, how- such situations is likely to be higher in larger and ever, is frequently analyzed from a benefit stand- more heterogeneous communities than in smaller, point. Fiscal equivalence in this context simply more homogeneous ones. means that local citizens get what they pay for and Sales taxes, collected from nonresidents as well pay for what they get. Where fiscal equivalence is as residents of a given community, are not directly found, local taxes can be construed as a collective amenable to analysis using the principal of fiscal expression of an economic demand for services, de- equivalence. In part, sales taxes can be conceived as fined as willingness and ability to pay for a local a fiscal device for capturing revenues from benefici- package of benefits.45 aries of local services who are not residents of a ju- A critical variable in establishing fiscal equiva- risdiction. This raises the question of whether non- lence is the incidence of a tax-who actually pays residents receive benefits that are proportional to the tax as opposed to those from whom the tax is the costs incurred in providing services to them and initially collected. The simplest sort of tax is a head to the commercial and industrial enterprises they tax or a flat tax per household or business unit; it is frequent. If this were true, sales taxes would exhibit understood as a public benefit tax on the assumption fiscal equivalence. If, on the other hand, sales tax that all members of the relevant public benefit alike. revenues exceed or are less than the cost of provid- Property taxes are also generally understood as local ing such services, they represent either a subsidy to benefit taxes where those who benefit from local residents by nonresidents or, in the case of insuffi- services by virtue of property ownership pay the tax, cient revenues, a subsidy of nonresidents and busi- and the level of benefit varies according to property ness firms by residents. Income taxes are generally ~alues.~e treated as a nonbenefit tax, but, again, can be em- The assumption of equal benefit in the context ployed in the form of an earnings tax (as in the City of a flat tax or a property tax is more easily sustained of St. Louis) to obtain revenue from nonresidents. in homogeneous than in heterogeneous communi- Again the question of the proportioning of revenues ties.47 As heterogeneity of population characteristics to the cost of benefits provided is raised. increases, or as the mix of interests in a community Under the principle of fiscal equivalence, local (e.g., residential interests vis-a-vis those of commer- governments must be able to obtain revenue from cial or industrial concerns) becomes more diverse, it the beneficiaries of local service provision. On the becomes increasingly less realistic to assume that other hand, local governments can appropriately be collective choices represent a common preference prohibited from employing fiscal devices that shift for benefit levels in relation to costs, or that all com- the burden of supporting local public services to munity members benefit alike from a given level of nonbeneficiaries. Drawing and policing the rather ambiguous boundary between these two types of Home rule counties and all municipalities over a revenue authority can be viewed as a basic fiscal population of 500 may also employ a sales tax of 112 task of the local government constituti~n.~~ or 1 percent, although sales tax authority has been specifically adapted to St. Louis County. Municipali- Revenue Redistribution ties have authority to levy taxes in conjunction with Another important function of fiscal rules is the the granting of licenses for various businesses and establishment of criteria for the redistribution of activities (the details vary with the class of munici- revenues among local governments. Redistribution pality). In the St. Louis area, municipalities and the of some sort occurs whenever relationships between county government are able to use their licensing taxes and services deviate from fiscal equivalence. powers to levy a tax on the gross receipts of public Although redistribution is most frequently thought of utilities. Missouri's two largest municipalities-St. in terms of a transfer of resources from those better Louis City and Kansas City-have authority to im- off to those worse off, the transfer can also occur in pose an earning's tax on both residents and nonresi- the reverse direction. In a multijurisdictional setting, dents. redistribution of revenues requires careful attention to fiscal rules. For example, where citizens of larger, Tax Rate Limits overlapping jurisdictions (e.g., counties, a state, or Limits on the type of tax that may be employed the national government) decide collectively that the by local government units are accompanied by limits operation of a local public economy under a princi- on tax rates. Rate limitations regulate the use of a ple of strict fiscal equivalence places undue tax bur- single tax base by multiple jurisdictions. A single dens on some members of this larger collectivity, household pays property tax in at least three local they may decide to redistribute some of their own government jurisdictions in St. Louis County- revenues to those in less fortunate circumstances.50 county government, a municipality andlor fire dis- Having made such a decision, they must incorporate trict, and a school district. Cities are limited to a rules appropriate to this redistributive goal into the maximum rate of $1.OO per $100 assessed valuation fiscal constitution.51 for general municipal purposes, and county government to a maximum rate of $0.35,52 although voters Lirnltations on Revenue Authority may approve temporary increases. Villages may tax The primary constraints on the taxing and property up to a rate of $0.50 per $100.53 School spending decisions of local officials lie within the districts, however, have no effective tax ceiling as body of fiscal rules enacted by state authority, both long as voters approve increases." in legislation and in state constitutional provisions. Variations in actual property tax rates among ju- Fiscal rules can be divided into tax rules and debt risdictions depend on a variety of factors, with cost rules. Dealing first with taxation, five types of rules of services being only one influence among many. are present: (1) limits on the types of taxes that can The benefit base of the property tax is dependent on be levied by each type of jurisdiction; (2) limits on full capitalization of property taxes into property val- maximum rates by type of tax and jurisdiction; (3) ues; the combination of housing cost and property earmarking requirements that link some portion of a tax (and other, location specific taxes such as utility tax, when levied, to a specific category of expendi- taxes) is the relevant number to use in comparing ture; (4) procedural requirements, focused largely prices across jurisdictions. When examining price on voter approval of tax and fee increases; and (5) competition among municipal jurisdictions that rely distribution rules, which come into play when a tax on the property tax, the relevant price is a package imposed by a larger jurisdiction is distributed among price that includes both housing value and the value constituent jurisdictions. Debt rules involve an allo- of local public services in the same bundle. Property cation of authority among different types of local tax rates alone do not disclose the full opportunity units, limits on the amount of debt that can be in- cost of locating in one jurisdiction as compared to curred, and procedural rules that require voter par- others. Location shopping cannot occur, at least not ticipation. The tax and debt rules found in St. Louis rationally, on the basis of comparative property tax County are discussed in detail below. rates because these reflect neither the full value of service provided nor the full cost of those services to Tax Rules the citizen-consumer.55 Price comparison among lo- All local jurisdictions in Missouri are authorized cal jurisdictions ought to be based on a comparison to levy a tax on real and personal property up to of total tax bills, not tax rates. some limit. The county government, municipalities, Property tax rates also vary with the yield from school and fire districts, and other special purpose other tax instruments used by a jurisdiction. If a tax, districts all have limited use of property taxation. such as the gross receipts from utilities tax, is skewed to impact more heavily on nonresidential (i.e., com- Voter Approval mercial) taxpayers, then both property tax rates and Procedurally, voter approval is an integral pan property values are affected to the short-term of the local fiscal process. Action by voters has been advantage of current residents. Whether there is a required in three ways: long-term advantage from such skewing depends on the mobility of the capital investment being taxed. 1. Voters are asked to approve initial tax Sales taxes, too, can alter the equation. If, as noted levies. Historically this requirement was not in place for general purpose property taxes, but it above, sales tax revenues exceed the cost of services was and is required for earmarked (special) necessary to provide services to nonresidents and property levies and for sales taxes. Some special firms, tax revenues raised from resident taxpayers levies require simple majority approval; others, may be below the value of their benefits. In the op- two-thirds majority approval. posite case, where sales tax revenues do not cover the cost of providing service to nonresidents or 2. Excess levies (i.e., levies that exceed the firms, the revenue raised from residents may exceed maximum property tax rate) also require voter the value of the benefits received. Either impact approval. In this case, approval is for a limited would be undesirable from a fiscal equivalence per- period of four years, as well as in a limited mag- spective. nitude-$0.30 for third and fourth class cities, Sales and excise taxes in Missouri are subject to $0.40 for a constitutional charter city. Two- even more stringent limitation than the property tax, thirds of the voters must approve. School dis- virtually determining the rate. Cities may choose, for tricts, however, are different. Excess levies in example, between a 112 and 1 percent sales tax. In this case may be approved by a simple majority St. Louis County, however, the sales tax rate is set of voters, with or without a time limit, up to by state law at 1 percent countywide. Municipal li- $3.75, and with the approval of two-thirds of the cense taxes, on the other hand, are not subject to voters, no limitation on the rate.56 rate limitation. As a result, municipal taxes on the 3. As of 1981, with the passage of a citizen gross receipts of public utilities in St. Louis County initiative known as the Hancock Amendment in vary from 0 to 10 percent. In addition, the tax can November 1980, the state constitution requires be applied selectively to receipts from nonresidential approval by a simple majority of voters for any users. Where present, this selective use of the tax increase of taxes or fees. This requirement ap- suggests a tax on capital without regard to benefit plies whether or not a jurisdiction is currently received, implying a departure from fiscal equiva- taxing at the maximum statutory rate.S7 lence. Villages are limited to a 2 percent levy on the Indeed, after Hancock and a supplementary bill receipts of electric utilities only, thus creating an in- passed in 19 83 (Section 137.073, Missouri Revised centive to alter the status of a municipality from a Statutes), voter approval is required in certain cir- village to a fourth class city if the population stan- cumstances even to hold tax rates at present levels. dard (more than 200 residents) is met. If, as a result of property reassessment or other fac- In addition to general fund revenues, certain tors, the assessed valuation of existing property in a portions of the property tax may be earmarked for jurisdiction increases in aggregate value at a rate that particular categories of expenditure. These special exceeds an increase in the consumer price index, levies are a means of dedicating revenues to specific that jurisdiction must roll back its tax rate, reducing purposes without creating special taxing and spend- it so as to hold nominal property tax revenues con- ing jurisdictions. Municipalities are authorized to stant or must request voter approval to continue at levy a tax on property specifically to support librar- its present tax rate, thus increasing revenues. ies, public health and hospitals, museums, and rec- The local government constitution thus contains reational programs. The fiscal rules for these ear- an important allocation of fiscal authority to voters marked revenues generally specify a maximum that applies to all local government units. Prior to rate-usually from $0.20 to $0.40 per $100 assessed the Hancock Amendment, the fiscal rules worked a valuation. County government is authorized to levy trade-off between constitutional-statutory restriction special rates for roads and bridges (not to exceed an and voter approval. Local officials could tax up to additional $0.35 per $100 assessed valuation), com- some maximum rate ceiling, after which they had to munity health and medical care (up to $0.38 per face the voters to obtain further revenues by taxing $100 assessed valuation to support a public hospital at higher rates, whether by special levies or excess and health center), and park maintenance (maxi- levies. After Hancock, the only remaining role for mum rate of $0.12 per $100 assessed valuation). legal rate ceilings is to require two-thirds majority Fire districts may employ a special levy to support an approval (in most cases) at four year intervals of any ambulance service. taxes that exceed the ceilings. No tax or fee increase may now occur, nor can local officials obtain addi- The use of distribution rules alongside a uniform tional revenues through property reassessment, with- tax rate thus provides a method of allocating reve- out voter approval.58 Voter approval is now the basic nues between the incorporated and unincorporated form of tax limitation with respect to local govern- portions of the county to support the provision of ments in Missouri. municipal-type services. Another method used is to The requirement of voter approval is consistent confine county government revenue authority to the with the concept of benefit taxation. If voters bene- unincorporated area. This is the case with the fit, they should be willing to approve tax measures. county license tax on gross receipts of public utili- Indeed, if local voters approve, one might ask what ties, which, like the county's cigarette tax revenues, need there is for any sort of rate ceiling (except per- is earmarked by statute (Sections 66.310 and haps to trigger greater voter scrutiny). In this argu- 66.370, Missouri Revised Statutes) for police. The ment, fiscal equivalence is assumed. However, if tax county gross receipts tax rate, unlike the municipal incidence instead places a burden on nonresidents, rates, is determined by state law and set at 5 percent voter approval can no longer be assumed to repre- (Section 66.300, Missouri Revised Statutes). The sent an economic demand for services. Voter ap- county motor vehicle license tax is also earmarked proval therefore cannot substitute for fiscal rules for police and law enforcement in the unincor- that limit the type of tax to be employed. Rate limits porated area (Section 66.330, Missouri Revised may also derive in part from uncertainties or reser- Statutes). vations about tax incidence. Thus the rate limit on The effect of the fiscal rules governing distribu- sales taxes could perhaps be construed as an effort tion of certain countywide taxes and the limitation of to compel some level of reliance on the more explic- other county levies and charges to residents of un- itly benefit-based property tax. incorporated areas is to create, informally, the functional equivalent of a municipal services district for Distribution Rules the unincorporated portion of the county. The objective is to maintain fiscal equivalence between the Some local taxes have special distribution rules. incorporated and unincorporated areas. Ordinarily, In St. Louis County, this feature applies to those the presumption is that services provided to resi- taxes levied on a countywide basis at a uniform rate, dents in an unincorporated area are subsidized by but distributed among local jurisdictions on a for- taxpayers living in incorporated areas who do not mula basis. Specifically, a special road and bridge benefit to the same extent from county government levy, a cigarette tax, and a sales tax are countywide services. In St. Louis County, however, the principal levies subject to distribution rules. The establishment services supplied by county government to residents of a uniform rate precludes competition among juris- of the unincorporated area, police patrol and resi- dictions on the tax side, but not on the benefit dential street maintenance, are supported either by side .59 taxes collected only in the unincorporated area or by Both the special road and bridge levy and the an earmarked tax source shared with the municipali- cigarette tax are closely tied to the financing require- ties. If there is a subsidy, it may be from the un- ments of municipal-type services in the unincor- incorporated area to the incorporated area of St. porated portion of the county. The first $0.18 of the Louis County .62 $0.35 authorized by law for the special road and Fiscal equivalence, nevertheless, does not imply bridge levy is divided between municipalities and the political equivalence. While being treated in effect county government according to the place of origin as a special tax and service district, the unincor- and its assessed property valuation, with the county porated areas have no special claim to representa- government receiving those revenues collected in tion in county government. Residents in unincor- unincorporated areas.60 The county Department of porated areas depend on county government as a Highways and Traffic maintains a separate account provider of basic municipal-type services, and pay for the county road system, restricted to the taxes to support those services, but they have no unincorporated portion of the county, as distin- greater ability to articulate demand for county serv- guished from the arterial road system, which is ices than the residents of municipalities. countywide. All of the county government share of the first $0.18 of the levy accrues to the separate The Sales Tax Controversy account for unincorporated areas. The cigarette tax The most controversial fiscal issue in recent St. is distributed between the county government and Louis history is the imposition of and distribution of municipalities on a population basis, with the county revenues from the countywide 1 percent sales tax. A portion earmarked for police services that primarily sales tax raises perplexing issues of fiscal equivalence serve the unincorporated areas.6' due to the fact that nonresidents are frequently a major source of sales tax revenue. Nonresidents who costs and the need for higher revenues. Both argu- do business within the boundaries of a unit of local ments have elements of validity, making the deter- government are to some extent beneficiaries of that mination of a fair distribution of sales tax revenues local government's services, but they are not voters difficult. Some point-of-sale cities are able to sub- in the taxing jurisdiction. Their influence is strictly stantially reduce taxes on residences (property tax of the pocketbook variety; if dissatisfied, they can and utilities tax), relying heavily on sales taxes for take their business e1sewhere.m As the sales tax has municipal revenue. On the other hand, most point- become the single largest revenue source for most of of-sale cities raise substantially more revenues from the municipalities in St. Louis County, arguments all sources than do most pool cities (see Chapter over whether its distribution is fair, in the fiscal Two). Further complicating the argument is the fact equivalence sense, have become increasingly vigor- that a number of small cities in the sales tax pool 0us.W that have no commercial or industrial activity within Any Missouri municipality with a population of their borders obtain significant proportions of their 500 or more, including those incorporated as vil- revenues from the pool distribution. These small lages, has been free since 1969 to enact a sales tax pool cities, even more than many point-of-sale cit- of 112 or 1 percent.= In 1977, at the urging of those ies, use sales tax revenues as substitutes for local cities without significant commercial sectors, the property taxes.71 state legislature authorized a countywide sales tax in An ability to rely primarily on sales and other St. Louis County that preempted municipal sales nonresidence taxes to supply services to residents taxes then in e~istence.~~The rate, approved by vot- can be interpreted as a subsidy of residents by non- ers in a countywide referendum, is 1 percent. In a residents, a redistribution of revenues. On the other complicated piece of legislation, a new system was hand, residents may argue that the subsidy is better established that created two means of distributing viewed as compensation for congestion and other the sales tax revenue: (1) point-of-sale and (2) pool nonamenities that accompany commercial develop- distribution.87 Point-of-sale cities receive all of the ment. From this latter perspective, suburban com- revenue from the 1 percent sales tax collected within munities may fall into one of two types: (1) those their jurisdiction. Pool cities, together with the un- that view themselves mainly as a place to reside and incorporated portion of the county, combine their which, therefore, seek to maximize the livability of sales tax collections and distribute the revenue ac- the community (subject to a cost constraint) impos- cording to population.68 ing strict zoning restrictions, for example, to exclude Initially, point-of-sale cities consisted of all those commercial activity, and (2) those that view them- that had enacted a sales tax pursuant to municipal selves, at least in part, as entrepreneurs whose prod- authority prior to the effective date of the new uct, offered "for sale," as it were, is business loca- county tax (a period of time elapsed between the tion, including a range of location-related services. legislative action establishing the system and its im- The point-of-sale distribution rule provides some plementation, allowing cities to choose to become municipalities with an incentive to compete-to give point-of-sale cities). All others became pool cities up some residential amenities in exchange for tax permanently. Beginning with the census in 1980, savings. These two types of communities will view and at ten-year periods subsequently, point-of-sale taxes quite differently. cities could choose to become pool cities, and subse- In Type One communities, taxes are collective quently, though only once, the same cities could payments for services rendered. The community is choose again to revert to their point-of-sale status. viewed as if it were a consumer cooperative for its The lack of sharing by point-of-sale cities has residents. In Type Two, taxes are a device for cap- created tension among the municipalities and be- turing the proceeds from the sale of a product-busi- tween county government and point-of-sale munici- ness location and the community is viewed in part as palities. Pool cities and the county contend that their if it were a for-profit enterprise, with the profits citizens contribute to the wealth of point-of-sale taken in below-cost services to residents. Type One communities by shopping in commercial areas that cities complain that the profits enjoyed in Type Two only happen to be located where they are.69 Point- cities derive mainly from locational rents, and that of-sale cities respond that (1) the location of a com- geographic location or prior development patterns mercial enterprise is not entirely happenstance, (e .g., extensive residential development which given that some municipalities discourage commer- leaves little or no room to accommodate commercial cial development by means of restrictive building or industrial activity) prevents them from exercising and zoning regulation,'o and (2) commercial areas the option available to Type Two communities. Type increase the demand for services, including police, Two cities respond that they are, after all, in busi- fire, and streets and, therefore, entail higher service ness: They provide a service-commercial location- and all parties to the transaction are left better off. utility revenues and other revenues from the county They are likely to discount the locational rent aspect government b~dget.~e of their revenues, and the geographic and developmental factors which contribute to that rent. Debt Rules77 In theory, a sales tax is tailor made for Type To some extent the debt rules in the local fiscal Two cities, especially if rates are uniform throughout constitution mirror the tax rules. Debt ceilings are a market area, as in St. Louis County. Jurisdictions couched in terms of a percentage of the total as- that choose to compete for commercial activity then sessed valuation of a jurisdiction. The ceilings vary compete, not on the tax side, but on the benefit among different types of jurisdictions. Villages are side for new business. To the extent that commercial limited to a total (general obligation) indebtedness development provides spillover benefits-both im- of 5 percent of the assessed valuation. Cities and mediate job opportunities and future development counties may incur an additional 5 percent debt. opportunities-that accrue to a broader economic Fire districts are not authorized to incur any general community, the entire county may be said to benefit debt beyond the current fiscal year. Additional debt from the entrepreneurial efforts of individual mu- is allowed, however, for specific, earmarked pur- nicipalitie~~2The competition is not entirely fair, poses. Cities may also incur a 10 percent debt for however, as some communities enjoy locational ad- street and sewer improvements and another 10 per- vantages and developmental opportunities (e.g., va- cent for the purchase or construction of waterworks cant land). Some communities that might wish to or electric plants.78 Any municipality, moreover, engage in this competition for commercial activity may issue revenue bonds without affecting their debt limit if the principal and interest are to repaid from are unable to exercise that choice because they lack revenue produced by a municipal utility. The pur- vacant land for development or are located at a dis- poses for which indebtedness can be incurred are tance from transportation arteries. Type One com- determined by state law in the case of counties and munities opt out of the competition entirely. municipalities incorporated under state law, but by County government planners and officials argue local charter in the case of constitutional charter cit- that professional planning provides a more rational ies. approach to economic development than does a Implicitly, going into debt is viewed as an ex- competitive dynamic, and is able to achieve an opti- traordinary fiscal measure, presumably because it mal balance of development potential and residen- creates a long-term obligation that taxpayers must tial amenitie~~3A purported lack of balance in rap- support. No unit of local government in the state idly growing unincorporated areas of the county, may incur a debt beyond the current fiscal year, combined with relaxed restrictions on annexation even within a statutory ceiling, except with the ap- and renewed interest in incorporation, has in recent proval of two-thirds of the voters, as required by the years given county government its greatest political state constitution.7Q The rule for revenue bonds is problems.74 If new annexations and incorporations somewhat more relaxed. Approval is required by occur, the predominant structure of economic de- only four-sevenths of the voters.80 These rules are velopment in the county will consist of competition consistent with the general approach of the local among municipal jurisdictions, rather than central government constitution to give citizens an opportu- planning. County revenues have and will continue to nity to veto new financial obligations. decline if this pattern continues, forcing cutbacks in county service production.75 RECENT ANNEXATION AND In 1983, reaction to new efforts at annexation INCORPORATION ACTIVITY and incorporation in St. Louis County led to state The mid 1940s to early 1950s was a period of legislation stipulating that all newly annexed and in- intense incorporation activity in St. Louis County. corporated territory in the county would remain By the late fifties and early sixties, however, the rate within the sales tax pool. This action protected the of incorporation had slowed, virtually to zero. Be- pool from potential erosion to the detriment of pool tween the Crueler decision of 1963 and Town and cities. Nevertheless, the basic dissatisfaction of the Country two decades later, only one new municipal- pool cities with the distribution formula remains. ity was formed-the City of Black Jack in 1970.8' Further, while sales tax revenues are protected in During the same period, few annexations occurred, the sense that per capita revenues of county govern- largely because of county government opposition in ment and municipalities in the pool will be unaf- court. Sometimes county officials lost, or did not fected, absolute revenues for the county will decline challenge, but infrequently so. The rate of annexa- as annexations or incorporations occur. In addition, tion is directly related to the Graeler precedent, but newly annexed or incorporated areas would subtract this case did not directly affect municipal incorporation. Drawing a connection between the two requires true, the continued enjoyment of these low-cost special care. services (to households) depends in part on the con- Those who argue that municipal incorporation is tinuation of the present sales tax distribution for- frequently defensive (i.e., to avoid annexation) mula. Voters in an area proposed for annexation might be inclined to draw a direct connection. Given must accept some risk that the favored position of that Graeler slowed the rate of annexation, the rate point-of-sale cities would be lost in the future. On of incorporation dropped also-so the argument the other hand, annexing municipalities may be would go. However, as noted above, the incorpora- more efficient units for service production and deliv- tion rate had slowed-apparently approaching zero- ery than the more heterogeneous, somewhat sprawl- prior to Graeler. A more likely explanation is found ing unincorporated area of the county. If this is so, in the development of alternatives to municipal in- changes in the sales tax distribution formula would corporation. The advent of county home rule in have a lesser effect on the bargain. 1950 brought an expanding service provision role for The county Department of Planning prepares a county government, including police protection and study and assessment of each proposed annexation. street services, as well as planning and zoning. Fire In addition to the projected impact on county gov- protection service was provided by established fire ernment revenues, the planning department also districts; sewerage, by the Metropolitan St. Louis points out features of an annexation it considers to Sewer District; and water, by a private water com- be unreasonable. Town and Country was scored for pany. Virtually all municipal services were available, its inclusion of a Western Electric Plant and from one local government unit or another, without McGraw-Hill Publishers and its exclusion of apart- municipal incorporation. Counsel for the county ments. The annexation vote succeeded. In its review government argued in Graeler that the county was a of a proposed annexation by the City of Florissant, de facto municipality in the unincorporated area, a the largest municipality in the county, the depart- position that would have completely terminated all ment concentrated on indicators of service quality annexations and incorporations. The court, how- alongside increased taxes, arguing that Florissant ever, did not accept this line of reasoning, and both could not match county government's tax-services annexation and incorporation remained as legal pos- package. The annexation vote failed. sibilities. Recently, the abandonment of the Graeler Cited by county officials as a prime example of doctrine in the course of a successful annexation of the abuse of municipal annexation powers was a re- a large tract by the City of Town and Country coin- cent annexation plan directed at a largely undevel- cided with a new burst of interest in municipal incor- oped parcel by the small City of Valley Park. The poration. area proposed for annexation bisected a subdivision under development, excluding the developed por- Annexation Actlvity tion by annexing along one side of a road. The area Many municipalities are apparently able to offer included 22 registered voters, only one of whom areas considered for annexation a tempting bargain, chose to vote. The annexation vote succeeded on namely increased service levels at the same, lower, the basis of this one citizen's vote.83 A proposed an- or only slightly higher tax burdens. Annexing mu- nexation by the point-of-sale City of Fenton was op- nicipalities often have low property tax and utility posed mainly on grounds of current high-quality rates, and may be in a position to reduce trash col- service by the county. The annexation vote failed. lection expenses in areas targeted for annexation. Despite an initial rush to file annexation plans, Annexing cities with their own police departments successful annexation activity in the county has so can usually promise faster response times to resi- far been modest. Town and Country added only dents in areas currently served by the county police. 3,356 residents in its annexation. Nine other munici- Whether the difference in tax-service packages palities have annexed small parcels, but only an- between the unincorporated county and municipali- nexations by Hazelwood and Florissant (3,057 and ties proposing annexations is due to a lack of fiscal 1,382 persons respectively) have included in total equivalence or to differences in the efficiency of more than 1,000 people. The remainder included service provision is difficult to determine. Mis- approximately 400 persons added to seven commu- matches between revenues and services may exist nities.84 Many proposed annexations have been re- between different parts of the unincorporated area jected by voters in the target areas. as well as between parts of the unincorporated area Several of the successful annexations, however, and a particular municipality. The annexation bar- have involved the movement of commercial and in- gain may be explained by the status of annexing mu- dustrial activity from the tax rolls of county govern- nicipalities as point-of-sale cities, able to keep their ment to those of annexing municipalities. County own sales tax collections.82 To the extent that this is government is concerned that annexations (together with new incorporations as discussed below) will erode the county government's tax base for financing services in unincorporated areas, as well as di- Interest in municipal incorporation has also in- minish its ability to control new development. When creased in recent years. A new City of Maryland an area is annexed (or incorporated), county gov- Heights was incorporated in 1985, as a third class ernment loses revenues that are distributed on a per city, with a population of 26,4 13. The first incorpo- capita basis (e.g., sales taxes, cigarette taxes, and ration since 1970, it was adamantly opposed by state highway funds), as well as losing revenues from county government, both in the incorporation elec- utility taxes in the area. The reduction in revenues tion and in the courts. When the incorporation and service areas has a direct impact on the county's cleared its last legal hurdle, county government with- urban service departments, police and highways in drew its services, including police, before the new particular, and potentially an indirect effect on other city could organize. Initially, Maryland Heights was departments that supply services countywide. patrolled by the state police. Incorporation brought From a service provision standpoint, the county no tax increases. Maryland Heights has no property government position is not the bottom line. While tax; its tax revenues consist mainly of a 5 percent revenues and service requirements do not match ex- utilities tax (equal to the county rate collected in un- actly, annexations and incorporations nevertheless incorporated areas), its share of the sales tax pool withdraw both service responsibilities and revenues (which the city was required by state law to join), simultaneously.~~One principal effect on the county and its share of other county and state revenues. is to require county government to make adjust- Included in the new city's boundaries is a highly de- ments in personnel and, eventually, equipment as veloped commercial area called Westport, previously needed to meet its remaining service responsibilities. a target of unsuccessful annexation attempts by two The net effect is a transfer of responsibility from adjacent cities. With little doubt, Maryland Heights county government to municipal governments. would have become a point-of-sale city had state law There is no a priori basis for judging this transfer to allowed it. be detrimental, either to those populations directly In November 1986, another incorporation issue affected or to the county as a whole.86 appeared on the ballot, to decide whether to create The county government is also concerned that a City of Chesterfield in the western part of the annexations and incorporations will leave isolated is- county. The incorporation issue was narrowly de- lands of unincorporated area to which it would be feated, receiving 48.6 percent of the voters' support. difficult to deliver services. The county has had a In December, however, the St. Louis Post-Dis- few such islands for a number of years (e.g., patch reported that supporters of incorporation were Elmwood Park and Meacham Park). In these par- again circulating petitions, but within a more circum- ticular areas, adjoining municipalities have assisted in service delivery, either without charge or through scribed set of boundaries. In April 1988, voters ap- a contract with county government, but the creation proved creation of a smaller City of Chesterfield (es- of many more such areas, especially if they were to timated population 33,000) by a 3-1 margin. In be low revenue, high service demand islands, could South County, citizens of the Oakville area also were be detrimental to county government service provi- exploring the possibility of creating a city as the field sion. work for this study was under way in 1986. Some municipal officials contend that the While this surge of interest coincides with a county increases its service levels in areas proposed change in the annexation rules, it is difficult to char- for annexation, an accusation denied by county offi- acterize these new efforts at incorporation as mainly cials. Whatever the methods used, the county has defensive. The queuing of incorporation and an- not been without significant success in opposing annexation proposals for a given parcel does create an nexations in the immediate post-Town and Country incentive to move quickly and may offer the appear- period. Yet county officials remain concerned. ance of defensive incorporation. One could as easily Clearly, the new annexation rules have exposed argue, however, that proposed annexations are de- county government to greater competition. Even the fensive, in anticipation of an incorporation. Incorpo- best prospect for county government may be no bet- ration and annexation proposals are interrelated ter than a slow, gradual erosion of its municipal serv- phenomena. Instead of viewing either one as mainly ice role. The worst prospect stems not from annexa- a response to the other, a more plausible explanation directly, but from the potential incorporation of tion is that substantially the same issues are driving the major portion of the county's direct service area. both annexation and incorporation. From this perspective, the overriding issue is as unresponsive to residents' desires to preserve the growth. This was especially apparent in the case of livability of their communities. In spite of their re- the new City of Chesterfield. Rapid commercial de- peated protests, planning and zoning decisions pro- velopment in this area clogged roads and eroded ceeded as before. County government established a residential amenities. County government was per- new service center in the Chesterfield area, but in- ceived by many residents in the area as pro-growth. terest in incorporation continued. Incorporation gives municipal officials full control of Responsiveness may be related to patterns of planning and zoning decisions. County government, representation. Citizen activists in Chesterfield re- together with the state highway department, how- ported little support from their county councilman in ever, retains jurisdiction over arterial streets. Never- articulating their concerns and turned instead to theless, the pace and character of development falls their state legislators-who represent much smaller under the regulation of the new city. districts than a county councilman-to obtain politi- Citizen leaders in the Chesterfield area antici- cal advocacy and support for their position. Incorpo- pated no need for residential tax increases to fi- ration brings with it an entirely different pattern of nance municipal services, even though the new city representation, with a much lower citizen-to-repre- would be restricted to participation in the sales tax sentative ratio. Citizens concerned about responsive- pool. In effect, citizens can simply choose to with- ness can be expected to take this difference into ac- draw from county government service provision, count. take their tax revenues with them, and establish Organized opposition to incorporation in Ches- their own local government. From the county gov- terfield emerged in border areas of the proposed ernment perspective, incorporation is tantamount to city. The collective decision rule for incorporation, secession-not secession from the county, but seces- recalling the discussion above, is a simple majority of sion from the county-as-municipality (i.e., from the those voting. Once incorporated, the procedures for county government's municipal services area). disincorporation are much more exacting. The Proponents of the Chesterfield incorporation Chesterfield incorporation as initially proposed was also argued that the new city would be in a better quite large by St. Louis County standards and was position to respond to growth by investing in appro- not uniformly supported throughout the area.87 priate infrastructure. County government efforts to Some citizens in areas adjacent to existing munici- raise money from countywide bond issues to finance palities may have preferred annexation, but the infrastructure development were unsuccessful be- neighboring municipality had not initiated an an- tween 1977 and 1986, despite overwhelming support nexation procedure in time. It could be argued that from voters in the rapidly growing areas of the the incorporation of a city is too easy, and that an county, such as Chesterfield. In effect, the county extraordinary majority ought to be required. Such a jurisdiction may be too large to aggregate the effec- rule might lead incorporators to draw their proposed tive demand for service provision in the Chesterfield set of boundaries more circumspectly, not including area. Voters can be expected to support new taxes fringe areas or subsections where majority support is when they derive a greater benefit, and oppose new perceived to be lacking (except where the contiguity taxes when the benefits flow primarily to others. rule would make exclusion difficult). Within such a Public demand is most easily aggregated when the smaller set of boundaries the Chesterfield incorpora- jurisdiction able to act is closely matched with the tion was approved by a 3-to-1 margin. population and territory where the demand exists. In 1986, the county was successful in obtaining voter CONCLUSION support for a $75 million road bond issue. Its suc- Constitutional choice in St. Louis County is a cess, after prior voter refusals to endorse such bi-level process, consisting of an enabling level and a bonds, may be in part a result of having established chartering level, linked by a set of rules of four special taxing districts and trust funds to finance types: rules of association, boundary change rules, road improvements in the Chesterfield area. If so, contractual rules, and fiscal rules. In the enabling this is a case where citizens will support increased process, the county delegation to the state legisla- taxation in circumstances where fiscal equivalence is ture, responding to local initiative and based on sub- maintained. stantial public agreement in the county, employs the The issues driving incorporation can also be un- device of special local legislation to make rules gov- derstood in terms of service responsiveness. Com- erning the creation and empowerment of local gov- munity activity in the Chesterfield area initially fo- ernments. Local citizens as voters in various refer- cused on an effort to secure greater attention from enda are the principal decisionmakers in the process county government for local concerns. County gov- of chartering local governments, as well as modifying ernment was perceived by some community leaders charters and boundaries. Conventionally, the analysis of the rules treated cial district formation are organized through local in this chapter would be based on a conception of citizen petition and election. Annexation requires two levels of government-state and local-in which majority approval in the area to be annexed. Con- one level is dominant-the state. Rules such as those solidation may take place only when approved by which establish tax ceilings or impose restrictive con- concurrent majorities, including any modification in ditions on annexation are viewed as limits imposed the relationship between St. Louis City and County. by a higher level of government on a lower level. (The rule for school district reorganization is differ- Two points are ignored in this conventional mode of ent, as discussed in Chapter Seven.) Especially since analysis: (1) state legislators represent local citizens, the 1983 Missouri court decision in effect removing not the state, and (2) most of the limits tend to em- a county government veto over proposed annexa- power local citizens while constraining local officials. tions, citizens throughout the county have been able Many of the limits on local government embodied in to choose from a broad menu of possibilities for state legislation are in fact an allocation of legal constituting and reconstituting local government ar- power to local citizens acting in their local communi- rangements. Whether a city-county board of free- ties. The function of this empowerment is to condi- holders may make changes within St. Louis County tion the actions of local officials on their being able that would not require concurrent approval by citi- to satisfy the preferences of local citizens. zens of the local governments affected is a subject of What would otherwise be viewed as a bewilder- current controversy. As suggested in Chapter One, ing array of occasionally contradictory rules makes the principle of local self-government may be viewed sense from a constitutional perspective. The basic as fundamental to the operation of a jurisdictionally rules of association in St. Louis County are relatively fragmented metropolitan area. permissive, in the sense of allowing fairly easy incor- While the legal tools for (1) transfer of munici- poration of municipalities and formation of special pal jurisdiction to county government, (2) consolida- districts-freedom to organize a local unit of govern- tion of municipalities, school districts, and fire dis- ment. At the same time, the fiscal rules are quite tricts, and (3) dissolution of municipal and special restrictive, in the sense of limiting the fiscal discre- district governments are available to local citizens, tion of local governments, or at least making the the chosen mode of adjustment in service provision process of raising revenue somewhat difficult. Free- arrangements tends to be cooperation among units, dom on the part of citizens to organize local govern- whether through contracting, joint service agreements does not imply freedom on the part of offi- ments, or the creation of new, overlapping special cials to tax. Similarly, ease of incorporation on the districts. These functional arrangements among units part of citizens is not accompanied by ease of an- of local government are the focus of discussion in nexation on the part of officials. The point of these Chapters Four through Seven. rules is not to make life easy for officials, but to give ENDNOTES citizens an opportunity to say no. Viewed as a single 'The original expression of this concept is widely known as configuration with the purpose of empowering local Dillon's Rule. In 1868, Judge John F. Dillon wrote that citizens, the various rules in the local government "any fair, reasonable, substantial doubt concerning the ex- constitution are coherent. istence of a power is resolved by the courts against the [local government]. " Merriam v. Moody's Executors, 25 This configuration of rules, together with the Iowa 163 at 170 (1868). processes that generate those rules, comprise the 2~ distinction can be made, however, between a legal subdi- functional arrangements for metropolitan govern- vision and its government. One can be a state creation, the ance in St. Louis County. In this form, metropolitan other a local creation. St. Louis County and St. Louis governance is consistent with a highly fragmented County government are not the same, though the boundaries are the same. Both St. Louis County and its govern- pattern of local governments. By explicitly address- ment, however, can be viewed as local creations. St. Louis ing the constitutional level of analysis, a metropoli- County was created with its present boundaries by the act of tan area that otherwise appears to be unorganized city-county separation. approved by local voters in 1876. St. Louis County government operates under a home rule on an areawide basis is found to be organized ac- charter adopted by county voters in 1950. cording to a set of rule-ordered relationships estab- 3~avidR. Reynolds, in his study of metropolitan St. Louis, lished on a democratic basis. The local government Los Angeles, and Detroit, made this same point: "Although constitution defines a decisionmaking arena in St. municipalities in the three areas derive their legal existence and powers from the constitutional and statute law of the Louis County that is both highly salient and very ac- respective states, it is not true in any literal sense that they tive. are the creations of state government. Instead, it is more The rules of association and boundary change correct to think of state government as providing the legal in the local government constitution consistently milieu which enables the existence of municipalities or, for that matter, of any unit of local government. The actual adhere to a principle of local self-determination or creation of municipalities in California, Missouri and self government. Municipal incorporation and spe- Michigan is reserved to the voluntary collective actions of 'citizens' in local areas." Reynolds, "Progress toward 16chapter 77, Missouri Revised Statutes. Achieving Efficient and Responsive Spatial-Political Sys- tems in Urban America, " in John S. Adams (ed.) , Urban 17chapter 78, Missouri Revised Statutes. Policymaking and Metropolitan Dynamics: A Comparative l8The lack of symmetry between the procedures for incorpo- Geographical Analysis (Cambridge, Mass. : Ballinger Pub- ration and disincorporation deserves comment. Incorpora- lishing Company, l976), pp. 470-471. tion by a simple majority of those citizens voting in an elec- 4Reynolds, pp. 485-486. tion establishes a relatively low barrier to the entry of new cities. Yet constitutional decisions of this order, creating a Swillis D. Hawley suggested that the procedures established government, are frequently associated with a requirement of by state law to govern changes in the powers and jurisdic- extraordinary majorities in order to promote consensus, as tions of local governments be thought of as "rules for con- in the case of villages. On the other hand, disincorporation stitution building. " Hawley, "On Understanding Metropoli- is potentially a very costly step, requiring liquidation of a tan Political Integration," in Willis D. Hawley, et a]., community's corporate assets. Petition by two-thirds or Theoretical Perspectives on Urban Politics (Englewood three-fourths of the legal voters can be a very high barrier to Cliffs, New Jersey: Prentice-Hall, 1976), p. 133. dissolution. Depending on the absolute number of voters involved, a petition rule creates a higher barrier than a voting %ee Larry Kiser and Elinor Ostrom, "The Three Worlds of rule. Gathering valid signatures on a petition is a time- Action: A Metatheoretical Synthesis of Institutional Ap- consuming, costly process. This means that a municipal proaches," in Elinor Ostrom, ed., Strategies of Political government can continue to exist (except for the possibility Inquiry (Beverly Hills, California: Sage Publications, of consolidation) with a relatively low level of support from 1982), pp. 179-222. its citizens. In St. Louis County, however, disincorporation 7~hisprocess, sometimes called "legislative courtesy, " char- might be viewed as a positive constitutional choice to asso- acterizes the consideration of local bills by state legislatures ciate with the unincorporated area of the county served by in perhaps one-third or more of the states. See Advisory county government. The county would have no choice but to Commission on Intergovernmental Relations (ACIR), accept such a decision. Measuring Local Discretionary Authority (Washington, 19chapters 79, 80, Missouri Revised Statutes. D.C. : ACIR, Report M-131, November, 1981). 20~ldermenin Kinloch were circulating a disincorporation pe- %ee James M. Brasfield, "The Great Land Rush: Annexa- tition over the objection of the city's mayor. The motivation tion and the Sales Tax in St. Louis County," a paper pre- for disincorporation is extreme community poverty, aggra- sented at the 1985 Annual Meeting of the American Politi- vated by Kinloch's location near Lambert International Air- cal Science Association, New Orleans, Louisiana, and the port. Recent airport expansion resulted in the relocation of discussion of the sales tax controversy later in this Chapter. hundreds of Kinloch residents out of the city and a further decline in the city's tax base. Those seeking disincorpora- 9~eethe discussion of boundary review agencies in ACIR, tion argue that the community would be better off as an un- The Organization of Local Public Economies (Washington, incorporated part of St. Louis County; those opposed argue D.C.: ACIR, Report A-109, December 1987), pp. 39-40. that disincorporation alone would do little or nothing to ad- lostatutory rules pertaining to municipal incorporation are dress the community's problems and could result in a de- found in Chapter 72 of Missouri Revised Statutes. For a cline in some service levels, such as police protection. See discussion, see Chapters 1 and 13 of Missouri Local Gov- Joseph Schuster, "Urban Suicide: A Depressed Kinloch ernment Law (Jefferson City: Missouri Bar, 1975, supple- Considers Taking Its Own Life," The Riverfront Times, July ment 1984). 16-22, 1986, p. 1. "A community of more than 200 residents already incorpo- 21~ora general discussion, see Richard R. Dohn, "Special rated as a village may also choose to become a fourth class Districts in Missouri and the Nation," in Richard J. Hardy city. and Richard R. Dohn (eds.), Missouri Government and Politics (Columbia: University of Missouri Press, 1985), 12~rticleVI, Section 19, Missouri Constitution. pp. 173-184. 13see Chapter 80, Missouri Revised Statutes 22~eeChapter 321, Missouri Revised Statutes. For a discussion see Chapter 12, Missouri Local Government Law. I4~hedifferences between cities and villages are worth not- ing. First, villages-the smaller unit-require a higher level 23~hiscould become relevant in a new municipality previ- of agreement within the constituting community. Secondly, ously organized as a fire district. county government can veto the incorporation of a village, 24~ountyclassification is based on the total assessed valu- but not a city. This can be construed as requiring a higher ation. level of agreement outside the constituting community. Yet areas proposed for incorporation as cities are more likely to 25~eeRobert A. Cohn, The History and Growth of St. Louis contain heterogeneous communities, as well as more likely County (Clayton, Mo. : St. Louis County Office of Public to affect other units of government adversely. In general, Information, 1974). the logic of constitutional choice suggests that higher levels 261n the language of contemporary political and administra- of required agreement would tend to be associated with tive theorist Vincent Ostrom, the rivalry is between "bu- higher levels of potential controversy. See James Buchanan reaucratic" administration and "democratic" administra- and Gordon Tullock, The Calculus of Consent (Ann Arbor, tion. Vincent Ostrom, The Intellectual Crisis in American Michigan: The University of Michigan Press, 1962). The Public Administration (University, Alabama: The Univer- differences between villages and cities in St. Louis County sity of Alabama Press, 1974). do not seem to be congruent with this logic. Instead, the 27~olicetraining standards were adopted countywide, how- rule structure seems to introduce a bias in favor of city, ever, through cooperative efforts of area police leaders rather than village, formation. If the same decision rule which resulted in state legislation applicable to St. Louis used for villages-two-thirds of the taxable inhabitants- County. For an account, see James C. McDavid, Police were applied to cities, city formation would become much Cooperation and Performance: The Greater St. Louis more difficult than village formation, due to the greater Interlocal Experience (University Park, Pennsylvania: The (probable) heterogeneity of an incorporating city. Pennsylvania State University Press, 1979), especially pp. 15chapter 80, Missouri Revised Statutes. 61-64. 28~hapter70, Missouri Revised Statutes. basis of deed covenants to provide services to residents. In St. Louis County, and to some extent in St. Louis City, sub- %ee Chapter 11, Missouri Local Government Law. divisions are alternative providers of street services (see 301f a municipality agrees to deliver service to another at a Chapter Six). Subdivisions have no tax powers; thus their price that equals or exceeds its marginal cost, yet is below only ability to derive revenue is from user fees and house- its average cost, the net effect is to reduce overall average hold assessments. While private covenants could be written cost of production. Such a situation would arise, for exam- to employ an equivalent to the property tax (by indexing the ple, when there were remaining economies of scale which subdivision fee to assessed valuation), none of the subdivi- could be obtained by the municipality supplying service. sions studied employ such a method. Instead, the most common method is a flat fee per household, followed by re- ='see generally Chapter 72, Missouri Revised Statutes. liance on a frontage or acreage based fee. These "tax" %ee discussion in Chapter 13, Section 13.8, Missouri Local mechanisms are unambiguously benefit based, i.e. , the tax Government Law. rate is determined by the cost of services provided. Vari- ation in the rates among subdivisions can be interpreted as 33~ayneMillsap, Chairman of the Board of Freeholders, price differentiation based on real differences in the value of personal communication, October 21, 1987. A board of services provided. freeholders was recently constituted in St. Louis, for the purpose of exploring local government consolidations within 48~eeRobert H. Edelstein, "The Fiscal Crisis of Cities: An St. Louis County (see Chapter Nine). Economist's Perspective, " in David Solomons, ed. Zmprov- ing the Fiscal Performance of States and Cities (Reston, 34~ohn,The History and Growth of St. Louis County, p. 62. Virginia: The Council of Arthur Young Professors, 1979), 35~hisaccount is based on Chapter 13, Section 13.9, Mis- pp. 139-160. souri Local Government Law, supplement 1984. 49~hateverthe choice of tax instruments, there may be im- %ee the discussion in Mark Schneider, Suburban Growth: portant advantages when local government units within the Policy and Process (Brunswick, Ohio: King's Court Com- same metropolitan region operate with substantially the munications, 1980), pp. 34-42. same set of instruments. If proximate jurisdictions employ tax instruments with widely varying patterns of incidence, 37~herewere two Graeler decisions, City of Olivette v. incentives are created for inefficient location decisions Graeler, 338 SW 2d 827 (Mo 1960) and City of Olivette v. solely to take advantage of incidence effects. See Wallace Graeler, 369 SW 25d 85 (Mo 1963). E. Oates, Fiscal Federalism (New York: Harcourt, Brace, %t. Louis County Department of Planning, Report of the St. Jovanovich, 1972), especially pp. 145-147. Louis County Annexation Study Commission. July 1985, 50~anyeconomists argue that revenue or income redistribu- p. 1. tion is best accomplished by jurisdictions with the broadest 39~hisprocedure, like several other features of the local gov- constituency possible, e.g., a national government (see ernment constitution in St. Louis, applies to St. Louis Oates, Fiscal Federalism, p. 8, for a typical statement). County only (Section 321.650, Missouri Revised Statutes). The American reality, however, is that redistribution is a phenomenon that characterizes overlapping jurisdictions at 40~eoffreyBrennan and James Buchanan, The Power to all levels of government. In a recent paper, Robert M. Stein Tax: Analytical Foundations of a Fiscal Constitution (New and Keith E. Hamm provide evidence that some state gov- York: Cambridge University Press, 1980). ernments may do better at redistributing revenues among 41~rticleX, section 11, Missouri Constitution. their local governments than has the federal government. 42~eeRhonda C. Thomas, "Recent Developments in Mis- See Stein and Hamm, "A Comparative Analysis of the Tar- souri: Local Government Taxation, " UMKC Law Review geting Capacity of State and Federal Intergovernmental Aid 49 (1981): 491-533. Allocations: 1977, 1982," Social Science Quarterly 68 (September 1987) : 447-465. 43~ora case study of interaction between the state legislature 51~onsumptiontaxes (e.g., sales taxes on goods, gasoline, and local officials in the context of a change in local fiscal or cigarettes in Missouri) can have redistributive effects rules, see James M. Brasfield, "The Great Land Rush," even in the absence of explicit attention to such rules. For 1985. example, distributing sales tax revenues on a per capita ba- 44For a discussion of this concept, see Mancur Olson, "The sis, as in the pool distribution formula (see below), most Principle of 'Fiscal Equivalence': The Division of Respon- likely results in a net transfer of revenues to lower income sibility among Different Levels of Government," American communities. Economic Review 59 (May 1969), pp. 479-487. 52~tatelaw authorizes cities to tax up to these limits estab- 45The concept of fiscal equivalence, thus, is similar to the lished by virtue of Article X, Section 11 (b), of the Missouri concept of individual economic choice. In the private econ- Constitution. omy, choices made by individuals are a function of their 53~hislimit is imposed by state law (Section 80.460, Mis- preferences for goods and services and of their budget con- souri Revised Statutes). straints. In a collective choice context, fiscal equivalence implies that communities, through collective expressions of M~ominally,the tax ceiling is $1.25 per $100 assessed valu- preference, choose public goods and services that they both ation. Any increase up to $3.75, however, regardless of the prefer and can afford. Just as private choice is made under current rate, requires approval by a simple majority of vot- constraint, so too is collective choice. ers in the district. Moreover, as discussed below, two-thirds of the voters may approve excess levies without limit. See 46~eeBruce W. Hamilton, "A Review: Is the Property Tax a Section 164.021, Missouri Revised Statutes, and Article 7, Benefit Tax," in George R. Zodrow, ed., Local Provision Section 11 (c)of the Missouri Constitution. of Public Services: The Tiebout Model after Twenty-Five 55~hetherlocation shopping on the basis of housing cost and Years (New York: Academic Press, 1983). For an alterna- service package is a reality is controversial. The first expo- tive point of view, see George R. Zodrow and Peter nent of this view. Charles Tiebout, in his article, "A Pure Mieszkowski, "The Incidence of the Property Tax: The Theory of Local Expenditures," Journal of Political Econ- Benefit View versus the New View," in the same volume. omy 64 (October 1956): 416-24, argued that this mecha- 470ne type of jurisdiction lies outside the scope of the fiscal nism would allow citizens dissatisfied with a given jurisdic- rules described here, that is, subdivisions organized on the tion's package to "vote with their feetw-move to a different jurisdiction with a more satisfactory package. If there were unincorporated areas does not guarantee fiscal equivalence. a large number of local jurisdictions, Tiebout argued, a One would need to carefully analyze both revenue sources sorting process would be available to enhance fiscal equiva- and expenditure patterns to reach a determination on lence. Another view is that this "exit" option, while present whether such equivalence is being maintained or whether to a degree, is less important in understanding local govern- redistribution between the unincorporated and incorporated ment finance than is the exercise of "voice," whereby dis- parts of the county occurs-and in which direction. The St. satisfied citizens can lobby for tax or service changes and Louis County Annexation Study Commission reported that, attempt to select local officials who will reflect their prefer- in 1983, the unincorporated county contributed 37 percent ences when making provisioning decisions. In reaching this of county revenues while service expenditures there were 28 conclusion, we take into account the relatively low propor- percent of the total, indicating that the unincorporated ar- tion of the housing service cost package which local service eas were disproportionately supporting countywide services costs represent in most local jurisdictions, the barriers to in that year. Report of the St. Louis County Annexation mobility caused by income differences and real estate mar- Study Commission (Clayton, Missouri: St. Louis County kets, the many extra-service cost factors which influence Department of Planning, July 1985). citizens' locational choices, and the relative inefficacy of exit threats as opposed to the exercise of voice when at- 63~urther,the level and quality of local public services are tempting to influence local officials. For a development of generally small influences on their decisions. Only in cases this argument, see Ronald J. Oakerson, Roger B. Parks, where shoppers feel unsafe or are unduly burdened by traffic and Henry Aaron Bell, "How Fragmentation Works," a pa- congestion are local public service packages likely to have a per presented at the 1986 Annual Meeting of the Midwest significant influence on their choice of where to shop. Political Science Association, Chicago, Illinois. The argu- 64The sales tax has increased in importance as a revenue ment, in summary, is that citizens are more likely to at- source due to the limits placed on other revenue sources by tempt to influence service benefit-cost bundles by remaining the Hancock Amendment and by roll back provisions in within their communities and exercising their voice in con- state statutes. Sales tax revenues increase as prices andlor junction with like-minded residents than they are to move to sales volumes increase without a rate increase. Sales tax an alternative jurisdiction in search of a more attractive revenues can increase without voter approval in contrast to package. Given this, opportunity to exercise voice becomes increases in revenues from most other sources. a critical factor in understanding the workings of a fragmented area. 85%. Louis County, along with the City of St. Louis, is also authorized to impose a "transportation sales tax" up to 112 56~ection164.021, Missouri Revised Statutes. percent. Both jurisdictions use this revenue, in part, to sup- 57~rticleX, Section 22, Missouri Constitution. For a discus- port the mass transit services of the Bi-State Transit sion, see Thomas, "Recent Development in Missouri: Lo- Authority on a metropolitan wide basis. cal Government Taxation." Also see, Chapter 4, Missouri 66~description of the legislative process leading to adoption Local Government Law, supplement 1984. of the current sales tax distribution formula can be found in 5811 is interesting that the Hancock Amendment applies to Brasfield, "The Great Land Rush," 1985. See also Andrew fees as well as to taxes. The immediate motivation was to D. Glassberg. "The Management of Uncertainty: The St. preclude local officials from shifting from tax finance to fee Louis County Sales Tax," in Carol W. Lewis and A. based finance. Fees, of course, are the most precise form of Grayson Walker, eds., Casebook in Public Budgeting and benefit financing available. But in cases where individuals Finance (Englewood Cliffs, New Jersey: Prentice-Hall, have no choice in using public facilities, or where individu- 1984), pp. 73-76. als lack close substitutes, fees take on a coercive character 67~ections66.600-66.660, Missouri Revised Statutes. much like taxes. Simply because a government employs a user charge, it does not follow that the price charged is mar- 68~hissystem discriminates against county government acting ket regulated. Voter approval of fees may therefore impose in its capacity as a "municipality" for the unincorporated a useful constraint on local officials. The prohibition of fee area. It is the only unit of government without a choice in increases without voter approval has had mixed results in the matter. The ability of pool cities to benefit in the aggre- Missouri, with marked increases in fees in some cities gate from this arrangement depends upon a redistribution of (e.g., St. Louis and Kansas City) and relatively little 'revenue from the unincorporated area to the pool cities. growth in others (e.g., St. Joseph, Independence, Floris- Point-of-sale cities are not required to share their revenues sant , and, especially, Springfield). See Elaine B . Sharp as county government is. and Dennis Elkins, "The Impact of Fiscal Limitation: A 6Q~ommercialor industrial location is, critics of the point-of- Tale of Seven Cities," Public Administration Review 47 sale distribution argue, influenced by the location of high- (September-October 1987): 385-92, especially Figure 2, ways, the airport, large parcels of land suitable for such ac- p.388. tivities, and other factors having little to do with the efforts 59~nargument can be made that uniform rates across multi- of point-of-sale cities to attract such activity. ple jurisdictions represent a form of economic collusion 70~tleast two of the larger pool cities explicitly rejected the among local government officials to hold tax rates at an ar- location of regional shopping centers within their borders. tificially high level. Of course all of the taxes in question, See James M. Brasfield, "Follow the Money: Fiscal Struc- and any tax increases, require countywide voter approval. ture and Metropolitan Reorganization in St. Louis County," Still, this requirement is not nearly as stringent as separate a paper presented at the 1987 Annual Meeting of the approval from voters in each jurisdiction. American Political Science Association, Chicago, Illinois. 60~ection137.558, Missouri Revised Statutes. This formula 710f the eight municipalities in St. Louis County that levy no applies exclusively to St. Louis County. property tax on residential property, seven are pool, not elsection 66.340, Missouri Revised Statutes. This section point-of-sale, cities. applies to first class charter counties only. A cigarette tax is 72~he1982 Censuses of Retail and Wholesale Trade and of not, of course, a benefit tax, but is simply a politically ex- Selected Service Industries report approximately 45 percent pedient way to augment revenue. of jobs in these areas to be located in identified point-of- 62The collection of specific taxes in the unincorporated area sale cities, nine percent in pool cities, and the remainder in and the earmarking of particular revenues for services in the the balance of the county. 73St. Louis County has experienced very strong economic eo~rticleVI, Section 27, Missouri Constitution, and Section growth in the presence of this competitive dynamic (see 91.010. Missouri Revised Statutes. Chapter Eight). Whether that growth has been "balanced" 81~heBlack Jack incorporation was quite controversial, as it and whether it might have been even greater under a system was in part motivated by a desire to exclude low income of county planning remain as issues, however. housing from the area. Without this partial anomaly, one 740ne of the salient issues has been the county's ability to could say that government organization in the county was support services, particularly road services, to unincor- essentially unchanged from 1960 to 1983. porated areas that have experienced rapid commercial 82~01all annexation proposals in recent years have been growth. In response the county has been successful in ob- made by point-of-sale cities, but a significant majority ap- taining voter support of a $75 million road bond issue, and pear to be so. three road trust funds, financed from developer exactions, have been established in the Chesterfield area to provide 83~hortlyafter the successful annexation vote, this citizen road improvements. moved out of the annexed area (Donald Clark, personal communication, August 27, 1987) 75~hecounty will have to cut back, unless, as other counties have done when facing similar circumstances (e.g., Los B4st. Louis County Department of Planning, St. Louis Angeles County, beginning in the immediate post-World County 1986 Fact Book, 1986, p. 114. War Two era and continuing today), the county were to en- 85Howeve1, most annexations would not be proposed unless gage in a vigorous program of service production for munici- the annexing community believed that revenues from the palities through the medium of service contracts. annexed area were at least sufficient to cover service costs 7e~npart as a response to this, the county executive, the local there. Where revenues exceed service costs by a significant citizen group Confluence St. Louis, and the St. Louis amount, the annexing community may reduce taxes and County Municipal League have advocated alternative gov- other revenue generation in the annexed area, thus reducing ernmental structures, including complete incorporation for total local government revenues countywide. This would the entire county. A board of freeholders, with representa- represent a move toward fiscal equivalence (see the discus- tion from both the City and County of St. Louis, is currently sion of fiscal rules above) in that some revenues from the considering this question, with the intent of developing a annexed area, prior to annexation and any reduction in lo- proposal to present to voters in a referendum during 1989 cal government revenues from this area, had been used to (see Chapters Two and Nine). support service costs in other parts of the county. 77~eegenerally Article V1, Section 26, Missouri Constitution, 86~oavert such adjustments in personnel and equipment, the and Chapter 95, Missouri Revised Statutes. For a discus- county government might, as governments like that of Los sion. see Section 4.22, Missouri Local Government Law. Angeles County have done in similar circumstances, engage in a more vigorous program of service delivery to munici- 78~ounties,cities, and villages with a population of fewer palities via contracts. If significant economies of scale exist than 40,000 persons may incur an additional 10 percent in the delivery of some services (a point which is frequently debt for the purpose of acquiring and furnishing industrial debated by urban economists), then the county would ap- plants. pear to have a competitive advantage in such contracting. 79~neffort in 1968 to amend the constitution, reducing the 87The St. Louis Post-Dispatch reported that border areas op- required majority to 60 percent, failed. posed incorporation by as much as 73 percent.

Chapter Four Police Services

INTRODUCTION government and funded, for the most pan, with earmarked revenues collected in unincorporated areas Police services are among the most important of (see Chapter Three). The county police are gov- all local public services to Americans.' Among local erned directly by a board of police commissioners, agencies, police departments rank second only to appointed by the county executive with the approval schools as public employers, and police budgets gen- of the county council and the circuit judges of St. erally comprise the largest single expenditure of gen- Louis County. The commissioners select the super- eral purpose local governments that do not operate intendent of police, the department's chief execu- school systems. Making provision for police services tive officer. is predominantly a local responsibility, and a signifi- The City of St. Louis is policed by the St. Louis cant majority of Americans report that their local metropolitan police department. Provision arrange- police do a good job.2 They also report a strong ments governing this department are quite interest- preference for maintaining local control of the po- ing, perhaps unique in the United States. The lice. In all of these respects, St. Louis area commu- department is governed by a board of police com- nities are typical. missioners appointed by the governor of Missouri- St. Louis County municipalities have arranged the mayor of St. Louis is an ex officio member of for the production of police services in ways similar this board.3 Formally, the commissioners have to most cities. A majority of the municipalities have authority to make decisions regarding the size and organized their own local police department, with a funding of the department, and submit a budget bureau chief accountable directly to local elected of- backing these decisions that must be funded by the ficials. A substantial minority of the municipalities, city.4 Although the board of police commissioners especially the very small ones, contract for service may make policy decisions without formal need for with an adjoining municipality or with the county po- agreement from city officials and citizens, such lice. In these cases, local elected officials serve as agreement is frequently sought and given. As in St. purchasing agents for their communities. Louis County, the appointed commissioners select Provision arrangements are somewhat different the city department's chief executive officer, the in both the unincorporated county and the City of chief of police. St. Louis. Although in each case, police protection is produced by a locally organized police depart- ARRANGING FOR THE PRODUCTION OF ment, an intermediate body has been interposed be- POLICE SERVICES tween elected officials and the bureau chiefs in A basic decision for a local provision unit (e.g., charge of police production. In both instances, the a municipality) is to choose how to arrange for the avowed intent of this interposition has been to sepa- production of a service. By far the most common rate the police from political influence. arrangement in St. Louis is in-house production-the Unincorporated St. Louis County is policed by supply of police services by a full-time bureau of the the county police department, organized by county local government (see Table 4.1) The City of St. Table 4.1 Provision of Policing in St. Louis Provision Units Units With Own Police Units That Contract With: Population Population Other Muni- Number (1984 estimate) Number (1984 estimate) County clpalltles None Municipalities1 in St. Louis County 89 589,983 62 566,810 72 17 32 With Population of: 0-399 9 1,924 0 0 1 5 3 400-1,000 13 8,902 1 71 5 5 7 0 1,001-2,000 14 22,944 9 15,513 0 5 0 2,001-5,000 17 56,718 17 56,718 0 0 0 5,001-10,000 17 112,720 16 107,448 1 0 0 10,001-56,000 19 386,775 19 386,775 0 0 0 Unincorporated St. Louis County 1 397,213 1 397,213 Total St. Louis County 90 987,196 63 964,023 City of St. Louis 1 429,300 1 429,300 Total City and County 9 1 1,416,496 64 1,393,323 Excludes Pacific. 2T~oof these municipalities employ part-time officers.

Louis, the unincorporated portion of St. Louis pressed sentiments of county residents reflected in a County, and 62 of 89 incorporated municipalities in poll conducted in 1982 by Market Opinion Re- the county have chosen this arrangement.6 Missouri search. In that poll, more than 80 percent of those Law (Section 70.800, Missouri Revised Statutes), interviewed said that "direct local controlw of police enacted at the behest of local police leaders, re- was "very important" to them.* quires all municipalities in St. Louis County with a Officials of the smallest communities having resident population of 400 or more persons to pro- their own police felt that the loss of federal revenue vide full-time police services. Of the 80 municipali- sharing funds might force citizens to reconsider this ties that fall under the jurisdiction of this statute, 62 stance, but that change would be difficult. If it oc- make provision by organizing their own local police curred, they felt, it would come in the form of joint force, 12 contract for services from adjacent munici- service production (i.e., functional consolidation palities, and 6 contract for services from the county creating one police department to serve two or more police. Among the 53 municipalities with popula- small municipalities) .Q Contracting with a large mu- tions exceeding 2,000 persons, all but one have or- nicipality was, they felt, a less likely alternative, and ganized their own police department. Of the nine contracting with the county police was unlikely. municipalities that are not required to provide police services by state law, six have contractual arrange- Police Contracting ments for service delivery (and two of these six have Most full-service police contracts in St. Louis part-time forces to supplement contract), two have County involve the supply of services by a larger mu- part-time forces only, and one, with a population of nicipality or by the county police to a small munici- 31, makes no explicit arrangement for the provision pality that is an enclave within or immediately adja- of police service^.^ cent to its boundaries. Of the 13 full-service contract In the interviews conducted for this study, suppliers, 7 obtain less than 3 percent of their police elected officials in municipalities with their own po- budget from contract revenues, and only 4 obtain lice forces were asked whether they had considered more than 10 percent. On an average cost per capita disbanding the local force and replacing it with a basis, all but one of the full-service contracts fail to contract with a neighboring municipality or the cover costs. That is, residents of municipalities that county police. Several officials in the smaller mu- receive services under contract incur a lower per nicipalities indicated that they may favor such a capita cost for these services than do residents of the change, but were quick to add that citizens of their jurisdictions supplying services. However, while not communities would be reluctant to give up their local directly calculable from the data available, it appears police. Their assessments are consistent with the ex- that the contract suppliers receive sufficient revenues to cover the marginal cost of supplying services full-service police contract in that year was $58,000, to contracting municipalities.10 while the lowest expenditure for a full-time police In this study, interviews with municipal and department was $108,000. An analysis of per capita county officials whose police departments supply expenditures for police services in 1985 indicates services to other communities under contract indi- that municipalities that arrange for service through cated one of two explanations for this activity, and contracts expend approximately $12 per capita less occasionally both. Officials of larger jurisdictions than those that organize their own police depart- supplying services to very small enclaves or border- ments." ing communities frequently expressed a sense of no- Officials of contracting municipalities were also blesse oblige. That is, the residents of their small asked whether municipalities other than their cur- neighboring communities needed police services, the rent supplier made bids. Approximately one-third of larger community could supply them at little addi- the officials responded that this did occur, with sev- tional cost to itself, and it had an obligation to do so. eral citing recent changes from one supplier to an- If the contract reduced the per capita cost of police other. The viability of such "shopping" for police for its own citizens, all to the good, but this was not services is a function of proximate location vis-a-vis reported as a primary motivation. alternative suppliers. Municipalities receiving serv- Mayors of the smaller communities that supply ices under contract that had several adjoining police services under contract frequently articulated municipalities with police forces reported such com- a different motivation. They argued that supplying petition. This was common among the small munici- services to adjacent municipalities allowed their palities in the Normandy area of North-Central St. community to have a larger police force than it Louis County, and was confirmed in interviews with could otherwise afford, and that their own citizens officials of municipalities that were police service as well as those of the contract municipalities bene- contract suppliers in that area. Officials of munici- fited from this larger force. In the Normandy area, palities in other parts of the county, particularly where competition among several departments for those that are enclaves within larger municipalities police contracts was keen, mayors generally offered or which had only a single bordering municipality, yet an additional explanation. They saw competing reported little or no competition for their contract.12 with nearby jurisdictions for police service contracts as a way of keeping their own departments efficient. PAlTERNS OF POLICE PRODUCTION That is, they had to hold their own costs down and There are 63 full-time municipal police forces in maintain quality service in order to be competitive. the City and County of St. Louis,l3 and one full-time Officials of municipalities without their own local county police department. These full-time police de- police were asked whether establishing a local force partments range in size from a low of 4 officers in 2 was a viable alternative to contracting. These offi- municipal departments to 537 officers in the county cials, for the most part, argued that the start up costs police and 1,702 officers in St. Louis City. Many of required to establish a local police force were too these full-time departments are small. Twenty-one high to give this option serious consideration. Their have between 4 and 10 full-time officers and an ad- assessment is supported by data on police service ex- ditional 20 have between 11 and 24 officers (Table penditures in 1985. The highest expenditure for a 4.2). However, the number of small departments

Table 4.2 Police Service Producers in St. Louis Number of Number of Percent of Total Percent of Total Departments Officers Officers Population

Municipal Departments2 62 1,368 37.9 41 .O With 4-10 Officers 21 146 4.0 3.8 With 11-24 Officers 20 347 9.6 8.4 With 25-77 Officers 21 875 24.3 28.8 St. Louis County Police 1 537 14.9 28.7 St. Louis Metropolitan Police (City) 1 1,702 47.2 30.3 All Police Departments 64 3,607 'Includes population served via contracts. 2~xcludesPacific. Table 4.3 Number of Full-Time Police Officers and Police Expenditures per 1,000 Population1 Police Officers aer 1.000 Po~ulatioq Percent of Real Estate Assessed Value Derived from Commercial-lndustrlal Property Department Size Median 25 Percent 25-50 More than (full-time officers) Department or Less Percent 50 Percent Small (4-10) 2.932 2.52 2.93 3.59 (21l3 (1 1) (4) (6) Medium (1 1-24) 2.71 2.38 2.71 3.37 (20) (6) (8) (s) Large (25-77) 2.34 1.79 2.31 2.43 (21) (5) (9) 0

Police Exaenditures Der 1.000 Poaulation Percent of Real Estate Assessed Value Derived from Commercial-lndustrlal Property Median 25 Percent 25-50 More than Department or Less Percent 50 Percent Small (4-10) $834 76 80 126 (11) (4) (6) Medium (1 1-24) 91 63 99 102 (20) (6) (8) (6) Large (25-77) 85 66 85 88 (21 (5) (9) 0 Resident and contract population. 3Number of departments in this category. 'Number of full-time officers in the median department. 4Expenditures by the median police department in this category. belies the fact that most police officers in St. Louis Among the municipal police departments in St. work in, and most citizens receive police services Louis County, the number of officers per 1,000 resi- from, larger departments. More than 85 percent of dents of own and contract jurisdictions ranges from the full-time police officers in St. Louis City and a low of 1.29 to a high of 7.98. The smallest police county work in departments with 25 or more full- departments in the county tend to have more offi- time officers. These larger departments supply police cers per 1,000 residents of the communities they services to 88 percent of the city and county popula- serve, though there is substantial variation in this ration. The 21 full-time departments with 10 or fewer tio among departments of similar size. Differences in sworn officers employ only 4 percent of all officers assessed valuation of real property, particularly dif- and serve fewer than 4 percent of the city-county ferences in the proportion of assessed valuation of population.14 real property that is derived from commercial and industrial property, help to explain the largest portion of the remaining variation not explained by size. Department Size and Costs Per Capita Departments serving areas with relatively larger There is wide variation in the number of police proportions of commercial and industrial property officers per 1,000 residents served among the full- tend to employ significantly more full-time officers time departments in St. Louis City and County. The per 1,000 residents than those serving communities two largest departments are quite distinct in this re- with little or no commercial or industrial activity gard. The St. Louis metropolitan police department, (Table 4.3) .15 A simple explanation for this is that serving the City of St. Louis, has 3.76 officers per communities with significant amounts of commercial 1,000 city residents. The county police department, and industrial activity have workday populations that on the other hand, has 1.32 officers per 1,000 resi- are larger than their resident populations.16 Police dents of the unincorporated county and the munici- services are supplied to persons working and shop- palities it serves under contract. ping in a jurisdiction, as well as to its residents. Police expenditures per capita also vary signifi- business activity, as discussed in Chapter Eight. cantly among the police departments. The St. Louis An analysis of police expenditures indicates that metropolitan police department expends $175 per some slight economies of scale or size are enjoyed city resident; the county police department expends by the larger police departments. In Table 4.4, the $76 per resident of the unincorporated county and presence of size economies is indicated for total po- its contract municipalities. The range of per capita lice expenditures when the equation has a positive expenditures among the municipal departments in and significant intercept, and for per capita police expenditures when the coefficient for population St. Louis County is $53 to $267. Here, too, differ- served is negative and significant." In the per capita ences in the extent of commercial and industrial ac- equations, the coefficients indicate that, after adjust- tivity help to explain this variation. As shown in the ment for other factors affecting police expenditures, lower portion of Table 4.3, higher expenditures per a 1,000 resident increase in the number of citizens jurisdiction resident are found for police depart- served is predicted to yield a 50 to 70 cent decrease ments that serve communities with higher propor- in police expenditures per capita, a decrease of tions of commercial and industrial activity. This ef- somewhat less than one percent for an average de- fect is stronger for the smaller departments in the partment .la county, but is significant even among the larger de- The effect of size is relatively weak, however. As partments. Higher police expenditures per capita are in the analyses of total revenues and total expendi- a significant contributor to the extra per capita tures, the presence of business activity has the costs of municipal government associated with strongest effect on police expenditures. The equa-

Table 4.4 Regression Coefficients for Police Expenditures1 in County Municlpallties- 1985 (N = 61) Police Expenditures Police Expenditures per Capita Unweighted Weighted* Unweighted Weighted Resident Population

Median Household Income

Percent Poor: 1979

Percent Nonwhite

Percent Over 65 Years Old

Density (number per square mile)

Percent Owner Occupied

Percent Housing Built 1970-80

Value of Commercial Property4

Village

Point-of-Sale

Intercept (280,995) (397,086) (30.83) (89.13) R2 0.94 0.94 0.75 0.75 * -Significant at p < .001. + -Significant at p < .05. Includes expenditures and populations for police contracts. 2Weighted by total population, including contracts 3Sta~darderror of regression coefficients. 4Assessed value in $1.000'~(total or per capita) . tions suggest an increase in police expenditures of $7 tance from the county police department and, less to $8 per capita for each additional $1,000 of as- frequently, from the major case squad, discussed be- sessed value in commercial and industrial property. low. As discussed in Chapter Three, St. Louis Service Components County government developed the functional equi- The range of services produced by area police valent of an urban services district in the unincor- departments in-house varies in rough proportion to porated county during the years where the borders their size. All departments produce local patrol and of the area were essentially frozen by the Crueler call response, traffic control, and initial investiga- decisions on annexation. The county police became tions of most reported crimes. On the other hand, the equivalent of a municipal police department for all departments in St. Louis County use the county this area, the largest such department in the county police crime lab (the city has its own lab), and all by a substantial margin. departments in the county and the city use the As a result both of this large size and of the Greater St. Louis Police Academy for training re- county government's responsibility to provide serv- cruits and other training activities.19 In between ices countywide, the St. Louis County police depart- these alternatives is a variety of other possibilities. ment became a supplier of many specialized se~ces Among the smallest departments-those with ten to the municipal departments. In addition to provid- or fewer officers-only one produces its own radio ing facilities for crime lab and training, a central reg- communications, doing so in conjunction with the istry for trained officers, the full county share of the local fire department and supplying an adjoining 9 11 dispatch system and (initially) the REJIS infor- small department with communications as well. The mation network, and assisting in investigations of remaining 20 small departments obtain radio com- unresolved homicides (indeed taking over such in- munications under contract from a joint dispatch vestigations for departments that lack skilled investi- center (supplying 4 departments), from larger mu- gators), the county police assisted in sexual assault, nicipal departments (supplying 12). and from the drug, and arson-explosives cases, provided follow up county police (supplying 4 departments). Five of the investigation for other serious crimes in small juris- 20 police departments in the 11-24 officer range dictions, and supplied crime scene investigation, produce their own radio communications. Two do so photography, identification, and prisoner assessment in conjunction with local fire departments, and one and transportation. This assistance was provided to supplies dispatch for an additional nine officers in a municipal departments on request, without charge. nearby department. The remaining 15 departments Municipal departments could also request the assis- obtain radio communications from a joint communi- tance of the county's police helicopter if needed. cations center (supplying five departments in this While many of the larger municipal departments size range in addition to four of the smallest), from produced most of these services for themselves, co- larger municipal departments (nine receive dispatch operating with the county as approximate equals from larger departments), and from the county po- when such cooperation was warranted,22 the smaller lice (one department). All of the departments with and medium sized departments relied heavily on the 25 or more sworn officers produce their own com- county police for this assistance. munications, though several do so in conjunction Since the Town and Country decision in 1983 with local fire departments. allowing municipalities greater freedom to annex un- Most police departments in the county, but par- incorporated territory (see Chapter Three), county ticularly the smallest ones, frequently rely on mutual government has moved to focus its resources on the aid from adjacent departments to service peak unincorporated county, including its police depart- demands. Such mutual aid is usually extended ment resources. The county police established new informally among departments that share radio fre- service centers in two parts of the unincorporated quencies, is quite common among the smaller de- county. Maryland Heights and Chesterfield, where partments in the county, and is not uncommon significant sentiment for new incorporation was among the larger. Unlike patterns found in many manifest, and reallocated officers to increase its metropolitan areas, where county or state police presence there.23 Because of this reallocation and, take responsibility for major traffic flows countywide especially, the loss of revenues that occurred as a (e.g., those found on Interstate highways),2O in St. result of successful annexations and the Maryland Louis County many municipal departments provide Heights incorporation.24 the county police depart- traffic patrol on major arterial highways.2' Investiga- ment has reduced services to municipal depart- tion of the most serious crimes that occur, especially ments, virtually eliminating some (e.g., arson inves- unresolved homicides in the jurisdictions of smaller tigation) and switching to service contracts for municipal departments, frequently involves assis- others, such as crime-scene investigation, photography, and identification. Some of the departments politan areas. However, St. Louis also exhibits sub- that received such-services free of charge from the stantially more concentration in the production of county police are now contracting with the county auxiliary services than do most similar areas. These police. Others are developing in-house capabilities, differences can be traced to historical patterns of lo- and still others are forming local consortia of depart- cal growth and incorporation in St. Louis County, ments to replace county services. citizen preferences for local control of policing, and, in the case of auxiliary services, entrepreneurial ef- Production Efficiency: forts by police leaders. A Comparative View The basic pattern of police production in St. A study of police organization in 80 U.S. metro- Louis County is characterized by a relatively large politan areas in 1975 found that in 13 of these met- number of small producers of police patrol and ropolitan areas with populations between 500,000 closely related services, combined with relatively and 1.5 million persons, the median area had 29 small numbers of larger producers of specialized, producers of police patrol services, and that 75 per- auxiliary services. This pattern was compared to oth- cent of these metropolitan areas had fewer than 39 ers in a study of the efficiency with which police de- patrol producers.25 The comparable figure for the partments in metropolitan areas convert resources City and County of St. Louis is 73 patrol producers (e.g., sworn officers, civilians, vehicles) into out- including, as the 1975 study did, patrol producers puts-crimes cleared and units deployed for patrol with specialized jurisdictions.= Only one of the 13 and immediate response. Metropolitan areas that, metropolitan areas similar in size to St. Louis (Pater- like St. Louis, exhibited substantial diversification of son-Clifton-Passaic, New Jersey) had more produc- the patrol function combined with substantial con- ers of police patrol. In per capita terms, however, centration of auxiliary services were found to be St. Louis is not as atypical. Dividing its 73 patrol more efficient compared to other patterns. The less producers by its population yields a ratio of 5.2 pa- efficient patterns were those with fewer patrol pro- trol producers per 100,000 residents. The compara- ducers or more producers of auxiliary services.28 A ble figure for the median metropolitan area in the possible explanation for this result is that smaller de- 1975 study was 4.0 per 100,000. In terms of relative partments are much more efficient at deploying offi- concentration of service supply, St. Louis, with 30 cers for patrol duties, while larger departments or percent of its population served by its largest pro- overlying joint investigation units more efficiently ducer, falls slightly below the concentration found in clear crimes that are reported to them.29 These find- the median metropolitan area in 1975, where 36 ings do not provide direct evidence that policing in percent of the population was served by the largest the St. Louis area is or is not efficient, but rather producer. that the functional arrangements for policing in St. In its arrangements for the production of auxil- Louis, particularly those arrangements found in the iary services, however, St. Louis exhibits substan- incorporated part of St. Louis County, have proved tially greater concentration than found in similar size to be comparatively efficient elsewhere. metropolitan areas in 1975. In radio communications, for example, only 45 percent of St. Louis's COOPERATIVE local service producers have their own dispatch cen- PRODUCTION ARRANGEMENTS ter. In the similar sized median area in 1975, 82 St. Louis affords many examples of cooperation percent of the local service producers supplied their and coordination in policing. In relation to the pa- own radio communications. Entry-level training is trol function, local police departments commonly also much more concentrated. In 1975, the median supply backup capabilities to one another as metropolitan area of St. Louis's size had eight pro- needed. Backup is facilitated by coordination in ducers of such training. In St. Louis, virtually all of- communications, such as shared radio frequencies ficers receive entry-level training from the Greater and joint dispatch arrangements, allowing officers in St. Louis Police Academy.27 St. Louis's concentra- any one jurisdiction to know what is occurring in ad- tion in auxiliary service production reflects the ex- joining jurisdictions on a real-time basis. All of the tensive use of cooperative and contractual agree- police forces in both city and county are linked to ments. the computer aided 91 1 emergency call routing sys- Production arrangements for policing in the St. tem. Investigation services are also coordinated in Louis area, therefore, differ from those in compara- various ways, including an areawide Major Case bly sized metropolitan areas in two different direc- Squad that draws trained investigators from many tions. The St. Louis area has more local police serv- separate jurisdictions to assist in the investigation of ice producers in absolute numbers and in proportion serious crimes. The Regional Justice Information to its population than do most comparable metro- System (REJIS) affords linkages to statewide and nationwide criminal information data banks. With ing dispatch operations, releases officers in many respect to training, all area police departments re- departments for on-street deployment. ceive recruit and in-service training from the Greater Regional Justice Information System St. Louis Police Academy. Many of these cooperative ventures-including This system, known by its acronym REJIS, pro- areawide mutual aid plans, specialized investigation vides a common data base for police relevant infor- resources, a shared information system, and a single mation to all of the police agencies in the St. Louis areawide police academy-are the fruits of initiatives area. REJIS terminals are located in each of the undertaken by local police leaders in the St. Louis PSAPs in the city and county, allowing dispatchers area. Local police officials from municipal depart- ready access for a variety of records checks re- ments, the county police department, and the St. quested by police officers. The REJIS system is Louis city department typically propose a plan, con- linked to statewide and nationwide law enforcement vince their colleagues to support it, and shepherd information systems as well. In addition to informa- the necessary enabling legislation through the appro- tion found in most law enforcement computer sys- priate decisionmaking bodies including, in the case tems (e.g., outstanding arrest warrants, arrest and of training standards, the Missouri legislature. conviction records, stolen vehicle reports), REJIS Prominent in this process has been the Board of maintains additional files on topics such as commu- Governors of the Law Enforcement Officials of the nity social services and business firms throughout the Greater St. Louis Area, a "peak" association of po- area. The city, county, and municipal departments lice chiefs from the largest departments in the area share the cost of maintaining and operating REJIS. County government pays 25 percent of the munici- and from a selection of smaller departments. This group meets monthly to discuss issues of concern to palities' cost share for REJIS operation^.^^ local police leaders. Cooperative ventures of less Mutual Aid among Police Agencies than an areawide focus, too, have come about Only the very largest police departments in the through the efforts of local police leaders who recog- United States deploy sufficient officers for patrol nize the wisdom of providing one another with and immediate response duties that allow them to emergency backup, of coordinating investigations, handle peak demands unaided. Virtually all police and of sharing dispatch and other specialized serv- departments that share borders with other police ju- ices.30 risdictions supply mutual backup aid. Such emergency backup is widespread in St. Louis, as in most Computerized Call Routing other metropolitan areas. Among many of the police Of major assistance to citizens needing police departments in the St. Louis area, backup is sup- and other emergency services in the City and County plied informally, as officers sharing radio frequen- of St. Louis is the availability of a 91 1 emergency cies aid officers in nearby jurisdictions when a per- telephone number combined with computerized ceived need arises. A number of the St. Louis routing of emergency service requests. A citizen departments have formalized backup assistance in needing emergency assistance anywhere in the City mutual aid compacts (a figure repeated by several and County of St. Louis can be connected automati- local officials was that 30 departments were mem- cally with the appropriate supplier for his or her lo- bers of a mutual aid pact), but aid appears to be cation by dialing 91 1. The 91 1 system routes re- provided as needed with or without formalization. quests for police, fire, and ambulance services, Police departments in St. Louis County have de- drawing on an elaborate geocoding system devel- veloped a formal plan to deal with the most serious oped jointly by the city, county, and municipalities. natural and man-made disasters. This plan, known A citizen's call is directed automatically to one of 29 as Code 1000, was developed originally to enable Public Safety Answering Points (PSAPs) located in areawide police mobilization in the event of civil dis- the 29 police agencies in the city and county that turbances. It now serves as a means of mustering operate dispatch centers. On receipt at a PSAP, a large numbers of officers to assist in the event of a call for police service can be dispatched immediately disaster. Under the plan, St. Louis County is divided to an officer in the relevant jurisdiction. Fire and into five zones, each of which has access to between ambulance calls are dispatched from the PSAP for 300 and 400 officers from the county police and some jurisdictions, and switched to dispatch centers municipal departments in the zone and, if necessary, specialized to these services in other jurisdictions. officers from other zones as well. The plan is initi- The development and operation of this system have ated at the discretion of a local police commander been funded by the county (two-thirds) and the city needing aid. On his or her request, adjoining police (one-third). In addition to its direct benefits to citi- departments switch from eight- to twelve-hour duty zens with an emergency, the system, by concentrat- shifts to increase their available personnel and, if re- quested, dispatch all but the minimum force neces- or "excellent" police services.33 These reported lev- sary to serve their own jurisdictions to the aid of the els of satisfaction with police services are character- requesting department. istic of surveys conducted in St. Louis for many years. The St. Louis Sample Survey of Citizen Opin- The Major Case Squad ion and Participation, conducted in 1959, reported The Major Case Squad of the Greater St. Louis 76 percent of city respondents and 89 percent of Metropolitan Area, formed in 1964-65, was mod- county respondents to be "fairly satisfied" or "very eled on a similar organization-the Metro Squad-in satisfied."34 Surveys of citizens in selected neighbor- the Kansas City area. It was organized under the hoods in the city and the county were conducted by auspices of the Board of Governors of the Law En- researchers from Indiana University in 1972 and forcement Officials of the Greater St. Louis Area, again in 1977. In 1972, 58 percent of respondents and currently operates with its own board of direc- in the city rated their police service as "good" or tors. The squad consists of investigative officers from "outstanding"; in the county, 76 percent rated local St. Louis County municipalities, the county police, police as "good" or outstanding."^ In 1977, 51 the St. Louis metropolitan police, and from police percent in the city and 70 percent in the county gave departments in surrounding counties in Missouri and these same answers.38 Illinois. Major case squad officers train together pe- Public opinion polls and general ratings from riodically, developing working procedures to be em- citizen surveys should not be taken as complete as- ployed when needed to investigate a serious crime. sessments of police performance. One reason is that Organizational details are prearranged so that the general ratings are usually more favorable than are squad can be activated immediately on request by assessments of more specific police actions-re- any police chief in whose jurisdiction such a crime sponding rapidly when called, providing assistance to occurs. Activation is subject to the approval of the victims and others, preventing crime and disorder in squad's board of directors, and is usually for a lim- a neighborhood.37 A second reason is that citywide ited time-five days-although extensions are possi- or countywide averages can mask significant internal ble if circumstances warrant. The squad's primary variations. The only studies available that go focus has been on homicides, although it investigates beyond general assessments and allow intracity and other serious crimes too. intracounty breakdowns are the 1972 and 1977 The Major Case Squad is a response to a com- Indiana University studies, which focused on specific mon exigency in policing-the fact that serious neighborhoods within the city and within county ju- crimes requiring large numbers of trained investiga- risdictions. tors are rare events for a single department. Yet, in Table 4.5 presents data from the Indiana studies a large metropolitan area, these crimes do occur and that illustrate intracity and intracounty variations in call for the immediate deployment of many investi- overall ratings of local police, perceptions of the gators. By drawing on officers from a large number speed of police response when called, perceptions of of cooperating jurisdictions, police in St. Louis are the trend of crime in the respondent's neighbor- able to meet this need. hood, and whether the respondent has been the vic- tim of crime in the neighborhood within the previous POLICE PERFORMANCE year. On each of these indicators, and in both years, No attempt was made in this study to assess the residents of neighborhoods in St. Louis County re- service delivery performance of local governments in ported, on average, more favorable perceptions of St. Louis. Performance measurement for local serv- local police and fewer unfavorable experiences (be- ices is a complex and costly undertaking, and fell ing victimized) than did city neighborhood residents. outside the scope of research. However, some indi- However, substantial overlap in assessments and ex- cators of police performance are available from pub- periences was found in examining variations across lic opinion polls, academic research on police serv- city and county neighborhoods. That is, there were ices in St. Louis, and local police reports. neighborhoods in the county where these assess- A 1982 survey conducted by Market Opinion ments and experiences were more negative than the Research found 79 percent of 1,000 respondents in average in the city, and vice versa. the city and county to be either "very" or "some- Within jurisdictions in these studies, assessments what satisfied" with their police protection.32 Citi- and experiences were patterned partially along racial zens in the county were more likely to report satis- and income lines. Neighborhoods with higher pro- faction than those in the city but the differences portions of low income and nonwhite respondents were slight. In 1987, Attitude Research Corporation reported assessments of local police that were, for surveyed 2,500 county residents and found that 91 the most part, below the average in their jurisdic- percent of the respondents reported either "good" tion. Similar patterns were found across jurisdic- Table 4.5 Assessments of and Experiences with Local Pollcel

Nelghborhoods Located in: Year of Survey City of St. Louis St. Louis County Lowest Average Highest Lowest Average Highest Score Score Score Score Score Score 1972 [N = 71 [N = 371 Percent Rating Police "Good" or "Outstanding" 20 58 73 19 76 94 Percent Rating Response "Very Rapid" or "Quickly Enough" 54 67 85 30 84 99 Percent Perceiving a Crime Increase 29 40 48 7 24 47 Percent Victimized in Previous Year 22 29 34 7 18 34 Neighborhoods Located in: City of St. Louis St. Louis County Lowest Average Highest Lowest Average Highest Score Score Score Score Score Score 1977 [N = 81 [N=17] Percent Rating Police 44 51 69 40 70 9 1 "Good" or "Outstanding" Percent Rating Response 53 6 1 69 50 82 94 "Very Rapid" or "Quickly Enough" Percent Perceiving a Crime increase 24 32 4 1 6 20 29 Percent Victimized in Previous Year 26 37 49 14 26 43 'Data from studies conducted by Indiana University- see endnotes for attribution. tions. Such variations and patterning have been ments in the county were more likely to rate their found in most research on local police and citizens' police this way than were citizens served by the assessments of and experiences with them. One rea- smallest departments there. In St. Louis County, son that the City of St. Louis exhibits lower satisfac- neighborhoods with high proportions of low income tion ratings in the opinion polls and scores lower on and nonwhite citizens are more prevalent in small these performance indicators is that the city has sig- communities than in large ones. Like the city-county nificantly more low income and nonwhite citizens comparisons noted above, these size comparisons than does St. Louis County.38 are, therefore, partially confounded by racial and Looking at neighborhoods within county munici- income differences across neighborhoods, and some palities only, the two studies provide partial informa- of the size effects noted in this paragraph are really tion on the question of department size and per- those of race and income. In addition, the results formance. Victimization rates were lower in both reported here are general tendencies, and some years in neighborhoods served by smaller depart- small departments out-performed some of the larger ments.39 Perceptions of crime trends reversed be- municipal departments.42 tween 1972 and 1977, with residents of larger juris- Neighborhoods served by the St. Louis County dictions reporting larger increases in 1972, but police were surveyed in the 1972 study, but not in residents of smaller ones reporting such increases by 1977. In 1972 citizens in neighborhoods served by 1977. In both years, citizens from neighborhoods the county-two in the unincorporated county and served by the larger municipal departments in the two small municipalities served under contract-re- county were more likely to say that their police reported assessments and experiences right at the sponded "very rapidly" or "quickly enough" when countywide average, somewhat less favorable than called than were residents of the smaller jurisdic- those in the larger municipal jurisdictions and some- tions.40 Speed of response when called is strongly what more so than in the smaller. correlated with citizens' perceptions of overall police In terms of crimes recorded by local police, St. effectiveness,41 and the pattern of citizens rating Louis County, with 40 Part I (generally serious) their local police as "good" or "outstanding" is the crimes per 1,000 residents, has somewhat less seri- same as that for speed of response. Citizens living in ous crime than found in the average metropolitan neighborhoods served by larger municipal depart- county.43 In the City of St. Louis, the recorded crime rate in 1984 was 104 per 1,000 residents, highly than any economies that might be somewhat above the average for large central cities captured by merging the smallest depart- in metropolitan area~.~4The recorded crime rate in ments. the unincorporated county was 34 per 1,000 resi- Competition for police service contracts is dents in 1985, while in the municipalities of St. found in some parts of the county, though Louis County, it averaged 44 per 1,000 residents. not countywide. Competition might become Among the county municipalities, recorded crime more widespread if the county police de- rates for jurisdictions with police departments of 4 to partment were to become more aggressive in 10 sworn officers averaged 34 crimes per 1,000 resi- seeking service contracts (adopting a strat- dents, the same as the unincorporated county. Juris- egy akin to that of the Los Angeles County dictions served by larger departments had somewhat Sheriff, for example). higher recorded rates on average, with 44 crimes per The Major Case Squad, the Code 1000 1,000 in those served by 11 to 24 officers, and 46 plan, and the areawide 911 and REJIS sys- crimes per 1,000 in jurisdictions served by 25 or tems indicate that local police leaders have more full-time officers. had significant success at metropolitan problem solving, adopting regional solutions to CONCLUSION police problems that spill over municipal Organizationally, policing in St. Louis is com- boundaries and, at the same time, remain- plex. A large number of autonomous police depart- ing fully accountable to citizens in their local ments produce a variety of services for the nearly jurisdictions. 1.5 million residents of the city and county. Jurisdic- tional fragmentation has been maintained at the 'See, for example, the review of surveys by Jiri Nehnevajsa, same time that a series of organizational overlays Crime in the Neighborhood (Pittsburgh: University of Pitts- was created to address specific functional needs. burgh, Center for Urban Research, 1977). The pattern that results is not unorganized chaos, 2~bid. but rather service delivery that is frequently coordi- 3This arrangement dates back to the time of the Civil War. nated among a group of independent agencies. This 4The arrangement raises an interesting question in light of coordinated effort has resulted in service delivery the Hancock Amendment to the Missouri Constitution (see that is, for the most part, highly rated by those most Chapter Three). Among other provisions, the amendment requires the state to fund any costs of mandates imposed on relevant to such an assessment-the citizens who live local governments. If one interprets the St. Louis board of there. The pattern is also one that has been found to police commissioners, given its appointment by the gover- be efficient in a comparative study of metropolitan nor, as an agency of the state, one could argue that the state, and not the city, is required to fund the metropolitan areas nationwide. police. This argument, while not accepted in Missouri juris- The evaluative criteria posed in Chapter One prudence, does appear to have given the city some addi- can be applied to police services in St. Louis as fol- tional leverage in bargaining with the board. lows: 5For the purposes of this chapter, the St. Louis metropolitan police department is treated as a bureau of the City of St. Police services in St. Louis appear to be rea- Louis. This is technically incorrect, as noted earlier. How- sonably well coordinated among separate ju- ever, the distinction is more technical than real, and the risdictions along a broad range of functional metropolitan police are in all aspects relevant to this chapter a police department organized by the city. dimensions. Coordination between county BThe City of Pacific, with only 12 residents in St. Louis and municipal police departments, however, County, is located principally in an adjacent county. It is has become somewhat more tenuous in re- not included in the discussion in this chapter. cent years. 71n this municipality, Champ, the county police enforce state law (but not local ordinances) and respond to requests for Through cooperative functional arrange- emergency assistance. ments, local police have captured economies *~arketOpinion Research, Public Opinion on Government of scale, especially in auxiliary service pro- Reorganization Alternatives for the City of St. Louis and duction utilizing joint dispatch centers and St. Louis County (Detroit: April 1982), p. 54. centralized crimelab and training facilities. g~epartmenlsin the Normandy area experimented with such a system in the form of the North Area Municipal Police Although some remaining economies of Association (NAMPA), which linked six small depart- scale might be captured by moving to a ments. While NAMPA is not presently active, it affords a larger scale of operation in the case of very model for such joint production arrangements. See St. small departments, the gains would be small. Louis County Department of Planning, Normandy Area Consolidation of Services Study: Police Services Element Local control of police is highly valued by (Clayton: August 1976). citizens in the incorporated portion of St. 1°~ompetition among departments would explain this ten- Louis County and, perhaps, valued more dency toward marginal rather than average cost pricing. "This $12 figure is the coefficient for a dummy variable that Missouri Highway Patrol until its own police department be- represents contracting in a regression equation predicting gan operations. police expenditures while controlling for local service conditions. The full equation is not shown. 24~hecounty government lost approximately $2 million in revenues earmarked for police services when Maryland 12~ompetitionfor police contracts in the Normandy area does Heights was incorporated. As a result, the county police de- appear to decrease per capita costs and increase service lev- partment reduced its personnel by 33 officers. Had the els somewhat. The average per capita cost of a full-service Chesterfield incorporation initially proposed in 1986 been contract from a municipal supplier in the Normandy area is successful, the county projected a further reduction of 60 15 percent lower than the cost of an equivalent contract in officers. James Hennessey, May 22, 1986. The modified other parts of the county, and the number of officers sup- Chesterfield incorporation approved by voters in April 1988 plied per 1,000 residents of a contract recipient is 7 percent calls for initial delivery of police services by the county po- higher. lice under contract to the new city. 13~heCity of Pacific, located principally in another county, is 250slrom, Parks, and Whitaker, Patterns of Metropolitan not included in the following discussion. Policing, Chapter Five. 14This skewing of the size distribution among local police de- 2e~henine additional police producers with specialized juris- partments is similar to that found in many U.S. metropoli- dictions are the Missouri Highway Patrol, the St. Louis tan areas. While there are many more small than large po- County Park Rangers, the Lambert-St. Louis Airport Po- lice departments, most metropolitan Americans receive lice, and campus police departments of Washington Univer- police services from departments of 20 or more sworn offi- sity, St. Louis University, the University of Missouri-St. cers. See Elinor Ostrom. Roger B. Parks, and Gordon P. Louis, and the junior colleges located in the area. Whitaker, Patterns of Metropolitan Policing (Cambridge. Massachusetts: Ballinger Publishing Company, 1978), pp. 27The Missouri Highway Patrol has its own training academy. 85-88. All other departments used the Greater St. Louis Academy at the time of our interviews. However, St. Louis County I5~hesecomments are based on a regression analysis that was in the process of establishing its own police-fire training uses total population served (own jurisdiction plus contract academy, a facility that would serve all departments in the municipalities), assessed valuation of commercial and resi- county. The impetus for this new facility appears to be a dential property, the number of Part I crimes reported in low demand for training by the city combined with the coun- 1985, and indicators of dispatch and contract service supply ty's desire to have a fire service training academy in addi- to explain variation in the number of full-time officers em- tion to one for police training. ployed by each department. Taken together, these variables explain 95 percent of the variation in number of full-time 28~ogerB. Parks, "Metropolitan Stmcture and Systemic Per- officers. formance: The Case of Police Service Delivery," in Ken- 16~enton'sworkday population, for example, is more than neth Hanf and Theo A. J. Toonen, eds., Policy Implemen- (Dordrecht, The ten times its residential population. tation in Federal and Unitary States Netherlands: Martinus Nijhoff Publishers, 1985), pp. I7see Chapter Eight for a discussion of the methodology used 161-191. here to test for economies of scale. 290n the deployment issue, see Ostrom, Parks, and I8~hesomewhat higher per capita costs in the smallest juris- Whitaker, Patterns of Metropolitan Policing, p. 89. In dictions may also result from additional duties assumed by larger departments nationwide, nine or ten officers are police there. In some small jurisdictions, police are the only needed to put one officer on the street 24 hours per full-time employees, and the police budget includes expen- day. In smaller departments, only four or five officers are ditures that are reported in other departmental categories in needed. As a result, patrol presence is much higher in the larger jurisdictions. See St. Louis County Department of jurisdictions of smaller departments. On the question of Planning, Normandy Area Consolidation of Services, p. 7. clearances, see Wesley G. Skogan, "Efficiency and Effec- tiveness in Big-City Police Departments," Public Admini- Ig~hiswill change once the new St. Louis County Police and stration Review 36 (May-June 1976): 278-286. Fire Training Academy is built. See Note 27 below. 30~nexcellent discussion of the emergence of cooperative po- 20~eeOstrom, Parks, and Whitaker, Patterns of Metropoli- lice arrangements through the efforts of local police leaders tan Policing, Chapter Six. in the St. Louis area can be found in James C. McDavid, 21Several of the small police departments significantly aug- Police Cooperation and Performance: The Greater St. ment municipal revenues with fines and fees collected in lo- Louis Interlocal Experience (State College: The Pennsylva- cal traffic courts and, even for some of the larger depart- nia State University Press. 1979). ments, the revenues raised are not insignificant (see ltThis is a reduction from county government's original 100 Chapter Two). percent funding of this service in the county. Whether this 22~eeSt. Louis County Municipal League, "Annexation will continue is unclear at present. The county executive Study Commission Survey of Selected County Police Serv- has announced plans to eliminate what he perceives to be a ices to Large Municipalities, " Clayton, Missouri, April 26, subsidy. Municipal, and some county, officials have argued 1984, p. 2. Chiefs of the larger municipal departments felt that the county obtains benefits by having access to infor- that "certain county services should be categorized as mu- mation supplied by the municipalities and that, rather than nicipal services that any municipal police department would representing a subsidy to them, the county's 25 percent make available to any other municipal police department." contribution to municipal costs is an appropriate recognition of these benefits. The county executive has also proposed =1nterview with James Hennessey, Director of Administra- elimination of the county's funding of the 91 1 system, shift- tion, St. Louis County Police, May 22, 1986. When, in ing to a telephone tax to support this service. spite of this reallocation, the Maryland Heights incorporation was successful, the county removed all police protec- 32Markel Opinion Research, Public Opinion in Government tion immediately, and the new city was patrolled by the Reorganization. 33St. Louis Post Dispatch, "Chesterfield and U. City Lead in 40~hisis principally a result of those very small departments Supporting McNary Remap, " November 27, 1987, Section that deploy only a single patrol unit. Even with backup sup- A, pp. 1 and 6. plied by neighboring departments, their response can be de- layed when the single unit is busy. See Roger B. Parks, 34John C. Bollens, ed., Exploring the Metropolitan' Commu- "Linking Objective and Subjective Measures of Police Per- nity (Berkeley: University of California Press, 1961), p. formance, " Public Administration Review 44 (March-April 432. 1984): 118-127. 35~ora description of the 1972 study, see Roger B. Parks, 47What police do after responding rapidly (or slowly) is also Assessing the Influence of Organization on Performance: A important. See Roger B. Parks, "Police Response to Vic- Study of Police Services in Residential Neighborhoods, un- timization: Effects on Citizen Attitudes and Perceptions," in published Ph.D. dissertation (Bloomington: Department of Wesley G. Skogan, ed., Sample Surveys of the Victims of Political Science, Indiana University, July 1979). See also Crime (Cambridge, Massachusetts: Ballinger Publishing Elinor Ostrom, "Size and Performance in a Federal Sys- Company, l976), pp. 89-104. tem," Publius 6 (Spring 1976): 33-73. This study and the one conducted in 1977 focused on neighborhoods that, on 42For data from the 1972 study, see Elinor Ostrom, "Size and average, had more crime and other police-related problems Performance in a Federal System." Data from the 1977 than did the city or county as a whole, thus explaining the study were reanalyzed for this report. Neighborhood and ju- somewhat lower ratings of police found in these studies. risdiction-specific data are reported in Law Enforcement in St. Louis Communities (5 volumes) for the 1972 study and 38~he1977 study is described in Stephen L. Percy, Citizen in Chief's Report for (name of jurisdiction) (12 volumes) Coproduction of Community Safety and Security, unpub- for the 1977 study. These reports were sent to the police lished Ph.D. dissertation (Bloomington: Department of chiefs of all departments studied by the Workshop in Politi- Political Science, Indiana University, November 1981); cal Theory and Policy Analysis, Indiana University, Debra L. Dean, Comparative Models of the Social Bloomington. Services Production Process, unpublished Ph.D. disserta- 43Crimes and rates for St. Louis County and its municipalities tion (Bloomington: Department of Political Science, Indi- are from 1985 Fact Sheet: Police Services in St. Louis ana University, October 1982). County (Clayton: Bureau of Management Services, St. 370n this point, see the review by Mervin F. White and Ben Louis County Police Department, 1986). Data for the City A. Menke, "A Critical Analysis of Surveys on Public Opin- of St. Louis are from Annual Report 1984-1985, St. Louis: ions Toward Police Agencies," Journal of Police Science St. Louis Metropolitan Police Department, 1985). and Administration 6, No. 2 (1978): 204-218. 44Recorded crime rates, and the underlying criminality they 38~hisis not the entire explanation, however. After careful imperfeclly reflect, result from a variety of factors only adjustment for neighborhood differences in respondent some of which are subject to police control. In addition to characteristics, Elinor Ostrom reported higher performance variations in criminogenic factors, crime rates reflect differ- levels in neighborhoods served by county municipalities than ences in the extent to which citizens report crimes to police in neighborhoods in the City. See Elinor Ostrom, "Size and and differences in the ways that police record those crimes Performance in a Federal System." that are reported to them. Thus, crime rates are quite imperfect indicators of police performance. Their inclusion 39~olice-recordedcrime rates show the same pattern. See be- here is only for purposes of rough comparison with other ar- low. eas.

Chapter Five Fire Services

INTRODUCTION districts in the county.2 All but one fire district produce fire services with a district organized fire de- This chapter describes the ways that citizens of partment. The remaining district, Meacham Park, St. Louis City and St. Louis County provide for fire contracts with an adjacent municipality. protection and organize the production of fire services, specifically the prevention and suppression of FIRE SERVICE fires.' All municipalities in the St. Louis area are PROVISION AND PRODUCTION authorized by state law to provide fire protection The provision and production of fire services in services. Municipalities may choose between estab- St. Louis are somewhat more concentrated than are lishing a municipal fire department and contracting police services. The number of provision units for with another government unit for these services. fire service is 52, fewer than for police, yet still a Citizens in St. Louis County have an additional substantial number (Table 5.1). There are 43 de- choice available, namely to create a separate fire partments producing fire services and 9 provision protection district. These districts, which are inde- units contracting for services.3 pendent of municipal governments, are governed by The City of St. Louis is the dominant provider elected, three-member boards of directors, and have of fire services in the area, serving just under one- authority to levy taxes on local property to provide third of the combined city and county resident fire services. Fire protection districts, too, may population.4 The city has 670 fire fighters, 36 per- choose between in-house production-the organiza- cent of the total employed in the city-county, mak- tion of a district fire department-and contracting ing it the dominant producer also. St. Louis City ex- with another government unit for the production of penditures for fire protection services in 1985, some fire services. Citizens in unincorporated St. Louis $30 million, constituted 40 percent of the total fire County must organize fire protection districts to pro- expenditures in that year.5 vide fire services for themselves, but the districts re- The municipal and fire district departments in tain the option of contracting. the county are substantially smaller and, individu- The City of St. Louis and 19 municipalities in ally, serve fewer citizens. The boundaries of their St. Louis County have chosen in-house production, jurisdictions are shown in Figure 5.1, together with establishing their own fire departments. Eight mu- the location of fire stations. The largest fire protec- nicipalities in the county arrange for fire service pro- tion districts are Mehlville, with a resident popula- duction through a contract with a fire department in tion of nearly 100,000, and Florissant Valley, with another unit of government. Five of these contracts more than 80,000. The largest municipality with its are with adjoining municipalities, and three are with own fire department is University City, with a resi- fire protection districts. In the remaining 62 munici- dent population of approximately 43,000. Fire pro- palities in St. Louis County, and in all of the un- tection districts serve approximately 50 percent of incorporated county, citizens provide for fire protec- the city-county resident population and employ 38 tion through one or more of the 24 fire protection percent of the full-time fire fighters. The fire district Figure 5.1 Fire Districts and Fire Departments in St. Louis County

Municipal Fire Departments '7 1 Berkeley 5 Des Peres 9 Hazelwood 13 Maplewood 17 St. Louis 2 Brentwood 6 Ferguson 10 Jennings 14 Olivette 18 Shrewsbury 3 Clayton 7 Frontenac 11 Kirkwood 15 Richmond Hgts. 19 University City 4 Crestwood 8 Glendale 12 Ladue 16 Rock Hill 20 Webster Groves SOURCE: St. Louis County, Missouri, Fact Book- 1986. St. Louis County Department of Planning. Table 5.1 Fire Services in St. Louis City of St. Louis County Fire St. Louis Municipalities Protection Districts Number of Service Providers 1 27 24 Number of Service Producers 1 19 23 Number of Full-Time Firefighters 670 485 704 (36%) (26%) (38%) Number of Fire Vehicles 42 52 78 (24%) (30%) (46%) Number of Ambulances 23 15 30 (34%) (22%) (44%) Number of Fire Stations 30 27 52 (28%) (25%) (48%) Population Servedl 429,300 277,000 710,000 (30%) (20%) (50%) Area Served1 (square miles) 61 76 43 1 (1 1%) (13%) (76%) Assessed Valuation in 19851 2,745 2,406 5,540 (in millions of dollars) (26%) (22%) (52%) Expenditures in 1985-fire Only2 30.2 13.1 31.6 (in millions of dollars) (40%) (1 8%) (42%) llncluding population, land area, and assessed valuation of areas served under contract. 2Does not include expenditures for emergency medical services. See discussion in text. departments range in size from Kinloch fire district, ditures per resident are higher in the city than in with five full-time fire fighters, to Chesterfield and either municipal or district departments in the Mehlville fire districts, with 59 and 75 full time fire co~nty.~Fire expenditures per firefighter in the city fighters respectively. University City has the largest are approximately equal to those for the average fire municipal fire department, with 50 full-time fire protection district, but higher than the average mu- fighters, and Rock Hill the smallest, with 11. nicipal fire department. Expenditures per $100 of The City of St. Louis employs approximately 1.6 assessed valuation are substantially higher in the city fire fighters per 1,000 residents (see Table 5.2). than the average expenditure for either municipal or Compared to St. Louis City, county municipalities district departments in the county.7 with their own fire departments employ, on the average, slightly more fire fighters per 1,000 residents, ECONOMIES OF SCALE? but fire protection districts in the county employ sig- A question of interest in St. Louis is whether nificantly fewer fire fighters. Fire protection expen- multiplicity, namely, a large number of producing

Table 5.2 Fire Service Ratios in St. Louis City of St. Louis County Fire St. Louis Municipalities Protection Districts Firefighters Per 1,000 Residents 1.56 1.75 0.99 Firefighters Per Square Mile 10.92 6.38 1.66 Firefighters Per Million Dollars of Assessed Valuation 0.24 0.20 0.13 Expenditures1 Per Resident $70.40 $47.30 $44.50 Expenditures1 Per Firefighter $45,000 $27,000 $45,000 Expenditures1 Per Hundred Dollars Of Assessed Valuation $1.10 $0.54 $0.57 Fire Stations Per 100,000 Residents 6.99 9.75 7.18 Fire Stations Per Square Mile 0.49 0.36 0.12 Fire Vehicles Per 1,000 Residents 9.79 18.77 10.99 Fire Vehicles Per Square Mile 0.69 0.68 0.18 (Does not include expenditures for emergency medical services. See discussion in text. units for fire services, is excessively expensive.8 Evi- such economies remain to be captured today, and dence from studies of fire protection found in the whether they are sufficient to overcome citizens' economics literature suggests that size economies preferences for local control of fire services cannot may be found for fire services up to a population be determined in this research.12 served of approximately lOO,OOO.Q Table 5.3 presents some evidence on this question, displaying av- COORDINATED FIRE PROTECTION erage per capita expenditures and expenditures per Coordinated production of fire protection serv- $100 assessed valuation by departments serving ices is found throughout St. Louis County. Fire serv- populations in several size ranges. ice providers, whether municipalities or fire protec- Average per capita expenditures for fire protection districts, have adopted ordinances that enable tion generally decline as the size of the population their fire departments to respond to requests for served increases, from an average of $66 per capita emergency aid from another department. These or- by departments serving fewer than 10,000 residents dinances authorize assistance to any jurisdiction that to $34 per capita in those serving 50,000 or more. adopts a similar ordinance. All of the municipalities By this measure, size economies are present in fire and districts with fire departments have adopted mu- protection, and economies might be captured by tual aid ordinances, allowing mutual aid to be prac- moving to larger service districts in St. Louis County. ticed in all pans of the county.13 It is not clear, however, that population served is the . Additional cooperative ventures have been best basis for such a comparison. Fire protection, worked out among the fire chiefs in the St. Louis unlike police or education, is a service focused area, including first-response agreements between somewhat more on property than on people per se, adjacent jurisdictions; the enactment of uniform although population size is also an important factor. areawide recruit training standards; recruit training Expenditures per $100 assessed valuation are not programs for entering fire fighters; joint training ex- lower in larger jurisdictions. Jurisdictions that are ercises for fire companies from different jurisdic- smaller in population tend to have greater concen- tions; joint dispatch centers for many of the depart- trations of property to be protected and, since fire ments; and the compilation of an inventory of services in much of the county are supported by apparatus and special equipment that can be re- property taxes, tend to have greater revenues from quested in an emergency. This coordinated produc- that property. On a property-protected basis, there tion of fire protection services has enabled the is no evidence of size economies.10 The data avail- county to obtain more favorable fire insurance rat- able on arrangements for fire protection in St. Louis ings than would have been obtainable otherwise. County, or rather the lack of data that would allow With the exception of the St. Louis City Fire questions of service demand and quality to be ad- Department, all of the fire service producers in St. dressed directly, preclude a definitive answer to the Louis are relatively small. Many have only a single question of size economies. Earlier studies of fire fire company on duty at any one time and, there- protection in the county, conducted during the fore, could be exposed to service deficiencies in the 1960s and, for a part of the county in 1981, found event of a large fire or multiple small fires occurring that economies could be achieved by merging simultaneously. Further, several of the fire service smaller departments and jurisdictions.11 Whether jurisdictions do not have a sufficient number of fire

Table 5.3 Size Effects in Fire Services: County Municipalities and Fire Protection Districts- 1985 Total Fire Assessed Equivalenl Population Number of Population Expenditures2 Expenditures Valuation1 Tax Served' Departments (in thousands) (in thousands) Per Capita2 (in millions) Rate3 Under 10,000 11 74 4,866 66 888 0.55 10,000-20,000 12 173 9,755 56 1,700 0.57 20,000-30,000 8 184 6,458 35 1,230 0.52 30,000-50,000 8 31 4 14,634 46 2,711 0.54 More than 50,000 3 234 7,958 34 1,418 0.56 Including contract populations, land area, and assessed valuation. 2Does not include expenditures for emergency medical services. See discussion in text. 3Equivalent tax rate is the rate per $100 assessed valuation required to raise the total fire protection expenditures of a department if the property tax were the only source of revenue. stations, or do not have them located, so as to meet the primary jurisdiction. As noted above, these first- insurance industry standards with respect to distance response agreements also help to lower fire insur- to the nearest fire station. In the absence of mutual ance rates. Through these agreements, fire service aid agreements, including first response where ap- producers have extended their service delivery propriate, citizens of these jurisdictions would pay boundaries beyond their jurisdictional boundaries, significantly higher rates for fire insurance. thereby increasing their capacity to respond to serv- Mutual aid agreements alleviate service defi- ice needs." ciency problems, especially in maintaining sufficient Fire chiefs in St. Louis City and County as well capacity to respond to emergencies. As phrased in as in adjoining counties, organized as the Greater St. one mutual aid agreement, "in the nature of fire- Louis County Fire Chiefs' Association, have been fighting work, it is always possible that a fire may instrumental in fostering cooperation and improve- rapidly grow so large that cooperation among differ- ments in the fire services of the area. Recruit and ent fire protection districts and fire services is a vital in-service training has been a significant focus of this necessity to the public welfare."l4 The agreements organization. Members of the association were in- among the municipalities and districts in St. Louis strumental in the adoption of an amendment to the County include provisions regarding compensation, St. Louis County charter that established minimum liability for loss or damage, legal status of fire fight- training standards for all recruits hired after January ers and fire equipment when responding to a request 1, 1984. These standards, recommended by the Na- for aid in another jurisdiction, command structures tional Fire Protection Association, require ten weeks in multijurisdictional fire operations, and notifica- of training. In St. Louis County, the training must be tion requirements. These agreements enable rapid completed within six months of employment. The response to requests for aid, avoiding the need for standards are enforced by the St. Louis County Fire negotiation or obtaining authorization in emergency Standards Commission, a body that the Chiefs' As- situations. sociation helped to establish. Calculations performed by the St. Louis County Association members arrange among themselves Department of Planning indicate that a number of for the recruit training classes and in-service train- areas in the county can be reached within three mining, using the facilities of one of their members. utes from as many as four fire companies. These They have been active in supporting the establish- back-up capabilities provide extra fire-suppression ment of a St. Louis County Police and Fire Training capacity for fighting major fires.'= Substantial back- Academy. The joint in-service training of cooperat- up capabilities exist in areas of high residential ing fire companies from adjacent jurisdictions has densities and commercial-industrial concentrations contributed to the reduction of fire insurance rates (e.g., Lambert-St. Louis Airport). in the county in recent years. The mutual aid pacts linking fire departments in Another area of cooperation is the annual com- St. Louis County have recently been extended to in- pilation of a Catalog of Apparatus and Special clude St. Louis City. The extended plan, involving Equipment, prepared for the chiefs' association by some 37 suburban fire departments in addition to Central County Emergency Dispatching Service. The the city department, received its first test early in catalog lists apparatus, equipment, and radio fre- 1988 when a major fire occurred in the city at the quencies used by all of the departments in St. Louis Royal Papers Building in midtown. Fire companies (including departments in surrounding counties) and from St. Louis County and other surrounding Mis- facilitates requests for aid and specialized equipment souri counties responded, providing a force of 115 when needed in emergencies. trucks to fight that fire, with additional trucks avail- Yet another form of fire service cooperation has able to deal with another fire in one of the county been the establishment of joint centers for the dis- fire protection districts at the same time.16 patch of fire and emergency vehicles. As discussed First-response agreements take mutual aid a step below, all of the fire protection districts and a few of beyond responding to requests for aid only when an- the municipal fire departments are dispatched by other department has an extraordinary emergency. one of three joint dispatch centers. These centers They are designed to place all citizens in the area were established through cooperative agreements within a reasonably close distance of a responding among the fire districts and departments involved. fire station, whether that station is in their own jurisdiction or not. These agreements specify areas FIRE AND EMERGENCY where a fire company from an adjoining jurisdiction SERVICE DISPATCH will respond automatically to a fire report, without Dispatching of fire and emergency service pro- the need for a request for aid, when that com- ducers in St. Louis City and County is linked by the pany is closer to the fire location than a station in areawide 9 11 emergency telephone system. Just as found in police services, this sharing of a common tion district, citizens vote separately to determine emergency network facilitates coordination of serv- whether they will remain within the fire district or ice production among the fire departments of the obtain fire services from the annexing municipality. area. In addition, like the organization of police If the annexing municipality does not have its services, fire and emergency service dispatching is own fire department, conflict is unlikely because the more concentrated than the production of basic annexed area will continue to be served by the fire services delivered directly to citizens. Twenty dis- district, and the district will continue to receive iden- patch centers supply communications service to the tical revenues, either from property taxes as before 43 municipal and district fire departments in the city or in equivalent payments from the annexing mu- and county. nicipality. If, on the other hand, the municipality This concentration is most evident for the fire proposing annexation maintains its own fire depart- protection districts, none of which maintain their ment, citizens may choose to receive services from own dispatch centers. The districts are dispatched by that department, removing their property from the one of three jointly established dispatch centers in tax base of the fire protection district with no com- the county. North Central County Fire Alarm Sys- pensatory payment required of the annexing munici- tems dispatches 13 district fire departments. Central pality. The boards of directors of some fire protec- County Emergency Dispatching Service dispatches tion districts have opposed annexations when they five district departments and one municipal depart- feared that the voters would choose to change fire ment, as does South County Fire Alarm. These joint service jurisdictions. In other instances agreements arrangements for the production of dispatch services have been worked out to avoid adverse impacts on allow the districts to deploy more of their personnel the fire district, and its directors have supported the for fire suppression activities, and contribute signifi- annexation. cantly to the implementation of mutual aid and first- A recent instance of opposition by fire district response agreements among the districts. directors involved proposed annexations by the City The remaining seventeen fire and emergency of Hazelwood of territory within the Robertson fire service dispatch centers are located in municipalities protection district. It appeared likely that if the an- that have their own local fire departments. With the nexations were successful the voters in the annexed exception of the City of St. Louis, which maintains areas would switch to Hazelwood for their fire serv- separate dispatchers for police and fire services, ices, because Hazelwood's property tax rate of 34 these municipal dispatch centers are shared across cents per $100 assessed valuation was less than one police, fire, and emergency services and are, for the half of Robertson's rate, and Hazelwood provided most part, operated by the local police departments. fire protection service to its residents. The This enables municipal departments in the county, Robertson directors opposed the proposal vigor- like their fire district counterparts, to deploy more of ously, threatening to withdraw from their mutual their personnel for fire suppression activities than aid agreement with Hazelwood if the annexations would otherwise be the case. Three municipal fire were successful. Hazelwood officials reported view- departments do not use these arrangements. ing this opposition as unwarranted, indicating that Crestwood and Frontenac are dispatched by joint Robertson had "plenty of work and plenty of reve- centers shared with adjacent fire protection districts, nue increase in unincorporated areas," and that the and Shrewsbury, through a contract with its neigh- small losses involved in the annexations should have bor, Webster Groves. little effect on the district. The annexations by Hazelwood were successful FIRE SERVICE CONFLICT and, as expected, the annexed area chose fire serv- Although cooperation and coordination among ice provision by the city. The revenue base lost by fire service producers in St. Louis County is wide- Robertson led the district to close one of its fire sta- spread, conflicts arise occasionally. The primary tions. Robertson also followed through on the source of such conflict is the revenue base of the fire threatened withdrawal of mutual aid. According to protection districts-real and personal property. The some reports, Robertson was able to coerce an ad- districts rely on property taxes levied against this joining district to do likewise, threatening to with- base for most of their revenues. Conflict can occur draw aid from it if it did not break with Hazelwood. when the tax base is threatened by an annexation This conflict appears now to have been resolved, proposal from an adjoining municipality. Under with restoration of mutual aid agreements, although Missouri law, citizens in an area that is the target of there are still some harsh feelings. an annexation by an adjoining municipality must ap- A similar instance was resolved without acri- prove the plan by a majority vote. If the area to be mony when the City of Town and Country success- annexed lies within the jurisdiction of a fire protec- fully annexed areas within the jurisdictions of the Manchester and the Chesterfield fire protection dis- absence of such data, proxy measures of perform- tricts. Prior to the annexation, Town and Country ance include the Insurance Service Organization had maintained its own combined public safety de- (ISO) rating and evidence from surveys of fire serv- partment, with officers serving both as police and ice consumers. fire fighters. Town and Country might have chosen A jurisdiction's IS0 rating is based on water to expand this department to supply all of the an- supply, equipment, personnel and their level of nexed area, substantially reducing the revenue base training, the distances fire trucks would have to of the Manchester district and, to a lesser extent, travel to reach a fire, and the number of trucks the Chesterfield district. Had this been done, how- available.19 The ratings range from 1 for the lowest ever, those districts might have been reluctant to risk jurisdiction to 10 for one with a high fire risk. continue mutual aid and first-response agreements. Ratings of 9 and 10 indicate a lack of fire hydrants To avoid this, officials of Town and Country ad- and more than five miles travel distance to the near- vised residents in the newly annexed portions of the est fire station. Three-fourths of the fire jurisdictions Chesterfield fire protection district to vote to remain in the county have an IS0 rating of 4, considered a in that district and advised residents in those por- good rating in the fire service. The remainder have tions of the Manchester district that were annexed ratings of 5 or 6, with some areas rated 9 in the most to vote to have fire services provided by Town and rural portions of the county where there are no hy- Country. To prevent a major loss to the Manchester drants.20 Ratings of most jurisdictions have improved district, Town and Country officials proposed to dis- during the 1980s, in part because of the new fire band the fire service portion of their public safety service training standards which have been adopted department and to enter into a contract for fire pro- in St. Louis County and also because of enhanced tection throughout the municipality. Two small mu- mutual aid and first response agreements coupled nicipalities lying within the borders of Town and with interjurisdictional training21 Country would also contract with the Manchester Citizens in the St. Louis area appear to be well district. New residents of Town and Country from satisfied with their fire protection services. In 1982, the Chesterfield district chose to ignore the advice, 87 percent of city residents and 89 percent of voting instead to have fire service provided by Town county residents said that they were "somewhat" or and Country. Otherwise the plan was implemented "very satisfied" with fire services, the highest rating as proposed, thus avoiding conflict with the Manch- obtained by any service." In 1987, more than 96 ester district. An interesting feature of the contract percent of county residents said that their current is that Town and Country retained title to its fire fire protection is "good" or "excellent."23 While not station and equipment. Its mayor argued that this measuring service performance directly, these per- maintained the city's options in case of any future centages suggest that fire services in St. Louis are disagreement with Manchester. very well thought of by their consumers. While annexations of a fire district jurisdiction are a potential source of conflict, these examples in- CONCLUSION dicate that conflict need not always occur. An agree- A variety of jurisdictions are used to provide fire ment like the one between Town and Country and protection in St. Louis City and County. Citizens in the Manchester district can avoid such conflict, al- different parts of the area have a different range of lowing the maintenance of the vital mutual aid pacts choices available to them-sometimes narrower, in the county. To further protect mutual aid, offi- sometimes broader. The formation of the City of St. cials might consider changes in the annexation laws Louis was, implicitly, a choice to provide fire serv- that would require compensation of fire districts for ices through a large municipal unit. Conceivably, capital investments made in service areas that are citizens could seek to alter this arrangement, per- subsequently annexed, perhaps involving purchase haps by dividing the city into several fire protection of the facilities and equipment by the annexing mu- districts, but such options are not currently avail- nicipality if it wishes to supplant the district as serv- able. New state enabling legislation would be re- ice producer for the annexed area. quired before citizens could exercise such a choice. Citizens of the unincorporated county, as long as FIRE SERVICE PERFORMANCE they remain unincorporated, have no choice but to Ideally, one would assess the performance of obtain provision by means of a fire protection dis- fire departments by comparing their fire incidence trict. District boundaries, however, are subject to and loss records to those of other departments serv- citizen choice. In the event of an annexation pro- ing comparable areas. Practically, this is not possi- posal by an adjacent municipality, citizens in the ble, as data on fire incidence and/or loss are not area to be annexed retain a choice between district maintained in any publicly accessible form.l8 In the and municipal provision even if they vote favorably for annexation. Citizens living in an unincorporated Number of area may also decide on the incorporation of a new Area Employees municipality that might provide fire services. Mu- Berkeley 23,041. Clayton 30,920 nicipal incorporation does not necessarily entail pro- Community 21,675 vision of fire services, however. Citizens of county Creve Coeur 31,715 municipalities still may choose, collectively, between Fenton 17,241 Hazelwood 19,706. municipal and district provision. Given the opportu- Maryland Heights 30,141 nity to choose, citizens in different areas tend to *Includes portions of Lambert Field and McDon- make different choices, depending on their circum- nell-Douglas stances. The result is a complex, varied, and far SOURCE: Quoted from East-West Gateway Coordi- from uniform pattern of jurisdictional arrangements. nating Council, "1980 Employment Data Base" in St. Louis County Department of Planning, Fire Pro- tection and Emergency Medical Services in St. In spite of, or perhaps because of, the complex Louis County, a report prepared for the St. Louis set of jurisdictions found in St. Louis County, the County Annexation Study Commission, June 24, functional arrangements for fire service production 1984. include a rich skein of coordination. Fire depart- this amount nearly $9 million was fire retirement payments in the county regularly assist one another with ments. Direct expenditures for fire protection and associated communications were approximately $21.5 million. personnel and equipment, arranging jointly to supply The cost of EMS services in the city are not included in services where they are needed. Rather than finding these figures. themselves hampered by jurisdictional boundaries, e~xpendituresdiscussed in the Chapter are for fire protection fire service producers have for the most part agreed only, not including expenditures for emergency medical to work together across such boundaries. Most re- services (EMS) by those departments that produce them. EMS and ambulance expenditures were subtracted from to- cently, fire service coordination has acquired a more tal fire department expenditures for those departments extensive metropolitan dimension, with the addition (principally fire districts) which listed them separately. For of St. Louis City to suburban mutual aid agreements. the remainder, the percentage of total expenditures allocated to fire protection was taken as the same as that com- The cooperative arrangements worked out by fire puted by the St. Louis County Department of Planning for chiefs and municipal and district officials appear to expenditures in 1982 (see Note 4). This estimation process, be serving county residents to their satisfaction. In while the best available, may not be wholely accurate, and all but the most rural parts of the county, the func- readers should interpret results with some caution. tional arrangements associated with jurisdictional '~xpenditures per $100 of assessed valuation may be interpreted as an "equivalent tax rate," that is, the tax rate fragmentation have resulted in quite respectable fire which would be required to support fire protection services if insurance ratings. all revenues were raised by a tax on local property. %ee Confluence St. Louis, Too Many Governments? A Re- port on Governmental Structure in St. Louis City and County with Recommendations for Change (St. Louis: Feb- ENDNOTES ruary 18, 1987), p. 22, citing 43 fire chiefs as "clearly inefficient" because of administrative costs incurred. However, fire chiefs of virtually all departments in the county are ac- 'Many producers of fire services in St. Louis also produce tive fire fighters as well as administrators, more akin to fire emergency medical services. The Chapter focuses on the company commanders in larger departments. fire protection activities of these producers only. gSee Werner Z. Hirsch, Urban Economics (New York: 2~acificand the Pacific Fire Protection District are not in- Macmillan Publishing Company, 1984), pp. 272-3. cluded in these counts, as Pacific has only 12 residents in St. Louis County. The reference to "one or more.. .fire pro- 1°~hereis wide variation in expenditures per capita and per tection districts" reflects the fact that several municipalities $100 assessed valuation among the departments in each in St. Louis are bisected (and even trisected) by the category. Expenditure and population estimates, together boundaries of two or more fire protection districts. with actual assessed valuations are shown in the Appendix to this Chapter. 3~hiscount does not include as a separate department the Louis County, fire district maintained at Lambert-St. Louis Airport by the llAnalysis of Fire Protection Services: St. City of St. Louis, nor fire companies organized by non- Missouri (Westchester, Illinois: Gage-Babcock & Associ- governmental entities such as the McDonnell-Douglas Cor- ates, December 1966) ; Douglas J. Harms, "Municipal Fire poration. Service: An Alternate Approach, " M. P. P. A. Exit Exam, University of Missouri-St. Louis, December 1981. 41f daytime population, including nonresidents employed in 120n the issue of local control, see Market Opinion Re- the city and persons entering the city for shopping and other search, "Public Opinion on Government Reorganization Al- purposes were included, as many as 25O,OOO-3OO,OOO addi- ternatives for the City of St. Louis and St. Louis County" tional persons should be added to the number served by the (Detroit: April 1982). city fire department. Similar additions to populations served would be appropriate for major county employment and 13st. Louis City has mutual aid agreements with municipali- shopping areas. In 1980, county municipalities and fire dis- ties and districts which surround the Lambert-St. Louis Air- tricts with large daytime employment (in excess of 15,000 port, however. The city maintains a fire company at the persons) included: airport (which is owned by the city), and this company co- operates with surrounding jurisdictions and the private de- dinated production results, just as it would in a large juris- partment maintained by the McDonnell-Douglas Corpora- diction (e.g., the city) that dispatched two or more fire tion. companies to a major fire. For a discussion of the structure of public service industries, including both coordination and 140rdinance No. 2 of the Eureka Fire Protection District. alternation patterns, see Elinor Ostrom, Roger B. Parks, dated October 18, 1984. and Gordon P. Whitaker, Patterns of Metropolitan Polic- l5For a graphic representation of fire service redundancy, see ing (Cambridge, Mass. : Ballinger Publishing Company, Confluence St. Louis, Too Many Governments? Appendix, 1978). p. 54. Confluence considered these back-up capabilities as lasuch data might be available from fire insurers, but is not evidence of service du~licationrather than of valuable re- regularly compiled by any public organization. dundancy providing eGra fire suppression capacity for major fires. leSee St. Louis Department of Planning, Fire Protection and Emergency Medical Services, p. 5. 16"~utualAid Pacts Pass First Fiery Test," St. Louis Post- 20~t.Louis County, Missouri Fact Book-1986 (Clayton: St. Dispatch, March 6, 1988, referring to fires that occurred Louis County Department of Planning, 1986), p.89. during the evening of February 7, 1988. 210n the latter point, see St. Louis Department of Planning, 171n the language of "public service industries," these agree- Fire Protection and Emergency Medical Services in St. ments create "alternation in space" that transcends jurisdic- Louis County. tional boundaries. That is, residents of a given fire jurisdic- *~arket Opinion Research, Public Opinion on Government tion may, as a result of such agreements, receive fire suppression services from one of two or more different fire Reorganization Alternatives. Approximately 1,000 citizens producers, depending on their location in the jurisdiction. were surveyed, two-thirds of whom were county residents. These producers do not duplicate one another's efforts, for 23"~hesterfield and U. City Lead in Supporting McNary they have agreed in advance on boundaries within which Remap, " St. Louis Post Dispatch, November 27, 1987, pp. they will respond. In the case of major fires, where two or 1A and 6A. Approximately 2,500 county residents were more companies from multiple jurisdictions respond, coor- surveyed. Appendix Table 5.1 Fire Protection Data for St. Louis Personnel-Equipment Assessed 1985 Equivalent Full-Time Munlclpality Resident Land Valuation Expenditures Tax Fire or District Population Area (thousands) (fire only)' Rate2 Fighters Vehicles Stationc City of St. Louis 453,085 61.37 2,744,652 30,208,208 1.10 Municipal Fire Departments Berkeley 4.87 124,698 781,674 0.63 Brentwood 2.16 92,903 526,543 0.57 Clayton 2.54 276,546 756,563 0.27 Crestwood3 3.15 1 12,571 544,654 0.48 Des Peres4 4.15 1 10,597 436,689 0.39 Ferguson 5.23 122,858 836,420 0.68 Frontenac3 2.90 77,503 375,150 0.48 Glendald 1.32 47,771 374,695 0.78 Hazelwood 4.90 188,044 1 ,177,975 0.63 Jennings3 3.78 74,352 748,362 1 .O1 Kirkwood 8.94 21 2,445 l,l3O,517 0.53 Ladue 8.57 203,125 952,064 0.47 Maplewood 1.55 60,570 409,129 0.68 Olivette 2.76 80,951 708,283 0.87 Richmond Heights 2.35 104,119 436,970 0.42 Rock Hill 1.08 33,110 248,578 0.75 Shrewsbury 1.42 44,700 196,053 0.44 University City 5.86 218,184 1,247,846 0.57 Webster Groves 5.79 168,927 918,486 0.54

Total Municipalities 269,994 73.32 2,353,974 12,806,652 0.54

Fire Protection Districts Affton 8.005 243,604 984,543 0.40 Ballwin 54.00 376,537 2,223,493 0.59 Black Jack 26.00 235,041 1,175,205 0.50 Chesterfield 63.00 555,931 3,859,795 0.69 Community 11.00 307,136 1,665,481 0.54 Creve Coeur 16.00 51 5,730 2,115,757 0.41 Eureka 43.00 51,149 524,901 1.03 Fenton 19.00 250,421 1,521,856 0.61 Florissant Valley 22.00 386,913 2,092,086 0.54 Kinloch 1.00 8,324 36,859 0.44 Lemay 5.00 102,697 487,061 0.47 ManchesteP 16.00 268,813 2,062,340 0.77 Maryland Heights 9.00 31 6,265 1,158,515 0.37 Mehlville 47.00 724,024 4,200,956 0.58 Mid-County 4.00 94,274 867,321 0.92 Moline 4.00 89,413 620,729 0.69 Normandy 7.00 127,081 965,363 0.76 Pattonville- Bridgeton Terr. 16.00 181,373 848,989 0.47 Riverview 7.00 100,858 605,148 0.60 Robertson 16.00 1 64,675 1 ,136,258 0.69 Spanish Lake 15.00 1 16,782 700,692 0.60 Valley Park 9.00 50,990 428,316 0.84 West Overland 3.00 102,716 582,584 0.57

Total Fire Districts 698,000 421.00 5,370,747 30,864,248 (continued on next page) Appenfix Table 5.1 (cont.) Fire Protect~onData for St. Louis Personnel-Equipment Assessed 1985 Equivalent Full-Tlme Municipality ResMent Land Valuation Expenditures Tax Are or Distrlct PopulaUon Are8 (thousands) (fire only)' Rat$ Fighters Vehlcles StaUonr

Contract Cities Country Club Hills Country Life Acres Crystal Lake Park Flordell Hills Huntleigh Oakland Town and Country Warson Woods

Total Contract Cities

NA - Not Applicable. 1Does not include expendires for emergency medical services. See discussion in text. 2Equivalent tax rate is the rate per $100 assessed valuation required to raise the total fire protection expendires of a department if the property tax were the only source of revenue. 3Department produces for additional population under contract. 4Combined fire-police public safety department. Expenditure estimated for fire protection only. 5Population and land area from 'Fire Protection and Emergency Medical Service in St. Louis County,' Clayton. Missouri: St. Louis County Planning Department. June 25.1984.

Chapter Six Streets and Street Services

INTRODUCTION incorporated area, the same responsibility is shared (to a lesser extent) between subdivisions and county Four different types of jurisdictions are involved government. Throughout the county, arterial street in street service provision in St. Louis: the state, the responsibility is shared by state and county govern- county, municipalities, and private subdivisions. ments. In St. Louis City, residential street responsi- Nearly all of the 90 municipalities in St. Louis bility is shared by city government and subdivisions County provide street services.' County government (though the city is the dominant provider), while ar- provides both residential street services in unincor- terial responsibility is shared by state and city gov- porated areas and arterial street services county- ernments. wide. The state is also a significant provider of street Provision units frequently are not involved in services in the St. Louis area by virtue of the numer- street service production. Subdivisions and small ous state highways routed into the region. When pri- municipalities tend to contract out privately for the vate subdivisions are included, the provision of production of most street-related services, while street services is by far the most fragmented of any larger municipalities (those over 5,000 population) local service in St. Louis. tend to produce most services in-house. Production Subdivisions, although legally private jurisdic- units include the St. Louis County Department of tions, function as collective provision units in a man- Highways and Traffic, the state highway department, ner very similar to a municipality. Ranging in size departments of streets or public works and depart- from a few to more than a hundred homes, subdivi- ments of planning in most mid-to-large municipali- sions are governed by a board of trustees elected ties, and an undetermined number of private firms. annually at a meeting of property owners. The legal basis for a subdivision is a "trust indenture" at- THINKING ABOUT STREETS tached to each parcel of property. Besides providing In economic terms, streets can be understood as for trustees, the indenture obligates each property a local public good and common pool facility. As a owner to an assessment for the maintenance of public good, streets are made available to all if made "common grounds," which in St. Louis County, and available to one. This implies nonexclusion-in the to some extent in St. Louis City, often includes case of streets, due to the high cost of excluding in- streets. Payment of the assessment is treated as a dividual users. Only a very weak form of exclusion lien on the property, thus providing an eventual, can be maintained by means of vehicle licensing. though not immediate, enforcement mechanism. Demand for streets therefore tends to be expressed Shared responsibility between overlapping juris- collectively through various nonmarket institutions. dictions for both residential and arterial types of As a public good, streets also exhibit jointness: streets, but not for particular streets, is the general many individuals use the same facility together, at pattern of provision. In the incorporated area of the the same time. Streets are, however, a "congestible" county, subdivisions and municipalities share republic good, that is, there is some maximum number sponsibility for residential streets, while in the un- of users after which each user begins to subtract from the use of others. Once supplied, streets can pletely separated in the absence of effective also be considered common pool facilities, and, as exclusion of individual users, given the interconnect- such, present a number of problems affecting the edness of a street system and (2) the congestibility ability of individuals to use street facilities jointly. of streets means that street uses sensitive to high vol- Users can subtract from one another's welfare in a ume traffic can be threatened as motorists seek out variety of ways, such as excessive vehicle weights less traveled streets to avoid congestion. Those who and speeds. Rules of the road, adjusted for the type derive a benefit from high volume streets, such as of street and its intended uses, are a response to the commercial developers, may be in conflict with common pool problem inherent in street use. De- those who feel negative spillover effects from high pending on the rules, any one user-or any one type traffic volume, such as abutting residents. of use-can potentially dominate others. Somehow The provision of streets and related street serv- the interests of truckers, motorists, pedestrians, and ices presents, therefore, a somewhat complex gov- street-side residents must all be accommodated to ernance problem. Provision includes decisions about one another in the use of a street system.2 Managing facility location, type, design, and subsequent regu- or governing the relationships among diverse users is lation of use. Various communities of interest, from an essential element in the provision of streets. those quite small in scale (e.g., a group of neigh- The uses of streets are more varied than is ordi- bors) to those of wider scale (e.g., commuters), narily supposed. Getting from one place to another must be taken into account. The economic demand by motor vehicle is a problem that varies somewhat for streets is far from homogeneous in a metropoli- between short trips and long trips. Different types of tan area such as St. Louis. Jurisdictional arrange- motor vehicles place varying demands both on the ments for the provision of streets might be expected facility and on other users. Use by heavy trucks is to develop in a highly differentiated fashion, includ- different than use by passenger cars. It is also useful ing both fragmentation and overlap, to the extent to distinguish access as a special problem distinct allowed by the rules for the formation of local juris- from traveling between two points. In the common dictions. law, the owners of "abutting property" enjoy a legal Street provision can be sorted into five basic right of reasonable access to a public street system. service components, each associated with somewhat A street that facilitates access, however, may not fa- different production processes: cilitate through-travel. (1) Development, including design, location, Streets are also used by pedestrians and nonmo- and level of capital investment; . torized vehicles, for whom the street right-of-way is (2) Surface maintenance-both preventive and frequently used to provide sidewalks or-in a more repair; recent innovation-bicycle paths. Outside the right- of-way, but affected by its use, are those who reside (3) Maintenance of the right-of-way, including or do business along a street. In the common law, services such as sweeping and cleaning, these abutting residents or users may have rights to mowing, tree trimming, snow removal, and be free of nuisance (e.g., dust or noise) caused by street lighting; the street and its users. More generally, a street has (4) Reconstruction-rebuilding an existing neighbors, and these neighbors can be the recipients street, sidewalk, or curb; and of either positive or negative spillover effects as a (5) Regulation of use, including access restric- result of street location, design, or use. From this tion, rules of the road, and parking. perspective, one of the uses of a street, to be accommodated by providers, is to make a residence or Streets can also be sorted into three basic facility place of business along its path. types: Streets are characterized, therefore, by a num- (1) Arterial streets-often limited access, usually ber of interdependent uses-in part complementary, higher speed and higher traffic volume; in part competitive. One response to the competi- (2) Residential streets-specialized to local ac- tiveness of uses is to sort different types of users cess, low speed and low traffic; and among different types of facilities. Between a limited access, median-separated highway and a residential (3) Collector streets-that link residential streets to arterial streets. cul-de-sac lies a considerable range of street facility types. Some streets are used mainly for the rapid Facility type and location serve as the primary movement of high volume traffic while others serve basis for distributing street provision responsibilities primarily for access to residences and to accommo- among various provision units. Service components date children at play. The separation of uses, how- tend to serve as a basis for distributing production ever, has difficulties: (1) uses can never be com- among production units. The analysis of street provision and production county, especially the rapidly growing parts. County- that follows is organized first by service component wide bond issues to support highway development and, within service component, by facility type. Sur- have sometimes been defeated at the polls while en- face maintenance and right-of-way maintenance are joying substantial support in fast growing areas of the collapsed in this discussion because the same provi- county, such as Chesterfield. Countywide road bond sion units, and frequently the same production units, issues were approved, however, in 1955, 1965, tend to be involved in both types of maintenance on 1969, 1977, and 1986. Incorporation activists envi- the same facilities. Collector streets are also col- sioned a significant highway development role for lapsed with either arterial or residential streets, while their projected city. Current law, however, gives a maintaining the distinction where salient to the dis- predominant position to state and county govern- cussion. ments, although a municipality is not precluded from constructing new streets and roads or improving ex- DEVELOPMENT isting streets. The legal position of a municipality is weak if the effort is to oppose a new highway devel- Arterial Streets opment, or to modify its design, but municipalities Arterial responsibility is shared between state remain free to add their highway development ef- and county (between state and city in St. Louis forts to those of the state and county, realizing that City). The state highways routed through the region county government could assume jurisdiction over a and the county's arterial road system both serve a newly constructed road. The immediate concerns of high-speed, high-volume, traffic function. State and Oakville incorporation activists are somewhat less county highway planners, while organized sepa- likely to be fully addressed by incorporation, insofar rately, also coordinate extensively out of necessity. as county government would retain its present juris- The resolution of traffic congestion problems in par- diction over arterial streets and roads in a newly in- ticular locations frequently requires a joint state- corporated city. county effort, especially where county and state One way to conceptualize the role of a munici- highways intersect. Because each jurisdiction may pality in street development is as a specialist in col- represent somewhat different communities of inter- lector streets. The newly expanded City of Town est, the arrangement may also require negotiation. and Country is beginning to view itself in this light, Both state and county enjoy preemptive powers. and the newly incorporated City of Chesterfield Neither the county nor other local units can veto a could move in this direction. Collector streets, nev- highway location decision made by the state. The county government also can transfer a municipal ertheless, are closely related to arterial streets. As street (but not a private street) to county jurisdiction new municipalities are formed in the midst of by incorporating it into the county arterial system. In rapid economic growth, the planning and develop- this sense, the wider scale interests associated with ment of arterial streets is likely to become a three- high-volume transit have a legally superior position. party concern, with the resolution of problems nego- Yet other communities of interest also have influ- tiated among state, county, and municipal officials. ence in county government. County government has begun to tie economic As provider of the county arterial system, the development to highway development by means of a county government has an important role, alongside "traffic generation assessment" fee, charged to de- the state highway department, in highway planning velopers as a condition for obtaining building per- and development. It is this function for which mits. Although the legal basis for denial of a building county officials have been heavily criticized in recent permit on these grounds is unsettled, developers years with respect to increasing traffic congestion in usually choose to pay the fee rather than litigate the parts of the county. Traffic congestion was viewed as issue. The county government recently came to an a major issue in the recent Chesterfield incorpora- agreement with area developers (represented by the tion. Moreover, a fatal traffic accident, plus per- Home Builders Association) that stipulates assess- ceived lack of responsiveness by county government ment rates to be assessed uniformly on each devel- to local demands for traffic regulation, was the im- opment. The agreement was negotiated through the petus for incorporation discussion in Oakville, an Road Development Advisory Board appointed by area in South County, ongoing during the field re- the County Executive. In this way, developers can search for this report in 1986. be required to pay for extensive improvements to Advocates of incorporation in the new City of the highway system in the immediate vicinity of their Chesterfield argued that the entire county was an in- project, based on the additional traffic being gener- appropriate collective unit to express the demand ated. County officials have unsuccessfully sought for arterial street development in parts of the state legislation that would enable them to use the revenue collected from traffic generation assessment ices are produced by the maintenance division of the fees anywhere in the county. county Department of Highways and Traffic. The The federal aid urban highway program is a ma- major funding sources are county sales taxes. Local jor source of funds for capital improvement projects. property taxes are not used to support the arterial Other funding sources include countywide bond is- system. State highways in the county also provide sues and a transportation sales tax. New highway de- arterial service and are maintained by the state high- velopment must sometimes compete with recon- way department. struction of existing facilities for funding from these sources. Capital improvement projects include inter- change and bridge improvements as well as resurfac- Residential Streets ing and concrete replacement. Responsibility for residential street maintenance is shared between subdivisions and municipalities in Residential Streets the incorporated portion of the county and, to a The development of purely residential streets is lesser extent, between subdivisions and the county a much different problem from the development of government in the unincorporated portion. Subdivi- arterial streets. The critical difference is that resi- sions account for a major percentage of street mile- dential street development can be more closely tied age in only six municipalities, and are the dominant to housing development. Those making the basic de- type of provider in two municipalities. Nine cities in velopment decisions with respect to residential the county spend more than $1 million on street streets tend, therefore, to be housing developers. services (and other public works) annually. This generalization holds both historically and at The total number of subdivisions providing present in St. Louis County. Initial decisions of de- street services in St. Louis City and County is un- sign and capital investment are private, subject to known. A total of 427 street-providing subdivisions market constraint and local government regulation. are reported in 27 municipalities, but these figures These development decisions include a choice of subsequent provision arrangements for obtaining are incomplete. Within the incorporated area of the street services. It is the developer who initially at- county, the greatest number of subdivisions are con- taches trust indentures to subdivision parcels and centrated in a cluster of municipalities consisting of thereby establishes the subdivision as a legal unit University City, Clayton, Ladue, Richmond Heights, with collective organization. Whether streets are Creve Coeur, Olivette, Kirkwood, and Town and public or private is also determined in most cases at Country. Ladue and Olivette are virtually blanketed the developer's discretion. by street-providing subdivisions. Ladue (population Various public interests in residential street de- 9,349) has 143 subdivisions; Olivette (population velopment are represented indirectly by means of 7,952) has 90 subdivisions. Town and Country, county and municipal subdivision regulation. The prior to a recent annexation of additional territory, relevant production units are the St. Louis County also consisted almost entirely of private streets in an Department of Planning (for the unincorporated estimated 90 subdivisions. University City, Clayton, county), the City of St. Louis Community Develop- Richmond Heights, Creve Coeur, and Kirkwood mix ment Agency, and other municipal planning depart- m'unicipal and subdivision provision, but with greater ments. In order for subdivision streets to be ac- reliance on municipal provision. The number of cepted by the county or municipality, construction county (nonmunicipal) subdivisions that provide has to be performed according to standard. Most street services could not be determined, although standards now apply whether streets are to be pri- county officials say that private streets are less com- vate or public, including pavement thickness in un- mon in the unincorporated area. A small number of incorporated St. Louis County. If the streets are to street-providing subdivisions can also be found in St. be private, the county requires the developer to pro- Louis City. Many of the smaller municipalities in the vide for a trust indenture and board of trustees. At county originated as subdivisions and are also en- least one municipality-Manchester-requires that tirely residential, having little or no state or county all subdivision streets be constructed to the municipal standard for acceptance by the city, thus pre- street mileage within their boundaries. cluding the development of private streets. To study street provision by subdivisions in St. Louis County, data on 53 subdivisions were collected from a sample drawn from Ladue and MAINTENANCE Olivette, where private streets are dominant, and from University City and Clayton, where private Arterial Streets streets are less prevalent. Information was collected At the end of 1985, the county arterial system by telephone interviews with subdivision trustees and included 397.22 miles of streets.3 Maintenance serv- mail-in questionnaires. holdq4The reported assessments vary widely, from a Table 6.1 low of $25 to a high of $800. All but two cases, Percentage of Subdivisions that Provide Common Street Services however, lie between $25 and $275; the mean assessment computed over this range is about $103. In Percent Sample a small percentage of cases, subdivision assessments Service Providing Size also support nonstreet services, such as trash collec- Snow Removal 94 53 tion, a private security patrol, or park maintenance. Street Repair 92 49 (Assessments may also vary substantially from year Sweeping 47 53 to year as subdivisions occasionally undertake ex- Mowing 41 51 traordinary projects best understood as capital im- Tree Trimming 36 53 provements and discussed below.) Street Lighting 3 1 52 Some interesting variation in subdivision assess- Other Street Services* 17 53 ments occurs among the four municipalities studied: Nonstreet Services 15 53 University City (N=6), Clayton (N=6), Olivette Access Restriction or (N= 18), and Ladue (N=23). For this portion of the Traffic Control 36 53 analysis, data for University City (population Recent Capital 42,690) and Clayton (population 14,306) are col- improvement 57 53 lapsed. Both cities are more urban than the others *Entrance lights and signs are the major 'other" services provided. and have a more similar socioeconomic composition. University City is the more heterogeneous of the two; Clayton is upper-middle in income. Olivette (population 7,952). on the other hand, is a rather Subdivisions as densely populated, but quite suburban, upper-mid- Street Maintenance Providers dle income community, while Ladue (population The most common street services-indeed the 9,369) is a large-lot suburb considered to be one of most common services of any kind-provided by the richest in America. subdivisions are street repair and snow removal. Subdivisions in Olivette tend to have the lowest More than 90 percent of the 53 subdivisions sur- assessments by far-averaging $67 per household. veyed for this study provide these two services. Assessments in Ladue are almost twice that Other street-related services include sweeping, light- amount-averaging $123 per household. Two factors ing, tree trimming, and mowing. Although only a may contribute to this difference: (1) the greater few subdivisions provide a full range of these serv- wealth in Ladue and (2) the lower density of popu- ices, the average number of different services relation, probably increasing the ratio of street miles to ported is almost four per subdivision (including non- households. The average assessment in University street services). City-Clayton, however, is still higher-$292 per As indicated in Table 6.1, subdivisions make household, though the mean score for this group is use of their ability to choose a mix of services in driven up by a single subdivision reporting an $800 accordance with their preferences. The result is con- assessment.5 The average total number of services siderable variation in service bundles across subdivi- provided (not counting access restriction) also var- sions. Provision by subdivisions allows for greater ies. It is highest in University City-Clayton-5.7 out variation in service bundles than would provision by of a possible 9, explaining the high assessments an overlying municipality. By and large, subdivisions found there, and lowest in Olivette-3.3. Ladue, are not supplementary service providers, augmenting somewhat surprisingly, has a score of only 3.8. Per- service provision by municipalities, but independent haps the greater number of services provided in Uni- providers who either provide for themselves or do versity City-Clayton is linked to the more urban without. The only general exceptions noted are character of those communities. Greater wealth in street lighting, which in University City is supple- Ladue does not lead to a greater number of services mentary to municipal provision, and nonstreet serv- provided, though it may lead to somewhat better ices, such as a security patrol. A few municipalities street conditions. The state of street repair is re- provide selected services free of charge to subdivi- ported to be somewhat worse in Olivette, better in sions-snow removal by the cities of Creve Coeur University City-Clayton, and better still in Ladue, al- and Webster Groves, a broader range of services in- though the differences reported are not statistically cluding minor surface repair by the City of significant .6 Kirkwood. A difference also exists in service provision The mean annual subdivision assessment, for a equity between subdivisions in Ladue and Olivette, group of 39 subdivisions, is about $130 per house- where street-providing subdivisions virtually blanket the city, and those in cities such as University City - ~~~~~ and Clayton, where street-providing subdivisions Table 6.3 cover only a portion of the city. In the latter cases, Relation between Munlclpal Population Slze and residents of street-providing subdivisions must pay to Production Arrangement support municipal street provision from which they For Mlnor-Routine Street Repair and gain only a partial benefit. The subdivision assess- Maintenance ment is paid on top of a municipal tax bill to support Number ot residential street provision elsewhere in the city. In Number ot In-house order for a subdivision to be cost effective, the Contracting Production added benefit from subdivision control in University Population Munlclpalities MuniclpallUes City and Clayton, as compared to municipal control, 500 or less 5 0 must be worth the entire amount of the subdivision 501-1,000 4 1 assessment. By contrast, in Ladue and Olivette, vir- 1,001-2,000 4 3 tually all residential street maintenance is private. In 2,001-5,000 1 7 this circumstance, the cost effectiveness of private 5,001 or more Q 24 streets requires only that the subdivision assessment 3 1 be worth the entire benefit gained from services provided.' A multivariate analysis of the 39 subdivisions for good, excellent. As Table 6.2 shows, the responses which data on household assessments is available are heavily skewed toward a favorable evaluation. shows that two variables are significantly related to More than 80 percent of the sample appears to be the amount of the household assessment: satisfied with the quality of street maintenance they w Whether or not the subdivision restricts ve- are able to provide. hicular access or traffic flow in any way; Subdivisions overwhelmingly tend to be pure Total number of services provided. provision units. Only a few of the smallest subdivisions produce street repair services for themselves, Access restriction is included here as a proxy for the the vast majority choosing instead to contract with "urban character" of a subdivision. Together, these either a private firm or the overlying municipality. two variables measure the demand of residents for Those that do self-produce tend to do so not by or- services-one somewhat more directly than the ganizing a separate production unit but by pooling other. The age of a subdivision and the number of their efforts as household "do-it-yourselfers." One households, on the other hand, are not significant in large subdivision has a full-time employee, hired to this equation. Both of these variables can be used to do snow removal and mowing. Olivette produces reflect the cost of ,supplying a given range of services both sweeping and snow removal for most of its sub- demanded. Subdivision age may increase the costs divisions under contract, except a few of the larger of supply while a greater number of households may ones. Clayton produces street sweeping under con- decrease the costs of supply per household. These tract to all 11 subdivisions. For street repair, subdivi- findings indicate that the level of subdivision assess- sions usually contract privately. ments within this small sample tends to vary with the level of demand for services, and is explained more Municipalities as on the demand side than on the supply side. Street Maintenance Providers The 53 subdivisions surveyed were asked to The smallest municipalities, like the subdivisions evaluate the condition of their streets. Their re- they closely resemble, also tend to produce few if sponses were coded on a four point scale: poor, fair, any street services and instead contract with private producers. Analyzing data collected from 45 municipalities,8 there is a clear relationship between popu- Table 6.2 Subdivisions' Evaluation of lation size and the use of contracting. In this group, Street Conditions no municipality with a population of less than 500 produces street repair or snow removal in-house. Evaluation Frequency Percent and no municipality with a population greater than Poor 2 3.77% 5,000 contracts to obtain these same services. Table Fair 8 15.09 6.3 displays a detailed breakdown of street repair Good or Average 28 52.83 production arrangements by the population size of Excellent Ei 2iUi2 municipalities. Total 53 100.00% For all street-related services, municipalities of more than 4,000-5,000 people employ a mix of contract services and production by their own crews, but this analysis, as well as the inability to determine tend to favor in-house production. The vast majority service quality, could mask the existence of size of all municipalities for which data is available re- economies. ported using contract production for resurfacing and major repairs. These services are followed, in the County Government as frequency of reported contracting, by street lighting Residential Street Maintenance Provider and tree trimming (but among municipalities with a At the end of 1985, St. Louis County govern- population greater than 5,000 only a few cases are ment was responsible for maintaining 1,142.46 miles found). The same group of municipalities reported of nonarterial county streets and roads. Its estimated no intermunicipal contracts for street services and maintenance expenditure on this system in 1985 was only a few joint service agreements. over $14 million.13 This makes the county govern- Municipal expenditures for streets or public ment by far the largest residential street mainte- works are reported annually to the state auditor. nance provider in the county. Each of the largest The average municipal expenditure for street serv- municipal providers spent in the neighborhood of $2 ices in 1985 (or the most recent year reported)@was million on roughly 100 miles of streets, or less. about $104 per household for the 61 municipalities County government was not, however, the highest in St. Louis County that reported street expenditures level provider in terms of expenditures per mile. As separately from public works.10 The average expen- an approximation, county government spent an esti- diture for the 18 municipalities that reported all pub- mated $13,000 per mile to maintain the county lic works was about $175 per household. The street (nonarterial) system in 1985.14 This compares to an expenditures figure is virtually equal to the average average of $10,694 per mile for 30 municipalities, annual subdivision assessment, noted above, of $103 over a broad size range, for which data were avail- per household.11 able. Considering only the seven municipalities with Analyzing data available for 34 municipalities, a population of 10,000 or more (for which data were those that contract out for street repair spend signifi- available), the average municipal expenditure was cantly less per household on street services than $18,424 per mile, ranging from a low of $7,985 to a those that maintain their own street departments, high of $30,604. Clearly, a number of municipalities controlling for the median household income in spent more per mile than did the county government each municipality (which is positive and significant to maintain residential streets, and a number spent in its relation to street spending per household for less. Expenditures may be a weak indicator of serv- this group). In other words, wealthier communities ice quality, however, and it is not known from this tend to spend more, but controlling for wealth, con- research what value in service county or municipal tracting communities spend less per household (though this finding should be treated with caution residents receive for their tax dollars. since data on contracting is not available for all mu- The county Department of Highways and Traffic nicipalities in the county). Furthermore, no signifi- produces all maintenance services through its divi- cant relationship exists between the population size sion of maintenance, except for some resurfacing of a municipality and street expenditure per house- projects and other projects classified as capital im- hold in a larger group of 61 municipalities for which provements and contracted out. The services pro- comparable data are available.12 Larger communi- duced by the maintenance division include preven- ties do not tend to spend more. These two findings tive maintenance (selective seal coating and are mutually supportive. Overall, size makes no dif- undersealing), snow and ice control, tree planting, ference in the level of spending per household, but bridge and culvert repairs, and storm sewer repairs. in the smaller group for which data are available, The division of traffic produces pavement striping, contracting (exclusively by small municipalities) and erects and maintains signs and traffic signals.15 does make a difference. The lack of a relationship between population size and spending per household RECONSTRUCTION suggests no apparent economies of scale in street The major problem that may face a street pro- service production. The absence of size economies viding subdivision is the need for reconstruction. can perhaps be explained by the negative relation- Sometimes the need for substantial reconstruction or ship between contracting and spending per house- replacement is an occasion for improving the streets hold: small municipalities (and subdivisions) are to a municipal or county standard and transferring able to take advantage of economies of scale in pro- jurisdiction to the overlying municipality or to the duction by contracting out. county government. A number of subdivisions in It is still possible, however, that the inability to University City have elected to do just that over the distinguish among street service conditions, such as years, at the invitation of the municipal govern- the original capital investment and traffic volume, in ment. 16 Nevertheless, many of the 53 subdivisions stud- to transfer any street to its jurisdiction as part of the ied are willing and able to engage in substantial capi- county arterial system. In so doing, the county not tal reinvestment. Fifty-six percent of the sample ex- only takes on the responsibility of maintenance but perienced some sort of extraordinary expenditure also assumes full control over traffic regulation (e.g., during the past five years, ranging from $24 to more speed limits) and the placement of traffic signals, as than $2,000 per household. Thirty-nine percent well as design. This capability on the part of county have done major reconstruction during the same pe- government potentially constrains traffic regulation riod, and 37 percent anticipate future expense of by municipalities that would seek to restrict traffic this sort. The average reinvestment-project cost, for flow. By removing the ability of municipalities to es- those subdivisions with reinvestment projects, was tablish speed limits on certain high-traffic streets, just over $650 per household. None of the subdivi- the county can also limit the use of speed traps, al- sions surveyed, however, report any indebtedness. though municipal police departments retain traffic Reinvestment seems to be cyclical. If the subdi- enforcement powers over the arterial system within visions are sorted by age and cut into four catego- their jurisdiction and can maintain the level of en- ries, a definite pattern emerges. For those subdivi- forcement they choose while keeping the revenue sions 20 years old or less, only 20 percent (1 out of collected from fines for the municipal treasury. 5) reported reinvestment expenses. For those between 20 and 40 years old, 82 percent (14 out of Residential Streets 17) reported reinvestment. Yet for those between 40 One of the major services provided by subdivi- and 60 years old, only 24 percent (4 out of 17) re- sions that control their own streets is access restric- ported any reinvestment. In subdivisions more than tion. Throughout much of University City and 60 years old, however, reinvestment activity is again Clayton, streets that would otherwise provide access higher, with 62 percent (5 out of 8) reporting such to subdivisions from city streets are chained off or an expense. barricaded. Residential streets within these subdivi- Relatively little interest was found among the 53 sions cannot conveniently be used for through traffic subdivisions studied for transferring their street juris- and become restricted to local access use. While not diction to the overlying municipality. A little more physically denying entrance to vehicle users, the ef- than two-thirds of those responding either indicated fect is to protect neighborhood streets from potential no interest in such a change or expressed opposition congestion and high volume traffic. Subdivisions also to it. Those recently experiencing higher capital re- may place speed bumps on their streets. Some also investment costs per household, however, appear restrict parking. Small municipalities often provide somewhat more inclined to want to shed jurisdiction, much the same sort of access restriction.18 or at least entertain the possibility.17 The ability of subdivisions to determine street Reconstruction can also pose special problems regulations is, from the perspective of residents, one for municipalities, frequently requiring bond issues of their main advantages. An analysis of the 53 sub- to be approved at the polls. Voters in the City of divisions sampled in this study indicates that subdivi- Webster Groves (population 23,097), for example, sions restricting access, or controlling traffic in other recently gave three-fourths approval to a $15 million ways such as using speed bumps, are more likely to bond issue to improve its 103 miles of streets. Pas- express a strong preference for retaining ownership sage of the bond issue was a major community un- and control of their streets. This capability is appar- dertaking enlisting both municipal officials and citi- ently of sufficient value to subdivision residents in zen volunteers. University City and Clayton that they willingly For county government, on the other hand, re- choose to pay the added cost of a subdivision assess- construction is a routine responsibility. The division ment.19 of construction supervised 28 completed projects in Oscar Newman, an architect and well-known 1985, totaling approximately $12.3 million. All were advocate of "defensible space" as an approach to contracted out. Reconstruction projects are funded crime control in residential areas, studied street pro- from countywide bond issues and the federal aid ur- viding subdivisions in University City and nearby in ban highway program. St. Louis City, comparing "private streets" to "public streets." He found that private streets, with re- REGULATION OF USE stricted access, had lower crime rates, higher perceptions of security on the part of residents, higher Arterial Streets property values for similar housing, higher assessed The legal power of the county government to act valuations, and higher rates of home ownership. as a provider of arterial streets is preemptive in na- Three factors may contribute to these results: (1) ture. The county council is authorized by state law street closure, (2) the existence of an association of which all residents are members, and (3) deed cove- access streets to local commerce and industry de- nants that restrict the conversion of property to fines a somewhat larger community of interest. multifamily use.20 The research was unable to sort Street connections to other parts of a metropolitan out the effect of one factor from another. The City area and arterial highway connections to other cen- of St. Louis has recently instituted an extensive ters of population define still larger communities of street closure program for the explicit purpose of interest. Each "box" can be associated with a par- controlling crime and promoting cooperation among tially distinctive set of values; yet the boxes compose neighbors. This program will offer an opportunity to an interconnected set and therefore impose recipro- study the effect of access restriction without formal cal constraints-limits as to how far certain values association and community control through deed can be maximized without infringing unacceptably covenants. Municipal programs and policies, how- on others. ever, may come and go. When subdivisions own Still further, the system is not static, but growing their streets, access control is an attribute of private and changing. New problems may arise that suggest property that cannot be taken for public use without new intermediate boxes. The general problem of just compensation. A municipality seeking to compel street organization is not simply how to govern that public access would be required to purchase an which exists, but how to shape future growth and easement from the subdivision. Under subdivision development. It is useful to summarize how the ownership, access restriction is a private prerogative overlapping responsibilities for street provision are that cannot be taken away through public regulation. shared among the various provision units in St. Yet all municipal subdivisions do not engage in Louis County. Table 6.4 contains such a summary. access control. Comparing across the four munici- Coordination is the dominant sharing mode only palities studied for this report, University City and for arterial street development. The responsibilities Clayton-the more urbanized areas-report the high- of state and county in this service area clearly, and est rate of access restriction, 75 percent of the 12 necessarily, overlap. If the state is to provide for ar- subdivisions in the sub-sample. By comparison, 35 terial highways to connect population centers and percent of the subdivisions in Ladue report access the county is to provide for arterial streets within a restriction, and only 11 percent in Olivette. Officials population center, then development decisions in the City of Creve Coeur report that only a few of within a county are inevitably going to involve the its street-providing subdivisions restrict traffic on linkage between these two systems. Still, if agree- their streets. Of course, traffic restriction can also be ment cannot be reached, decisions can be made achieved by means of the street pattern, without any separately. It is the high level of functional inter- need for closure. Most of the street-providing subdi- dependency between the two road systems that visions in the City of Town and Country, for exam- makes coordination both necessary and mutually ad- ple, obtain restricted access through physical design vantageous. and layout that avoids the grid pattern common in A somewhat different sort of coordination oc- earlier suburban developments. curs between private developers and county government in relation to traffic generation assessment BASIC PAlTERNS OF STREET PROVISION fees. The regulatory authority of the county is used No provision unit or single type of provision unit to require developers to make a contribution to in St. Louis County has a monopoly over any street street (and other infrastructure) developments. In facility type or street service component. In these contrast to the basically complementary relationship terms, responsibility is broadly shared among a vari- between state and county, the relationship between ety of jurisdictions. At the same time, with only mi- county government and developers is somewhat ad- nor exceptions, no single stretch of pavement is the versarial. responsibility of more than one unit at a time. Over- Residential street development, by contrast, in- lapping responsibility occurs in relation to street fa- volves less direct coordination and is instead primar- cility types and service component types, not par- ily a regulatory relationship between either the ticular service components on particular streets. county or a municipality and private developers. In- Street provision, like many public services, instead of coordinated decisionmaking, decisions are volves a diverse set of problems perhaps best viewed sequenced, with the relevant public jurisdiction es- as "Chinese boxes."21 Local access streets and their tablishing a set of rules and private developers mak- distinctive values-ease of access coupled with peace ing decisions within those legal constraints. Some co- and quiet-are associated with a relatively small- ordination must also occur in order for developers to scale community of interest, multiplied many times understand the applicable rules and gain approval over wherever people live apart from commercial for their proposals from county authorities. The and industrial uses of land. The connection of local linkage between residential and arterial street sys- Table 6.4 How Street Provision Responsibilities Are Shared and Distributed among Provision Units Service Facility Provision Units Sharing Mode

Development Arterial StatelCounty CoordinationIAiternation CountyIDeveloper Fee Assessment Residential CountyIDeveloper Regulation Municipality/Developer Regulation Malntenance Arterial StatelCounty Alternation Residential Municipality/Subdivision Alternation CountyISubdivision Alternation Reconstruction Arterial StatelCounty Residential MunicipaIityISubdivision Use Regulation Arterial Alternation Residential Alternation Alternation tems is also handled mainly through regulation-by of residential streets, is based in some part on street means of a county or municipal subdivision code. design. The county, and larger municipalities, agree While development responsibility is associated to accept responsibility for maintenance if, and only with high levels of coordination (in one form or an- if, certain construction standards are met. If a subdi- other), maintenance responsibility is characterized vision chooses to improve its streets to the accepted mainly by alternation-dividing responsibility street standard of the overlapping jurisdiction, it may also by street on the basis of legal jurisdiction. Alterna- choose to transfer all subsequent responsibility. tion characterizes both the division between arterial Use regulation, like maintenance, is mainly a and residential street responsibility and, within each matter of alternation. Subdivisions are sometimes street type, between relevant provision units. While characterized by municipal officials as "kingdoms development decisions bearing on location and de- unto themselves" on the subject of traffic and park- sign characteristics tend to be highly interdependent ing control. County and municipal police do patrol between street jurisdictions, maintenance decisions subdivision streets, but mainly for the purpose of are much less so. Maintenance schedules need not more general law enforcement. be coordinated in the way development plans are. In Notably absent from the modes of sharing re- the few cases of overlapping service provision on the sponsibility for streets and street services is extensive same streets-those few municipalities that provide duplication. Duplicate provision of the same service selected right-of-way services to subdivisions free of for the same street does not tend to occur. Duplica- charge-alternation still occurs, only by service in- tion in a broader sense-duplication of effort-may stead of facility. Municipalities and subdivisions do occur, but it is difficult to show that it does. When not tend to duplicate snow removal and surface re- two adjacent subdivisions or municipalities go pair, for example, but alternate between these serv- through simultaneous processes of decisionmaking ices. The coordination that occurs in residential with respect to maintenance services, is this "dupli- streets (i.e., contracting between subdivision and cation of effort?" What if preferences are different? municipality for selected street services) is a rela- What if one finds a better contractor than the other? tionship between the subdivision as a provider and What if street conditions differ between them? If the the municipality as a producer. two jurisdictions share information, some degree of Interjurisdictional relationships with respect to coordination is occurring. reconstruction are somewhat more complicated than The arrangement of street provision units in St. in the case of maintenance. This is because recon- Louis County follows closely the image of Chinese struction may provide an occasion for transfer of re- boxes. From this perspective, the structure of street sponsibility from one provision unit to another. The provision problems accounts for much of the pattern division of maintenance responsibility between of street organization. The only major anomaly in county or municipality and subdivision, in the case this explanation is the extent of county government responsibility for residential street provision in the Production arrangements in street services make unincorporated area. The absence of municipal or- much less frequent use of intergovernmental con- ganization means that a large and heterogeneous tracting and joint service production than in police, provision unit-county government-is responsible fire, and education. Small municipalities, and to a for tending to a large number of small-scale con- somewhat lesser extent subdivisions, seem to prefer cerns, the streets in particular subdivisions. A partial contracting with private producers to contracting explanation of this anomaly is found in the annexa- with municipalities. To some extent, the larger mu- tion rules that protected the unincorporated area nicipalities may discourage contractual relationships, from municipal expansion (see Chapter Three). in part because the more competitive nature of street service production, as compared to police and ARRANGING FOR PRODUCTION fire services, may lead to greater variation in pro- On the production side, the most interesting duction levels as purchasers, responding to price question concerns the choice by provision units competition, shift their business among vendors. either to organize their own production unit in- house or to enter into contractual relationships with CONCLUSION independent producers, whether private or public. The highly fragmented and overlapping jurisdic- The choice appears to depend substantially on po- tional arrangements for street provision in St. Louis tential economies of scale in production. The very County seem to have emerged largely for develop- smallest units-whether a subdivision or municipal- ment reasons and seem to be sustained largely for ity-contract out for virtually all street-related serv- reasons of street regulation and traffic control. The ices. Larger municipalities organize in-house pro- development factor can be seen in both the large duction units for certain routine services-those for role played by county government in arterial high- which equipment can be kept in continuous use- way planning and the significant reliance on subdivi- and contract out for occasional services, such as ma- sions to maintain residential streets. That subdivi- jor repairs, resurfacing, and reconstruction. County sions retain control of their streets even after and state agencies contract out only for particular municipal incorporation seems to be a function of projects-mainly construction-that make both labor traffic control-the ability to restrict vehicular access and equipment demands beyond their regular capac- by closing streets to traffic. New municipal incorpo- ity. Much the same pattern prevails for the planning ration efforts are also motivated in part by both traf- component of street development. fic control and highway development considerations. The frequency of private contracting for street Such a diverse set of jurisdictions enables the citi- services in St. Louis County is explained almost en- zens of St. Louis County to address widely varying tirely by the size of provision units. Heavy use of interests simultaneously. County government pro- private producers is found only in subdivisions and vides for arterial traffic flow interests, while small small municipalities (2,000 people or less). The municipalities and subdivisions tend to the interests emergence of a sizable private sector in the produc- of neighborhood access and residential amenities. tion of residential street services can be traced to the The level of service provision-measured either by existence of small units of provision, including both expenditure per household or expenditure per subdivisions and small municipalities. Despite the mile-appears to vary widely among municipalities, availability of private producers, however, larger subdivisions, and the country road (nonarterial) sys- municipalities make relatively little use of private tem. Small units-the large number of street-provid- contracting, even on a selective basis (service by ing subdivisions and municipalities with populations service). Aside from major repairs and resurfacing of less than 2,000 people-tend to function as pure projects, only street lighting appears to be con- provision units in street services, contracting pri- tracted out by the larger cities with any notable fre- vately to obtain production. Service contracting quency, followed to an even lesser extent by tree makes these small units economically feasible, in- trimming. creasing the degree to which small scale, neighbor- One relatively new forum has emerged with an hood interests can be represented in street provi- interest in promoting greater cooperative ventures sion. among the area's streets and public works depart- ENDNOTE- ments. The St. Louis Metro branch of the American ' or 1985, three nonproviding municipalities were found Public Works Association has organized a joint pur- (based on street expenditures of $0 reported to the state chase of road salt for 20 communities (including one auditor): Champ (pop. 28), Country Life Acres (pop. 77), in Illinois). Equipment sharing possibilities are also and Huntleigh (pop. 428). A fourth municipality, MacK- enzie Village, reported $0 street expenses in 1983, its most being explored. Some equipment sharing is reported recent report. The absence of street expenditure in a single among municipalities in the Normandy Council area. year by municipalities as small as these should not necessarily be construed as evidence of nonprovision over a or just above the mean in the distribution of municipalities longer period of time. by street expenditure per household. In other words, the 2~eeRobert Kanigel, "Improving City Streets to Death," City largest municipalities tend to spend about the same-or a 6 (Winter 1972), pp. 45-47. little more-per household than the average-spending municipality, but a number of smaller municipalities spend %ource: St. Louis County Department of Highways and considerably more per household than do the largest mu- Traffic, 1985 Annual Report. nicipalities. The data include the 61 municipalities that re- 4~ssessmentsper household were not available for the reported 1983-86 street expenditures to the state auditor sepa- mainder of the sample of 53 subdivisions. rately from public works and that report expenditures greater than zero. 5~hedifferences among the three municipalities are statistically significant in an analysis of variance. 13~ource: St. Louis County Department of Highways and Traffic, 1985 Annual Report. The Annual Report lists ex- =The lack of statistical significance in an analysis of variance penditures on the county road system totaling $13.9 million indicates that differences among subdivisions within mu- (p. A-7). This figure, however, does not include $2.37 mil- nicipalities are too great compared to differences across mu- lion in revenue sharing and transportation highway fund nicipalities to conclude that the municipal location of a sub- supplements (used for snow and ice removal), which were division makes a difference in the state of repair. not allocated in the report between the county road system 711 should also be noted that subdivision assessments, unlike and the arterial system. The proportion of operating ex- local property taxes, are not deductible for federal income penses allocated to the county road system exclusive of tax purposes. This increases the relative cost of provision by these unallocated funds was 57 percent. If the unallocated subdivisions as compared to provision by local governments funds were allocated in the same proportion, expenses for and is a consideration that applies to all subdivisions wher- the county road system would total $15:25 million. ever located. 14Using an estimated figure of $15.25 million in total expen- ~ACIRmailed or delivered questionnaires dealing with street ditures (see note 13) as the base, the expenditure per mile is services to all St. Louis County municipalities in the sum- $13,353. Using a figure of $14 million as the base, the ex- mer of 1986, and received 45 responses-a return rate of 50 penditure per mile is $12.259. The true figure probably lies percent. somewhere between these two estimates. As a rough ap- Q~omediscrepancy exists in the year of reporting for munici- proximation, $13,000 per mile does not seem unreason- pal street or public works expenditures, but the discrepancy able. makes little difference in the computation of mean scores. l5source: St. Louis County Department of Highways and Of the 61 municipalities reporting non-zero expenditures traffic, 1985 Annual Report. separately from public works, 46 reported in 1985. Five reported in 1983, eight in 1984, and two in 1986. If computed l6County government also stands ready to assume responsi- for only the 46 municipalities reporting in 1985, the mean is bility for private streets if improved to standard. One ar- $109. Of the 18 municipalities reporting public works ex- rangement used for this purpose in the unincorporated penditures, 14 reported in 1985, while 3 reported in 1984 county is to create a special road district for the purpose of and 1 in 1986. If computed only for the 14 municipalities levying a property tax to finance street improvements as a reporting in 1985, the mean is $177. means of transferring maintenance responsibility to the 10Excluded from this calculation are four very small munici- county. palities that reported expenditures of $0.00. 17subdivisions that engage in access restriction or traffic con- l1This comparison should be viewed with some caution. Con- trol are more inclined to want to retain jurisdiction. See the siderable variation exists among both subdivisions and mu- discussion of use regulation below. nicipalities in precisely what services are provided. Some 18pasadena Hills, for example, has chained off its access subdivisions also provide nonstreet services. Although some streets along a traffic artery on one of its borders. subdivision budgets would resemble a municipal public works budget, others are strictly a street services budget. lQ,One other factor found to have a significant positive rela- Another reason for caution is the comparison of expendi- tionship to expressed willingness to consider transfer of tures in the case of municipalities with revenues in the case street ownership to the overlying municipality is the amount of subdivisions. This comparison is, nevertheless, the best of recent expense per household on capital improvements. available. Subdivisions have no other source of revenue. Municipal street revenues are only partially earmarked for 2O0scar Newman, Community of Interest (Garden City, New streets. The main source of potential distortion is the possi- York: Anchor PressJDoubleday, 1980), pp. 124-156. bility that subdivisions do not spend all the revenue they col- 21Robert A. Dahl uses the "Chinese Box" imagery to charac- lect. terize the relationships among overlapping territorial juris- 12~hestatement should be qualified as no significant linear dictions in After the Revolution: Authority in a Good Soci- relationship. The largest municipalities do tend to cluster at ety (New Haven: Yale University Press. 1970). Chapter Seven Elementary and Secondary Education

Local school districts are not, however, the only provision units relevant to public education in Mis- This chapter focuses on relationships among the souri. The state constitution assigns responsibility for 24 elementary and secondary school districts in St. provision of public education to the state General Louis City and County and the emergence of multi- Assembly [Article IX, Sec. l(a)]. This presents a jurisdictional arrangements both for providing and sharp contrast to provision arrangements for police, producing education services. These functional ar- fire, and street services. The authority and responsi- rangements include: (1) joint enterprises organized bility of local school districts derive exclusively from by separate school districts to produce auxiliary serv- state statutes, and the state legislature is directly ices; (2) coordination between regular school dis- charged by the Missouri constitution with a funda- tricts and a separate countywide district for special mental responsibility to "establish and maintain free education, and (3) cooperative efforts for public public schools." State regulation of local authorities school desegregation among the city and county is greater in public education than in any other local school districts. The jurisdictional fragmentation of function. Arranging for the governance of school school districts in St. Louis County has not fore- districts-the enactment of rules governing provi- closed the development of joint functional arrange- sion-can be viewed as a state government function. ments through coordination and cooperation. How- The state also participates directly in the provision of ever, fiscal problems continue to frustrate efficient education through state financing, providing roughly and equitable public education in some parts of the one-fourth of total funding for public schools in St. county. Louis County, and one-half in St. Louis City. School Dlstrlct Development In St. Louts County PROVISION ARRANGEMENTS During the last half-century, public school dis- In St. Louis City and County, public elementary tricts across the nation have undergone extensive and secondary education is provided through 25 consolidation. In 1942 there were more than autonomous school districts. (Figure 7.1 displays a 108,000 public school districts in the United States; map of the school districts.) Twenty-four of the dis- in 1987 there were approximately 14,741. Although tricts are located in St. Louis County. One is a school district consolidations have taken place for countywide Special School District organized to pro- the most part in rural areas, a good deal of this con- vide special education services and vocational-tech- solidation has occurred in metropolitan areas. Be- nical education. For regular elementary and secon- tween 1905 and 1985, the number of public school dary education, the county is divided into 23 school districts in St. Louis County fell from 88 to 24. districts ranging in student population from less than Public education in St. Louis County is mark- a thousand to nearly 20,000. St. Louis City, by con- edly less fragmented than police, fire, and street trast, is served by a single school district with more services. In fact, school districts have developed in a than 50,000 students. pattern that is the reverse of development patterns Figure 7.1 School Districts in St. Louis County

SOURCE: St. Louis County, Missouri, Fact Book- 1986. St. Louis County Department of Planning. for municipal services generally. As the population pal consolidation, which requires voter initiative and of St. Louis County increased over the past century, approval in each municipality. In some part, there- the number of local units that provide police, fire, or fore, the less fragmented pattern of school district street services grew as the number of school districts organization in St. Louis County, as compared to declined. This different pattern of development municipal organization, may be explained by the dif- might be explained as simply a local manifestation of ferences in constitutional rules between the two the larger, national pattern of public school consoli- types of local organization. dation that occurred over the last half-century. What the national trend may best explain, however, The Special School District of St. Louis County is not the pattern of provision that has come to exist in St. Louis County, but rather the pattern of gov- The Special School District to provide for edu- ernance created by the state legislature-the rules cation of the handicapped and for vocational-techni- that govern the creation and dissolution of school cal education encompasses all of St. Louis County. districts. Each elementary and secondary school district in Unlike the rules for the creation of municipali- Missouri is required by state law (Section 162.670, ties in Missouri, rules for the creation of school dis- Missouri Revised Statutes) to make provision for tricts have placed substantial authority in the hands special educational services to any handicapped or of county officials (originally a county superinten- severely handicapped child residing in the district. dent of schools) to initiate procedures for school dis- In St. Louis County, however, these services are trict reorganization. The 23 regular school districts provided in large part through the Special School in St. Louis County are distinguished by state stat- District, approved by the voters in 1957. In that utes as either common, reorganized, or consolidated same year the special district began operating pro- districts. Common districts are those originally estab- grams for educable mentally retarded children in the lished to erect and maintain common schools county. In 1965, St. Louis County voters approved throughout the state. To reorganize two or more dis- the addition of vocational-technical education serv- tricts into a single larger district, the county govern- ices to the special district's responsibilities. ing body (the county council in St. Louis County) The use of a countywide special district to pro- may submit a reorganization plan to the state board vide education for handicapped students is consis- of education. (Originally this power was vested in a tent with the nature of the service. Special education county board of education, since abolished.) If the for handicapped children is much more expensive state board approves a proposed reorganization than regular education, and it is provided for only a plan, the plan is then submitted to voters in the pro- small minority of students. The Special School Dis- posed new district. A simple majority of all the votes trict spent $16,737 per student in 1984-85 (this fig- cast within the new set of boundaries, not necessar- ure does not distinguish handicapped from vo-tech ily concurrent majorities in each of the existing dis- students), more than twice the expenditure level of tricts, is sufficient for approval of the reorganization. the highest spending regular district in the county Common school districts may also be merged and more than four times the mean level of spend- through consolidation, a different procedure than ing per student averaged over all students in the 23 reorganization. Consolidation proposals must be in- regular districts. Efforts to increase the quality of itiated by citizens-at least 10 percent of the total education for the handicapped by any single school number of eligible voters in each of the districts to district in the county would significantly increase av- be consolidated-and require concurrent majorities erage costs per pupil, and increase tax bills within of the voters in each district for approval. the district at the same time. Moreover, parents of Seven of the 23 districts in St. Louis County handicapped children would be attracted to a school were created using the reorganization procedure. district that excelled in these services, thus increas- There is only one consolidated district-Parkway- ing still further the average cost of education per pu- formed in 1954, largely in an unincorporated area, pil in the district. Districts that sought to improve but now the county's largest district in number of education for the handicapped would in effect be students. The remaining 15 districts are common penalized for their efforts. Moving to a larger scale districts. Most school district mergers, therefore, of provision thus increases the feasibility of improv- took place at the initiative of the county board of ing education for the handicapped. To some extent, education rather than local citizens. Approval was the costs of special education in St. Louis County not required from voters in each existing district be- may still be affected by families that, attracted by fore their district could be merged with others.' high quality services, relocate from other parts of the The rules for school district reorganization are St. Louis metropolitan area.2 The broader the juris- significantly different from those that govern munici- diction is geographically, however, the more difficult interjurisdictional mobility becomes, and the less se- that provide street services are legally private, only vere the problem should be.3 education (of the four service areas studied) has an Vocational-technical education also is more ex- independent, fully private sector-a market that pensive than academic programs, although the bene- links individual households as consumers with pri- ficiaries compose a larger segment of the student vate schools as producers. population than in the case of the handicapped. The Presently, there are more than 30 private, non- case for areawide provision is accordingly somewhat parochial elementary and secondary schools in St. less compelling, and, not surprisingly, some of the Louis City and County, 22 of which participate in regular districts have chosen to provide their own the Education Conference, an association of private vocational-technical program. This type of training schools. Numerous parochial schools are also oper- may also exhibit greater economies of scale in pro- ating in the area, among which are over 30 Lutheran elementary and high schools, consistently enrolling duction than academic programs. over 4,000 students each year during the past dec- The St. Louis City School District ade. Both nonparochial and Lutheran enrollments are dwarfed, however, by enrollments in Roman The City of St. Louis is organized as a metro- Catholic schools in the area. The St. Louis Archdio- politan school district, legally defined as a school cese operates the largest school district in the entire district in a city that is not also within a county. Be- metropolitan area (also in the state), serving stu- cause St. Louis City is the only city in the state that dents in six counties in addition to St. Louis City is not also within a county, the St. Louis school dis- and County. During 1984-85, the archdiocese pro- trict forms a legal class of one and is the only metro- vided services through 123 elementary and 34 sec- politan school district in the state. Since it is not a ondary schools for approximately 45,000 elementary metropolitan district in a functional sense, it will be and 18,000 secondary students. referred to here simply as the city school district. Private school enrollments are substantial, but In 1968 the board of education in St. Louis City vary across the 23 county school districts. Table 7.1 suggested the merger of all school districts in the city - and the county. This proposal was received unfavor- Table 7.1 ably in both the city and county. Not surprisingly, Percent of School Age Children in the proposal was killed soon after its introduction in Private Schools, by District, the state legislature. Although several of the school St. Louis County- 1980 districts in St. Louis County have merged since Percent Percent 1968, no additional proposals for a unified city- Districts K-8 Private 9-12 Private county school system have been advanced. The St. Louis City district much more closely Merged approximates the traditional reform ideal of large- Hazelwood scale, comprehensive public organization than any Ferguson-Florrisant of the county school districts. Besides providing Pattonville regular elementary and secondary education, the St. Rockwood . Kirkwood Louis City school district has organized an internal Lindbergh special education division, although its scale of op- Mehlville eration is not as large as the county Special School Parkway District. The scale of operation for regular elementary and secondary education in St. Louis City, how- Common ever, greatly exceeds that for any single district in Amon the county. Student population in the city district, Bayless just over 50,000 in 1984-85, was more than twice as Brentwood large as in Parkway, the largest school district in the Clayton county. One would have to combine the student Hancock Place populations in each of the four most populous Jennings county districts to match the population in the city Ladue district. Maplewood Normandy Private Alternatives to Public Provision Ritenour Riverview Gardens Education also differs from police and fire and, University City to a somewhat lesser extent, from streets, in the ex- Valley Park tent of the opportunities available for private provi- Webster Groves sion. Although production of street services by pri- Wellston vate contractors is common, and the subdivisions displays the percentage of all school-age children county, members are elected at large, despite the within each district who were enrolled in private or greater size and heterogeneity of the city district. parochial schools in 1980. Private enrollment in elementary schools, averaged across the districts, was PATTERNS OF PROVISION about 25 percent of all students, while private high Provision refers to basic decisions of what serv- school enrollment was slightly lower, at about 21 ice to provide, how much, and how to pay for it. percent. The range is from 43 percent of all elemen- Included are arranging for production (but not pro- tary students and 27 percent of all secondary school ducing), monitoring service quality, and holding pro- students in the Bayless district to no elementary ducers accountable for their performance. Provision school students and only 2 percent of high school of public education in this broad sense is shared be- students in the Wellston district. tween local school districts and the state government The existence of an independent private sector in Missouri. The basic unit of provision is, neverthe- has the potential to enhance responsiveness among less, the local school district. This is so not simply the districts forming the public ~ector.~State aid for- because most (but not all) local districts supply the mulas based on enrollment potentially place public major portion of school funding but also because schools in competition with private schools for stu- they have a primary role in all the basic provision dents.5 Whatever the degree of competition, the pri- activities. Local electorates approve or disapprove vate sector in education generates alternatives that property tax rates. Locally elected school boards allow greater diversity of education services than the employ a superintendent whose job it is to link pro- 25 public school districts in both city and county can vision with production. Accountability for school provide alone. Since private schools are widely dis- performance is obtained primarily within each dis- persed across the area, citizens need not necessarily trict. For most direct services to students, produc- move their residence in order to exit from the public tion is also organized by the local school district. In- schools. Of course, opting out of public school atten- direct services in St. Louis County, however, tend to dance does not entail opting out of tax support for be organized through joint production units that public schools. Yet, for those willing and able to pay, serve a number of school districts. the private sector in education increases the range of As provision units, the 23 elementary and sec- choices available to St. Louis residents. ondary school districts in St. Louis County are quite different, varying in a number of measurable ways. This analysis will examine variation in the following School District Organization respects: Members of the board of education in the 23 w District size, measured by number of stu- six-director districts in St. Louis County are elected dents in average daily attendance; at large, two per year, for staggered terms of three w School size, averaged by district; years. School board elections are conducted simulta- w Student social background, by district; neously with municipal elections. Each board ap- Level of district financial support; points a full-time professional superintendent who is in the broadest sense chief executive/administrative w Total and current expenditures per student; officer for the district, though specific duties are de- Student-educator ratios for elementary and termined by the terms of a contract negotiated with secondary schools; the board. Superintendent contracts may differ w Percentage of district students attending pri- across districts, but in accordance with state law no vate schools; and superintendent may be appointed for a term exceeding three years, subject to possible reappointment. w Percentage of eighth-grade students passing School districts thus differ from St. Louis County standardized tests. municipalities in their greater and more exclusive re- The comparisons are among districts, not among liance on professional managers, as compared to the schools. This should be kept in mind because larger citizen mayors who function to varying degrees as districts will tend to have greater within-district vari- chief executive and sometimes chief administrative ation, while smaller districts show greater variation officer in municipalities. The board of education for from one district to another. That is, smaller districts the Special School District is also composed of six can be expected to be "more different" from one members, with one member elected from each of six another, while larger districts are "less different." regions of equal population. In St. Louis City, the Financial data are for 1984-85.6 District demo- board of education is composed of 12 members who graphics are taken from 1980 census figures, except serve six-year terms. Like the regular districts in the for the poverty measure, which is a 1979 figure. Variation in Local Provision Figure 7.2 plots each district on two dimensions: assessed valuation per student and levy rate. The There are several points of interest in examining solid line shows the predicted levy rate given as- the variation among St. Louis County school dis- sessed valuation per student. Districts above the line tricts. One major source of interest is the variation in are, in effect, digging deeper into their pockets than size. The total student body of the smallest district districts below the line. These differences would ap- was 582 in 1984-85, while the largest was 19,259. pear to reflect differences in voter preference, given The mean district size is a little more than 5,100 ability to pay. Some communities in St. Louis students. Only three districts have more than 10,000 County seemingly want to invest more of their re- students, while eight have fewer than 2,000. School sources in public education, others less. One should district consolidation-although it has reduced the point out, however, that individual families can ex- number of school districts in St. Louis County sub- press demand for education by choosing to live in a stantially over the years-has produced neither a sin- high property value district, as well as by voting for a gle large district nor a set of districts similar in size. higher levy rate in a lower property value district. Do the differences in district size cause some parts This makes it impossible to completely sort out will- of the county to be better served than others? ingness from ability to pay. Would a set of similar size districts, or perhaps even No school district in the county is required to be a single large district, serve the county better? completely self-sufficient because the state provides Answers to these questions depend on what is aid to local schools. Twenty out of 23 districts, how- meant by better served. Educators are apt to look ever, raised more than half of their total spending in mainly at educational quality and to argue that communities with better quality schools, by some set of 1984-85 from their own sources. Four of the dis- standards, are better served. Others might argue tricts were self-supporting at a level of more than 80 simply that those districts spending more per student percent. The smallest proportion of district funding better serve their students. On the other hand, there was 48 percent.' are questions of efficiency to be considered. Effi- The 23 school districts in St. Louis County differ ciency criteria suggest that education should be pro- dramatically in the demographic characteristics of vided at a level at which the relevant communities their resident populations, with important implica- are willing and able to pay for it; but are the relevant tions for the educational task each faces. These are communities school districts, the county, the state, extremely important variables because of the weight the nation, or all of these? Increasing marginal costs ordinarily attributed to the social background of stu- mean that, at some point, greater educational qual- dents as a factor related to student achievement. ity is not worth the price; but who should make this Three indicators of student background are used in judgment? Finally, there are questions of equity. this analysis: (1) rate of poverty in the district in Those communities least able to afford good educa- 1979 (percentage poor), (2) percentage of district tion may be those most in need of it in order to residents who are white (1980 census), and (3) me- provide an opportunity for their residents to improve dian household income (1980 census). The percent- their lives. The analysis in this chapter can do no age of poor ranges from a little more than 2 to 35.5. more than array some limited evidence and offer a The percentage of whites varies from 0 to 99.4, and few preliminary observations on these questions. median household income, from $9,043 to $38.874. The districts also vary in their economic demand In still other ways, the districts encompass quite for education as indicated by willingness and ability different communities. Parkway, the largest district to pay. The amount a district spends per student in number of students, serves mostly an unincor- from its own tax sources is a function of the total porated, but developed, upper-income area. Rock- assessed valuation of property in the district per stu- wood, the far western district in the county, and the dent and the property tax rate. The relationship be- largest in geographic area, is the most rural. Some of tween total assessed valuation per student and the the smaller districts-Brentwood, Clayton, University rate at which property is taxed in the 23 St. Louis City, and Wellston-have boundaries that are coter- County school districts is negative. The higher the minous, or nearly so, with municipal boundaries, but assessed valuation per student, the lower levy rates most of the school district boundaries do not corre- have to be in order to raise the same amount of spond closely to the boundaries of cities and villages. money, and the lower they tend to be. The relation- Wellston is nearly all black, while University City is ship is far from perfect, however, and variation in racially mixed. Ladue and Webster Groves are this relationship provides an indicator of variation in somewhat larger school districts that encompass the the willingness of district taxpayers to spend money cities of Ladue and Webster Groves, respectively, for public education. but that also extend to include a larger area contain- St. Louis County School Districts

1 Hazelwood 9 Affton 17 Normandy 2 Ferguson-Florrisant 10 Bayless 18 Ritenour 3 Pattonville 11 Brentwood 19 Riverview Gardens 4 Rockwood 12 Clayton 20 University City 5 Kirkwood 13 Hancock Place 21 Valley Park 6 Lindbergh 14 Jennings 22 Webster Groves 7 Mehlville 15 Ladue 23 Wellston 8 Parkway 16 Maplewood

Assessed Valuation Per Student ($) ing other small municipalities. Most of the small to the second lowest district on this indicator). mid-size, as well as some of the larger, districts serve Wellston also taxes itself at the highest rate in the a fairly homogeneous population, some (e.g., county. In total spending per student (excluding Ladue) more so than others (e.g., University City). debt service), Wellston ranks second from the bot- Normandy district, on the other hand, encompasses tom. The bottom district in terms of expenditure per 23 small municipalities, but is highly urban in char- student, Hancock Place, is 99.42 percent white, has acter, as well as having a very heterogeneous popula- a poverty rate of only 4.2 percent, and a median tion in terms of both race and income. household income of $16,353-fourth from the bot- Total expenditures per student from all sources tom. Its assessed valuation per student, however, is (excluding debt service) in 1984-85 varied among second from the bottom, and its tax rate, in sharp the 23 districts from a low of $2,544 to a high of contrast to Wellston, is third from the bottom. $7,005-a range of nearly $4,500. The mean is It is less clear that there is a single most highly $4,110 and the standard deviation, a measure of the advantaged district. In terms of assessed valuation dispersion of spending among districts, is $1,145. and total expenditure per student, the answer is The distribution of spending per pupil by school dis- Clayton, followed by Brentwood; but in terms of the trict is skewed positively-the distance from the demographic indicators, Parkway, Ladue, and mean to the highest funded district is much greater Lindbergh are the leaders. than the distance from the mean to the lowest The State Role In Provision funded district. If the distribution is shown by student, as in Figure 7.3, the positive skew is even The provision role performed by state govern- more pronounced, since most students are found in ment is one of supplementing local funding, in part larger, moderately funded, districts. The mean to maintain minimum funding levels (equalizing in spending per student, when averaged over students the sense of making school funding more equal), rather than districts, is less-$3,909. The standard and in part to support certain portions of the educa- deviation, considerably less, is $746, indicating sub- tion budget. The Missouri Constitution (Article IX, stantially less variation in spending levels among stu- Section 3(b)) demands that at least 25 percent of dents than among districts. state revenues be earmarked for support of public The smallest districts-those with fewer than schools. State monies for public education are dis- 2,000 students-include both the highest spending tributed to school districts through several different and lowest spending districts in the county. The funds, which can be divided into two types: (1) allo- greater variation among small districts is to be ex- cative funds distributed primarily on the basis of pected as a statistical property. There is some ten- daily attendance or student population,a and (2) dency for the smallest districts to be better funded redistributive funds intended to correct for fiscal dis- (indeed all of the extremely well funded districts are parities among local districts. The most important quite small). One of the motivations commonly at- state funds of the first type are the Missouri School tributed to citizens who seek to maintain a relatively District Trust Fund, the Fair Share Fund, and the small school district is to preserve a higher than av- Free Textbook Fund. The most important fund of erage level of school funding. The residents of small the second type, and, in fact, the most significant wealthy districts, however, cannot escape state taxa- source of state aid to local school districts, is the tion for the purpose of redistributing wealth to dis- Foundation Program. tricts that are less well off. Smaller districts are also Allocative Funds. The school district trust fund was more homogeneous with respect to population char- created when voters approved Proposition C in acteristics. Some are wealthy; others are middle in- 1982, imposing an additional 1-cent state sales tax come or poor. Some are nearly all white; others are earmarked for education. Use of these funds was mostly black. One district-Wellston-is virtually all tied to local property tax relief. When a district re- black and quite poor. ceives money from this fund, the revenues are reck- Table 7.2 ranks the school districts by median oned in the Missouri state budget not as state household income while also showing their scores on monies, but as local tax revenues obtained during other demographic and fiscal variables. Considering the current year. Each district accepting funds must the various factors that affect the provision of educa- use 50 percent for local property tax reduction, and tion, the most seriously disadvantaged district ap- 37.5 percent must be applied to teacher salaries. pears to be Wellston: 100 percent nonwhite, 34.5 The remaining 12.5 percent can be used for district percent poor, a median household income of $9,043 operating costs. (more than $5,000 less than the second lowest dis- The Fair Share Fund was also introduced in trict on this indicator), and a total assessed valuation 1982 when the General Assembly approved a 4-cent per student of $19,306 (more than $8,000 less than increase in the cigarette tax, raising it from 9 cents Figure 7.3 Variation in Per Student Spending, St Louis County Schools - 1984-85 Number of Students

1.00 -.75 -.50 -.25 Mean .25

Standard Deviation from Mean of $3,908.92 per Student to 13 cents per pack. Funds from the cigarette tax the state Foundation Program. This program is ex- were previously earmarked for education, but reve- plicitly redistributive, its central purpose being to nues from the additional 4-cent tax are specifically provide equity to public school students by reducing earmarked for teacher salaries and are distributed to fiscal disparities across school districts. school districts strictly on the basis of the districts' Some of the funds within the Foundation Pro- average daily attendance. The state also provides gram-those designated for student transportation revenues to school districts from a Free Textbook and to programs for gifted and exceptional stu- Fund that is generated through a foreign insurance dents-are allocative, distributed to the districts gen- tax. This tax is levied on all out-of-state insurance erally according to cost per student. The largest part companies that operate enterprises within Missouri. of the Foundation Program, however, is distributed Money froin the fund is distributed to districts annu- on the basis of a complex equalizing formula that ally and must be used to purchase textbooks ap- seeks to compensate for local factors that tend to proved by the state board of education. advantage some districts and disadvantage others. This formula takes into account criteria estimating Redistributive Funds. Even though state allocative each district's ability (as reflected by assessed valu- funds do not discriminate among districts on the ation) and willingness (as reflected by school-tax basis of wealth, the effect is at least mildly levy rates) to support public schools. redistributive. By collecting and distributing revenue The formula first determines a minimum guar- on a larger scale, state government achieves a de- antee of state financial aid to school districts. The gree of redistribution by reducing the effect of vari- amount of the guarantee is calculated by adding one ation in tax base on variation in spending per stu- factor, based roughly on average daily attendance, dent. The preponderance of state aid to school to another factor, based on the number of AFDC districts, however, comes through the various funds, (Aid to Families with Dependent Children) entitled drawn from state general revenues, that comprise and orphaned students in attendance. Deductions

Table 7.2 County School Districts Ranked by Median Household Income, with Fiscal and ~emographicCharacteristics Assessed Total Median Valuation Spending Household Per Per Number Income1 Percent Percent Student2 Student2 of Dlstrlct (in dollars) Poor1 White1 (in dollars) (in dollars) Students3 Wellston $9,043 34.49% .oo% $ 19,306.43 $2,634.77 1,043 Maplewood 14,520 7.97 85.52 81,487.46 4,565.98 1,268 Jennings 15,882 6.15 64.16 40,594.16 2,725.76 2,165 Hancock Place 16,353 6.76 99.42 27,660.31 2,544.43 1,530 Normandy 16,408 13.75 35.99 28,482.83 3,392.16 5,844 Brentwood 16,654 6.77 85.67 125,430.89 5,951.68 769 Valley Park 16,819 11.74 96.81 73,606.03 4,610.91 582 Ritenour 17,438 7.32 89.70 49,507.07 3,706.78 5,979 University City 17,941 10.36 55.26 46,253.22 3,639.92 4,682 Riverview Gardens 19,282 5.13 83.48 42,151.62 3,379.54 4,299 Affton 19,576 3.38 98.82 101,317.02 4,787.84 1,784 Ferguson-Florrisant 20,672 6.37 81.29 62,634.03 4,321.15 10,571 Bay less 21,024 4.04 99.33 58,364.74 3,127.10 1,233 Pattonville 22,068 3.44 96.51 90,674.40 4,789.67 5,844 Webster Groves 22,388 4.22 88.90 68,897.95 5,489.38 3,497 Clayton 23,036 6.47 94.1 1 193,075.74 7,005.94 1,644 Mehlville 23,618 2.94 98.40 56,578.22 3.488.09 8,870 Hazelwood 23,739 2.78 89.87 53,286.65 3,606.42 15,054 Kirkwood 24,813 3.59 92.80 90,616.93 3,920.62 4,043 Rockwood 25,277 3.96 97.79 58,628.19 3,445.20 9,857 Lindbergh 25,996 2.50 98.62 91,583.05 3,618.92 4,718 Parkway 31,197 2.15 96.02 74,643.37 3,832.26 19,259 Ladue 38,874 2.71 91.81 168.961.59 5.936.11 2.889 '1980 Census. 21984-85. =Students in average daily attendance. are then taken from each district's minimum guar- dicted with reasonable accuracy from a knowledge antee that are intended to differentiate respective of the district's assessed valuation per student. amounts of state aid to local districts on the basis of Those that fall above the line are being treated more two criteria: (1) differences in district wealth taxable favorably than would be predicted from assessed for public schools, or assessed valuations, and (2) valuation alone and those that fall below the line, differences in district tax efforts, or the rates at less favorably. which districts tax themselves to support schools. Fi- The least favored district by this measure is Jen- nally, a district entitlement is determined by adding nings, a small district adjacent to Wellston and St. the district's minimum guarantee (less deductions) Louis City. Its tax effort is relatively high-at $3.66 to the amount of guaranteed tax base, which is the per $100 assessed valuation-and the taxable wealth assessed valuation per pupil guaranteed each school per student is relatively low.12 The most favored dis- district by the state in computing Foundation aid. trict is Normandy, a racially and economically het- The terms "minimum guarantee" and " guaran- erogeneous district encompassing an urban-type teed tax base" are better for understanding the com- area. Another apparently well-treated district is Fer- pensatory nature of the Foundation Program than is guson-Florissant, a large middle-income district that the term "equalizing formula." The purpose of state includes most of the county's largest municipality aid to education in Missouri is essentially to furnish (Florissant) and makes a high tax effort relative to a minimum provision standard for all public school its assessed valuation. districts in the state. This is different from a goal of Overall, the Foundation Program seems to per- absolute equalization, which could be accomplished form fairly well in terms of its redistributive purpose, only by stripping school districts of all authority to as indicated by a strong negative correlation (-0.74) make key provision choices.9 between assessed valuation per student and Founda- Twenty of the 23 regular school districts in St. tion funding per student in each district. The stan- Louis County raised more than half of their total dard error of the estimate in this equation, $192.94, spending for 1984-85 from their own sources.lO is also fairly small, indicating a low variation among Four districts raised 80 percent or more, and nine, the districts in how they are treated. This conclu- more than 70 percent of the total funds spent. Three sion, it should be emphasized, relates to the way the districts raised less than half-Normandy, a hetero- program distributes aid and not to the total amount geneous district with urban characteristics, Hancock of aid distributed. Nor does this general conclusion Place, a small South County district adjacent to St. address how any particular district is treated. Louis City, and Wellston, a mostly black area between Normandy and St. Louis City. Wellston, how- Fiscal Disparity- How Big a Problem? ever, also levies by far the highest property tax rate Per pupil spending does not vary widely among in the county, at $4.13 per hundred, while Hancock the great majority of students in St. Louis County. Place is among the lowest taxing districts and Nor- Referring to Figure 7.3, 79.5 percent of the total mandy is just above the median. Local funds are county student population (in 13 districts) falls generated mainly from property tax revenues, in- within one standard deviation ($746) of the mean cluding both residential and commercial-industrial expenditure per student (averaged over students property, within each district. There is also, how- rather than districts, as discussed above). Only 6.4 ever, a countywide railroad and utilities tax ear- percent (7,563 students in four districts) lie more marked for schools, distributed on the basis of stu- than one standard deviation below the mean. Fiscal dent population. Like the state allocative funds it disparity does not adversely affect most county stu- resembles, the county tax (collected and distributed dents to any great extent. by county officials) is mildly redistributive." A small minority of the student population is, The distribution of Foundation funds per stu- nevertheless, a source of concern. Three of the four dent ranges from a low of $268 and $330 in Clayton districts whose students lie more than one standard and Brentwood, respectively, to over $1,000 in deviation below the mean are also among the bot- Riverview Gardens, University City, Wellston, and tom four districts in terms of median household in- Normandy. An indication of the responsiveness of come, and two of the four are among the four dis- the Foundation formula to differences in district tricts with the greatest percentage of nonwhite wealth can be gleaned from Figure 7.4, which plots population. Students in these districts are likely to assessed valuation per student against Foundation be in greatest need of remedial education, yet they funding per student in each district. The relationship attend schools in districts least likely to be able to is strongly negative. Districts that fall close to the afford the extra attention to individual students. regression line in the center of the figure are receiv- Greater variation exists among students in ing Foundation aid in amounts that could be pre- schools that are above the mean, where 14 percent Figure 7.4 Relation between Current Expenditures Per Student and Foundation Funding Per Student, by District, St. Louis County- 1984-85 Foundation Funding Per Student ($)

Assessed Valuation Per Student ($) I I St. Louis County School Districts 1 Hazelwood 9 Affton 17 Normandy 2 Ferguson-Florrisant 10 Bayless 18 Ritenour 3 Pattonville 11 Brentwood 19 Riverview Gardens 4 Rockwood 12 Clayton 20 University City 5 Kirkwood 13 Hancock Place 21 Valley Park 6 Lindbergh 14 Jennings 22 Webster Groves 7 Mehlville 15 Ladue 23 Wellston 8 Parkway 16 Maplewood fall outside one standard deviation and 7.5 percent alone would bring its students to within $302 of the are outside two standard deviations. Students in the hypothetical standard. Hancock Place, the most Clayton school district, the most highly funded dis- poorly funded district in terms of total spending per trict in the county, lie 4.15 standard deviations off student, was much better treated in the Foundation the mean.13 program and fell only $26 short of the amount of aid County, state, and federal funding reduces the predicted on the basis of assessed valuation. variation in spending per student among districts, Wellston, on the other hand, is relatively well while increasing the level of provision in each of the treated by the Foundation formula, though not as districts. These extra-district sources increase the well treated as one might expect.15 Bayless also cur- spending per student in the lowest expenditure dis- rently fares relatively well in the Foundation pro- trict by $1,425 and in the highest expenditure dis- gram. Thus the four districts that end up at the bot- trict by $851. This more than doubles the spending tom of an expenditure-per-student ranking are level of the lowest expenditure district. Yet the mean treated quite differently by the major state spending level (averaged over the 23 districts) in- redistributive program. To address the diverse situ- creases by $1,299-almost as much in absolute dol- ations of three such districts by adjusting the formula lars as the increase for the lowest expenditure dis- would be a complex undertaking and one that would trict, though less than a 50 percent increase likely produce similar unanticipated consequences proportionately. Most of this increase is from state for other districts. funding. Clearly, state policy reduces the variation in There is no approach to interjurisdictional redis- spending among districts below the mean. The distribution that can guarantee that each dollar of assistance between the mean level of spending and the tance will help both students who are the lowest lowest expenditure district decreases slightly in abso- funded and taxpayers who have the heaviest tax bur- lute dollars and substantially as a percentage of the dens. This leaves policymakers in the position of de- mean, from 60 percent to 38 percent.14 liberately following a mixed strategy: (1) targeting Still, redistribution efforts have left a small left- some assistance to those school districts with rela- hand tail in the distribution of spending per student. tively high tax burdens and (2) providing other assis- Virtually eliminating the left-hand tail of the distri- tance across the board so as to provide assistance in bution would be relatively inexpensive. To raise all some absolute amount to students in low-expendi- students to a level of district spending no more than turellow tax-burden districts. This is the basic strat- three-fourths of one standard deviation below the egy currently followed by the state of Missouri. 1984-85 mean would cost about $3.6 million, or It is also impossible to define a permanent stan- about $30.65 per student averaged over the entire dard in statistical terms, any more than all students county student population. Targeting funds on those can be made above average. The discussion above, districts where per-student spending is the lowest, in terms of the 1984-85 distribution, is illustrative however, entails certain difficulties. Such an ap- only. Fiscal standards ultimately ought to be defined proach would create an equity problem at the same in terms of performance (i.e., the amount of money time that it reduces disparities. Districts currently required to attain some level of performance consid- making a very high tax effort (e.g., Wellston) would ered desirable). Unfortunately, as the discussion be- be treated the same as districts making a relatively low suggests, both measuring performance and link- low tax effort (e.g.. Hancock Place). In the long ing performance to expenditures are exceptionally run, the equity problem could become an efficiency difficult undertakings in the field of public educa- problem. Rewarding districts for low tax effort could tion. encourage similar tax strategies in other districts. Thus, even though it would be relatively inexpensive PATTERNS OF PRODUCTION to establish a funding floor, doing so raises an equity Student-teacher ratios are often used as a partial question, and maintaining that floor could become indicator of school quality, assuming that lower ra- increasingly expensive. tios facilitate greater individual attention to students. The four districts involved fare much differently Although student-teacher ratios were not readily under the state Foundation program (see Figure available for this study, the number of certificated 7.4). One of the four is Jennings, discussed above as personnel, that is, teachers and administrators as the district least favored, in terms of assessed valu- opposed to nonprofessional support staff, was used ation, in the distribution of Foundation funds. In to compute a related index, here termed a student- 1984-85, Jennings received $321 per student less in educator ratio. Considerable variation exists on this Foundation money than its assessed valuation would indicator, in both elementary and secondary educa- predict. Simply bringing Jennings up to the level of tion, among the 23 St. Louis County districts. Such Foundation aid predicted by its assessed valuation ratios are, at best, however, only an intermediate indicator of school performance, since the ultimate districts diverge from the estimated values shown on product is an educated student, not the placement the regression line. of professionals in schools or teachers in classrooms. If the school districts in Figure 7.5 are distin- Student achievement might be considered a bet- guished by size, small and medium size districts ap- ter indicator of performance, except that the level of pear to be as capable of translating their resources achievement, as measured by standardized tests, is into favorable student-educator ratios as large dis- known to be related more to student background tricts, given various levels of spending. Neither side than to school-related variables. The school districts of the regression line is dominated by districts of a in St. Louis County also vary widely in the percent- particular size range. There is no evidence from this age of students who passed three basic skills tests (in analysis that even the smallest districts in the county mathematics, reading, and government) admini- necessarily suffer, with respect to elementary stere$ to eighth graders in 1982. This variation schools, from diseconomies of small scale. All dis- should not be equated with variation in school per- tricts in the county, however, benefit from a variety formance, given the influence of student back- of joint production efforts that may account for the ground on student achievement and the substantial efficient performance of small and medium-size dis- variation in student background observed across districts. These efforts are discussed in the next major tricts. section of this chapter. Figure 7.5 can also be used to ascertain an esti- Variation in Production mated spending level required to produce, effi- Many observers would expect that widespread ciently, a given student-educator ratio. To establish variation in district size would be associated with dif- a funding floor at an amount estimated to produce ferences in performance. Frequently it is assumed 15 students per educator, for example, would re- that larger districts are able to draw on economies of quire an estimated current expense of about $3,600 scale, including the services of specialized profes- per student. A ratio of 16 students per educator is sionals, that increase productivity and efficiency. estimated to cost $3,100 per student. Figure 7.5 plots 1984-85 current expenditure per Figure 7.6 plots spending per pupil against stu- pupil against student-educator ratios in elementary dent-educator ratios for high schools, by district. A schools, by school district. The downward sloping somewhat different picture emerges here. The fit of line on the graph is a regression line that statistically the scatter of points, each representing a district, to summarizes a linear relationship between current the regression line is not as good. While there is a spending per pupil and student-educator ratios. The relationship between spending per pupil and stu- relationship is negative, as expected. The more dis- dent-educator ratios in high schools, it is not as tricts spend per pupil, the lower the student-educa- strong as the relationship in elementary schools.17 tor ratio. Eighteen of the 23 districts hug the regres- For this reason, the ratio is a less compelling indica- sion line closely, lending credence to the hypothesis tor of performance in high school education. that districts strive to translate their financial re- Breadth of program, for example, also deserves con- sources into more educators per student.16 Included sideration, but no indicator of performance on this among these 18 districts is a broad size range, en- criterion was readily available. compassing both some of the smallest and some of Unlike elementary schools, a relationship be- the largest districts in the county. Controlling for tween district size and student-educator ratios spending per pupil, size of district is unrelated to stu- emerges among high schools. The very small districts dent-educator ratios. tend to have fewer students per educator. All three The vertical distance of each district from the of the districts positioned well below the regression regression line could be interpreted as a measure of line in Figure 7.6 are quite small, operating a single, the relative efficiency with which a district translates small, high school. Valley Park, with a 142-student its fiscal resources into a favorable student-educator high school, reports a student-educator ratio of 5:l. ratio, assuming this is what they strive to do. Outly- The low ratios for very small high schools may, in ing districts, it should be emphasized, should not be some cases, suggest the presence of potential econo- assumed to be more efficient or inefficient because mies of scale not currently being captured. All high we are unable to determine what accounts for their schools, in order to meet accreditation require- position relative to the others. Nonetheless, those ments, must maintain some degree of program districts falling below the line are relatively more breadth-some menu of courses. This means hiring successful with respect to this particular measure, teachers with diverse specializations. In very small and those above the line, relatively less so. The stan- schools, the result may be a decrease in the student- dard error of the estimate is 1.86 students per edu- educator ratio. Larger high schools with similar lev- cator; the standard error is the average amount the els of spending per student do not seem to put their Figure 7.5 Relation between Current Expenditures Per Student and Student-Educator Ratio for Elementary Schools, by District, St. Louis County- 1984-85 ementary Students Per Educator

3500 4000 4500 5000 5500 Current Expenditures Per Student ($) St. Louis County School Districts 1 Hazelwood 9 Amon 17 Normandy 2 Ferguson-Florrisant 10 Bayless 18 Ritenour 3 Pattonville 11 Brentwood 19 Riverview Gardens 4 Rockwood 12 Clayton 20 University City 5 Kirkwood 13 Hancock Place 21 Valley Park 6 Lindbergh 14 Jennings 22 Webster Groves 7 Mehlville 15 Ladue 23 Wellston 8 Parkwav 16 Figure 7.6 Relation between Current Expenditures Per Student and Student-Educator Ratio for High Schools, by District, St. Louis County- 1984-85 High School Students Per Educator

I I I I I I I I 1 I I I I I I I 1 1 I 2obo125O0 3000 3500 4000 4500 5000 5500 6000 6500 7( Current Expenditures Per Student ($) St. Louis County School Districts 1 Hazelwood 9 Affton 17 Normandy 2 Ferguson-Flonisant 10 Bayiess 18 Ritenour 3 Pattonvllle 11 Brentwood 19 Riverview Gardens 4 Rockwood 12 Clayton 20 University City 5 Kirkwood 13 Hancock Place 21 Valley Park 6 Lindbergh 14 Jennings 22 Webster Groves 7 Mehlville 15 Ladue 23 Wellston 8 Parkway 16 Maplewood money into matching the low ratios found in many ratios. On the other hand, one might also expect smaller schools. Instead, they may invest in still that smaller schools, by allowing less flexible assign- greater program breadth, or in better educated fac- ment of teachers to grade levels, would tend to pro- ulty, while allowing the student-educator ratio to in- duce smaller class sizes. Neither view is reflected in crease. the data from St. Louis County. School size is not Two small districts, Hancock Place and related to student-educator ratios in the case of ele- Brentwood, shift from well above to well below the mentary education. Small schools may of course be regression line between Figures 7.4 and 7.6, that is, preferred for other reasons (e.g., a more intimate between elementary and secondary education. Han- school community), and appear to be consistent cock Place is the most poorly funded district with favorable student-educator ratios. Webster in the county, while Brentwood is one of the best Groves, which reports the lowest such ratio for its funded. Both districts evidently trade off the level of current expense of any district, also operates student-educator ratio in their elementary schools relatively small elementary schools, averaging 286 in favor of a better ratio for their small high schools. students per school. (It should be noted, however, that Brentwood also ranks first among the districts in student achieve- Variation in Student Achievement ment; see Table 7.3 and related discussion below.) Student achievement on standardized tests is an A larger, more rural district, Rockwood, shifts in the alternative indicator of school performance, dis- opposite direction, apparently maintaining a more played by district in Table 7.3. The major difficulty favorable elementary school ratio as opposed to re- with this indicator is that variation in student ducing the ratio at its two high schools. achievement has been shown to be mainly a function The school districts also vary in the average size of social background rather than school variables. of their elementary schools, ranging from 225 stu- To a large extent, therefore, differences in student dents to 61 1. One could argue that larger schools, achievement among school districts reflect variation by generating economies of scale at the school site, in student population characteristics, rather than would allow districts to improve student-educator variation in school performance. In large data sets, it

Table 7.3 County- School Districts Ranked by Student Achievement on Standardized Tests of ~asicSkills- 1982 Overall Math Reading Government District Percent Passing Percent Passing Percent Passing Percent Passing Brentwood 100% 100% 100% 100% Affton 97 98 100 97 Clayton 95 96 100 99 Lindbergh 93 94 99 96 Ladue 92 95 100 96 Kirkwood 89 93 98 92 Bayless 86 88 100 95 Parkway 86 89 98 94 Mehlville 84 90 99 9 1 Rockwood 83 87 97 90 Hazelwood 82 89 97 88 Webster Groves 81 88 96 86 Ferguson-Florrisant 80 86 96 89 Pattonville 79 88 97 87 Ritenour 73 78 9 1 90 Maplewood 70 n 92 87 Hancock Place 70 74 95 87 Jennings 68 78 97 79 Riverview Gardens 66 71 95 86 University City 64 71 94 78 Valley Park 59 61 95 68 Wellston 57 75 9 1 76 Normandy 48 62 84 64 is possible to control statistically for the effects of If Wellston's performance on the separate tests of students' social background in order to estimate the mathematics, reading, and government is used as an effect of different types of schools and districts, but indicator (see Table 7.3), its ranking is still higher. this approach can be used with only limited success The nearly all-black district is also one of the small- in an analysis of 23 districts. est districts in the county, and one where taxpayers Nevertheless, student achievement in St. Louis dig deeply into their pockets to support their County is clearly related to poverty. The percentage schools. of the population of a district defined as poor accounts for more than half of the variance in the per- JOINT PRODUCTION ARRANGEMENTS centage of students who pass their eighth-grade tests The production of direct services to students, of basic skills when the Wellston district, an outlier, with the exception of special education and voca- is removed from the analysis.'B (Wellston is an out- tional-technical training, is organized separately by lier because it performs much better than its poverty each school district and carried out, for. the most rate would predict.) The three districts with the part, independently. Education for the handi- highest poverty rates, including Wellston, also have capped, vocational-technical training, and indirect the lowest rates of student achievement overall. Pov- services that benefit all students are organized, how- erty rate is a more consistent predictor of student ever, in a much different manner. Instead of acting achievement in St. Louis County than either race or independently, the 23 regular school districts and median household income. one special district have entered into a variety of No relationship is evident between student- joint production arrangements that enable each dis- educator ratios in the elementary schools and trict to draw on the combined resources of all of student achievement by district. Despite the evident them. These arrangements include both bilateral strategy of most districts to use financial resources to ties, such as those that link the special district to secure a lower ratio, the differences in number of each of the regular districts in the county, and a students per educator do not seem to account for number of consortia and similar associations, some much, if any, of the differences in student achieve- of which include school districts outside the county ment among the districts. This result, while disap- as well as private school systems. pointing, underscores the complexity of educational Special Education and processes and the difficulties of relying on any single Vocational-Technical Training indicator of school quality. The Special School District engages in two basic On the other hand, spending per pupil appears production tasks with respect to handicapped stu- to be related to student achievement, although it is dents: evaluation and instruction. In performing difficult to know how much weight to attribute to this these tasks, the special district is connected to each factor. Spending is also correlated highly with stu- of the 23 regular districts in the county. These pro- dent social background. The relationship does hold duction arrangements entail patterns of both spe- up, however, as statistical controls are introduced cialization by clientele and coordination with respect serially for students' social background, including to the same clientele. race, poverty, and median household income in the The referral and evaluation procedures through district .lQ which a student gains access to special district serv- Size of district is not consistently related to stu- ices require coordinated performance on the part of dent achievement. The smallest districts, however, administrators in the regular district where the stu- are clearly capable of performing quite well. The dent resides and the special district staff. Once a stu- single best academic achievement of any district is dent is referred by his or her home district, special registered by Brentwood, with a total student popu- district counselors, the student, and sometimes the lation of 729 students. Brentwood is a relatively low student's home district jointly produce an evaluation income community (median household income is session. The regular districts vary in the extent to $16,654), but it enjoys substantial assessed valuation which they contribute to the production of evalu- and is the second highest spending district in the ations. A few districts go so far as to complete all of county. the paperwork and some of the evaluation proce- Wellston, discussed above as the most seriously dures before calling for special district intervention. disadvantaged district in the county, finished next to Other districts wait for special district intervention last in overall student achievement. Because its after completing only the necessary referral and ap- demographic standing is by far the worst in the plication procedure. county, it is significant that Wellston's academic per- The evaluation session determines whether and formance is not the worst. Instead, it finished 9 per- to what degree a student is learning disabled, and centage points ahead of the lowest achieving district. signals commencement of instructional service, which may be either produced independently by the to deal with administrative problems at the school special district in its own schools or coordinated with site. The special district would become purely a pro- the regular school districts, depending on the degree vision unit in this case, and the regular district would of a student's learning disability or handicap. Stu- become the only production unit. dents are accepted by the special district if they are In the current arrangement, regular districts evaluated as being Phase I (mildly), Phase I1 (mod- contribute space and facilities for education con- erately), Phase I11 or Phase IV (severely) handi- ducted by the special district with their employees. capped. Those who are evaluated as being Phase I11 The special district is obligated, at least in principle, or Phase IV handicapped are referred for enroll- to pay rent for the space it uses in any regular dis- ment in one of 12 schools operated independently trict school. At least one district, Clayton, does not by the special district. Students who are evaluated as collect the rent as a matter of policy, and another being either Phase I or Phase I1 handicapped are reports receiving no payment over a two-year pe- mainstreamed, integrated into educational programs riod. On the other hand, the Riverview Gardens dis- within their home districts. The special district is the trict rents an entire building to the special district for only public school district delivering services to use in its Phase I11 and IV programs. Phase I11 and Phase IV handicapped students in the While the special district is an exclusive provider area. However, with respect to Phase I and I1 handi- of special education services (in the public sector), capped students, the special district coordinates with several of the 23 regular school districts in St. Louis the regular districts to produce direct services for County offer vocational programs for resident high students. school students. The Clayton district, for instance, Within the regular district schools, a Phase I stu- offers its students a part-time high school, and the dent, one handicapped with a milder form of learn- Ritenour district operates a vocational preparatory ing disability, speech or behavior disorder, may school. Nonetheless, high school students from regu- spend no more than 50 percent of each school day lar school districts may choose to pursue vocational- in a resource room with a special education teacher. technical training at one of the three Special School The larger part of the student's school day must in- District technical high schools. Admission to the volve him or her in the school's regular curriculum. technical high schools is granted on a first-come, Phase 11 students spend more than 50 percent of first-served basis. However, under Sections 12B and their time, but less than 100 percent, with a special 12C of the Voluntary Interdistrict Desegregation education teacher. In both cases, the special educa- Agreement, discussed below, the special district tion teacher is an employee of the special district must reserve at least 25 percent of the enrollment who nevertheless works as part of an education team spaces in its vocational-technical schools for transfer in a school operated by a regular district. Teachers students from St. Louis City high schools. The Spe- with two different employers jointly deliver services cial School District accepts students from school dis- to the same students on a coordinated basis. A spe- tricts in surrounding counties as well, providing vo- cial education teacher must work as pan of a team cational and technical training to these students on a of teachers and within a common facility supervised tuition basis. The special district also coordinates by a principal who works for a different school dis- ~4ththe St. Louis County Department of Human trict. Services to operate a summer occupational training The regular district superintendents interviewed program established under the auspices of the Job for this study expressed overall satisfaction with their Training Partnership Act. working relationships with the special district. Some voiced a concern over delays in evaluation and re- Indirect Services ferral, but also noted improvement in this regard. Several types of production relationships link Most saw no particular problem involving the place- the school districts in the St. Louis metropolitan ment of special education teachers in their schools. area. Although these districts are authorized by state The Ferguson-Florissant district was seeking, how- law to engage in cooperative endeavors to deliver ever, to work out different arrangements for serving elementary and secondary education (Sections Phase I and I1 students who are speech and lan- 70.210, 70.220 Missouri Revised Statutes), they guage disabled. Instead of using teachers employed nonetheless remain, for most purposes, autonomous by the special district, Ferguson-Florissant proposed production units. Production of direct education that it hire its own special education teachers with services to public school students continues to be or- special district funds and perform its own evaluation ganized by individual school districts. Yet, the ability and referral. The objective of this plan is to decrease of the local public school districts to function effi- the lag time currently involved in evaluation and re- ciently as autonomous units depends in part on their ferral through the special district staff and to be able ability to develop joint, interdistrict arrangements, intended to capture economies of scale in the deliv- rial vendors are supplying and about the discounted ery of various types of indirect or auxiliary education cost per unit for the supplies at various quantities services to public schools. purchased. Districts then decide whether they wish to purchase supplies from the vendor secured by The Cooperating School Districts of the Subur- CSD, and at the stated price. Once a district chooses ban St. Louis Area. Perhaps the broadest arrange- to place an order, it is given a 10 percent margin for ment involving public elementary and secondary error in the number of units ordered. Suppose, for schools, including the Special School District, is instance, that CSD arranges for a vendor to supply their joint participation in the Cooperating School microscopes and other pieces of laboratory equip- Districts of the Suburban St. Louis Area (CSD). ment at given prices per unit. If a particular district CSD originated in 1928 when the school superinten- is not satisfied with the quality or price of the micro- dents in St. Louis County began holding informal scopes that are offered by the vendor, the district meetings to discuss issues of common concern. may look elsewhere, as it chooses. If the district Since that time, the member districts have expanded does choose to purchase microscopes at the ven- to include all of the districts in St. Louis County as dor's price, then it can submit an initial purchase well as a few districts outside the county. The serv- order that varies as much as 10 percent from its ices and facilities now include an extensive film li- eventual purchase without penalty. brary and audiovisual services, a cooperative pur- Compared to many purchasing arrangements chasing program, a joint data processing service, a within large school systems, the CSD program has pre-college guidance program, and an annual considerable flexibility. In 1974, the executive com- countywide music festival. mittee created an ad hoc committee to explore pos- Since the time of its formal creation, CSD has sible advantages of constructing a central warehouse been a nonprofit organization funded entirely for CSD educational supply purchases. The commit- through dues paid by member school districts. Dues tee concluded that the costs, especially costs for for full or participating members are determined an- maintaining a supply warehouse and for hiring addi- nually on the basis of the district's assessed valuation tional administrative staff, were prohibitive. Instead (50 percent) and average daily attendance (50 per- CSD chose to retain the original, decentralized, bid/ cent). Associate (nonvoting) members (which in- no bid structure of the cooperative purchasing pro- clude the St. Louis city school district) pay a flat gram. $250 that entitles them to participate only in the co- The audiovisual education department of CSD is operative purchasing program. The CSD executive reported to be the largest such facility for grades committee meets monthly, except during summer K-12 in the nation.Z0 Current library holdings are in months, and makes all decisions that bind CSD con- excess of 26,000 programs with an estimated value tractually. The organization has a separate adminis- of $6 million. Services to member districts and trative staff and an executive director. schools include instructional television, video dupli- New members of the eight-member executive cation, and the delivery of instructional materials. committee are selected annually during the confer- All of the districts in St. Louis County do not ence of superintendents and school board members. benefit to the same degree from their membership in Nominations for committee membership may come CSD. The larger districts in the county-those with from the floor, but no nominations have been made several thousand students and a dozen or more from the floor since 1975. Instead, the executive di- school buildings to maintain-can, in many in- rector and the current executive committee draw up stances, purchase in lots sufficiently large to capture a slate of candidates and submit that slate to the su- independently any price discount available through perintendents in member districts before the confer- joint purchasing. As is the case for all the school ence. Generally the slate is composed in accordance districts in St. Louis County, the largest districts con- with three criteria: geographic balance, district tinue to participate as full members of CSD, paying a wealth (balancing wealthy and poor districts), and fee based on enrollment and assessed valuation. previous service on the committee (those who have One reason for this could be that even the largest not been represented in some time being favored districts in St. Louis County are not large enough to over those who have). produce, efficiently, the other services supplied Under the cooperative purchasing program, the through CSD, especially the 26,000 volume film li- executive committee and director of CSD arrange brary and audiovisual services. The St. Louis city vendors for classroom, sports, office, and other district does maintain its own audiovisual depart- types of supplies for which districts have a common ment, and may be large enough to maintain it effi- demand. Once vendors are arranged, each school ciently, that is, large enough to capture separately district receives information about the types of mate- any discount that it would receive by virtue of joint purchasing through CSD. This would account for the topics of interest to users, and an electronic bulletin decision by the St. Louis city district to hold only board for student users as well as teachers and associate membership in CSD, allowing for selective school administrators. The consortium also sponsors participation in the cooperative purchasing program. a number of roundtables, conferences, and in- One of the services that is still produced for service workshops centering on the uses of computer some of the districts by CSD, but that may revert to technology for educational purposes. in-house production among many of the districts in the area, is data processing. As computer technology Municipal and Community Relationships has become less expensive to purchase and operate, Various arrangements link the regular school a number of districts in the area have purchased mi- districts with the municipalities that they overlap and crocomputers and have begun to process their own the communities they serve. The school superinten- payrolls and so forth, instead of using the data proc- dents who were interviewed described the general essing services offered through CSD. However, ad- character of these arrangements as informal, but vances in computer technology, which have reduced generally considered such relationships to be good. the costs associated with in-house data processing, Some districts do have formal contractual rela- have created demand among microcomputer users tionships with municipalities for specific purposes. for operator training, machine repairs, service sup- The Jennings district, for example, in cooperation port, and so forth. This demand is being met by an- with the City of Jennings, operates an adult educa- other organization discussed below. tion program and a youth summer recreation program. As part of this arrangement, the district pro- The Regional Consortium for Education and vides school facilities at no cost to the city. While Technology. The Regional Consortium for Educa- the city hires employees, including district teachers, tion and Technology (RCET) is a new organization to operate both programs, the school district collects that specializes in producing information about and tuition fees for the two programs and turns these assistance in the utilization of computer technology funds over to the city. for educational applications. RCET integrates busi- Several of the school districts in St. Louis ness interests with institutions for public and private County also permit municipalities to use school fa- education, being an organization open to public cilities for various types of public functions, includ- school districts, private schools, colleges and univer- ing municipal elections, town meetings, charity sities, nonprofit organizations, and private business events, and community service programs. The enterprises. Currently, 20 of the regular school dis- Hazelwood district, for instance, allows smaller mu- tricts in the county, the special district, and the city nicipalities within its boundaries to use school facili- district are members of RCET. Consortium members ties for public meetings and as polling places, paying also include several public school districts outside of rent only to cover utility and clean-up costs. The the St. Louis area, as well as the St. Louis Archdi- larger municipalities that lie partially within the ocesan school district, the Lutheran schools of the Hazelwood district-Ferguson, Florissant, and St. Louis area and the Educational Confederation, Hazelwood-have sufficient municipal facilities for an organization of 22 private schools in the area. public meetings and, therefore, do not rely on dis- The Cooperating School Districts (CSD) also main- trict facilities. tains a separate membership in RCET. The Ritenour district permits several churches School districts wishing to participate in the con- that have pooled funds to operate a meals-on-wheels sortium may choose to join as either associate mem- program for the elderly to run their program from bers or general members. Within each membership Home Heights, a closed elementary school. While classification there are five levels of dues based on the kitchen used by the church organization is lo- student enrollment. However, the services that a cated in the basement of the school, the Ritenour school district may receive under the alternative district also operates a community education center classes of membership differ only in terms of the fee for GED, arts and crafts, and other courses in the that school districts must pay for computer repair classrooms above. The district buys the teachers' services provided by RCET. time and sells it for tuition. Besides repairing machines. RCET provides a The Ferguson-Florissant district is currently cooperative purchasing program for computer hard- leasing space in a closed elementary school to the ware and software, a computer lab where members City of Florissant, which is operating a similar meal can receive instruction in using microcomputers, a service for senior citizens. Ferguson-Florissant also resource center and software library where member leases school facilities to the smaller municipalities in districts may send individuals to preview hardware its area for public gatherings. The City of Berkeley. and software, technical workshops organized around which lies partially in the Hazelwood district and partially in the Ferguson-Florissant district, leases uses Clayton High school and pays a rent of approxi- facilities for public meetings from both districts. mately $10-15 per classroom to the Clayton district The most tightly knit and wide-ranging relation- for use of the facility. The Community College dis- ships between school districts and municipalities in trict also hires one of the Clayton district's assistant St. Louis County seem to occur in those instances superintendents as building administrator during the where a district is more or less coterminous with one evenings when classes are offered and employs some or two municipalities. The superintendent of schools of the Clayton teachers as instructors for the adult and city manager in University City, for example, courses. The Clayton district has also articulated meet regularly and share agendas for school board community preferences for the types of courses to and city council meetings. The district and the city be offered by the Community College, shifting from have a history of cooperating in a number of more a more academic type to arts and crafts and recrea- substantive areas also. Around 1950 the city gave tion.21 property to the school district to construct a high school, stipulating that the district would also have to ARRANGEMENTS FOR PUBLIC build and maintain an indoor swimming pool. The SCHOOL DESEGREGATION school district did so, and continues to maintain Article IX of the Missouri Constitution of 1945 both. The school district and the municipal street originally required two sets of public schools and maintenance department routinely share ma- throughout the state, calling for "separate but equal chine parts and some equipment. In addition, city schools for children of African descent." Since the maintenance crews remove snow from school en- U.S. Supreme Court ruled in 1954 that legally sanc- trances and lots during the winter. In 1986 the dis- tioned equal but separate public facilities are uncon- trict and city began investigating the possibility of stitutional, the racial composition of public schools combining their separate grounds crews into a single has been a source of conflict in central cities and unit. metropolitan areas. In the St. Louis Metropolitan area, the Voluntary Interdistrict Coordinating Coun- Bilateral Arrangements cil (VICC) operates as the linchpin in an areawide Between School Districts plan for public school desegregation. Since 19 8 1, Only one formal interorganizational arrange- school desegregation throughout this area has gone ment relating two school districts in the production forth in two phases: first under an Intracity Plan, of a direct education service was identified during centering exclusively on the St. Louis city school dis- interviews with 13 of the 23 school superintendents trict, then second, under an Interdistrict Plan involv- in St. Louis County. This relationship involves the ing the city district and 16 county districts. Affton and Lindbergh school districts. Affton and Lindbergh cooperated for the first time during the The lntracity Plan 1985-86 school year to provide an early childhood In 1972, a group of black parents residing in St. education program. The program, which is head- Louis City, who collectively have come to be called quartered in Truman, a closed elementary school in the "Liddell Plaintiffs," filed suit against the St. the Lindbergh district, enrolls regular preschool-age Louis City board of education. The parents charged children. However, some special preschoolers, that school board members were in violation of the mostly with language disabilities, are enrolled in the Fourteenth Amendment to the U.S. Constitution by program. The program has been well received by perpetuating a segregated school system in the city. parents in both districts, and the districts planned to When the suit was brought, black students outnum- increase funding for the program in the 1986-87 bered white students in St. Louis City schools by school year. Although the early childhood program nearly four to one, and 92 of the 120 schools in the provided by these districts enrolls some special stu- city district were attended exclusively by children of dents, the program is not formally linked with the one race. At the same time, 30 of the city schools Special School District. While the preschool pro- employed all white faculties, and more than 90 per- gram employs its own special education staff, the di- cent of the faculty members in the remaining 90 rector of the preschool program shares information schools were black. After four years of litigation, the with Special School District staff. initial outcome of the suit was a court order requir- Many of the school districts, independently or in ing St. Louis City's board of education to propose conjunction with municipalities, also provide adult some acceptable method for desegregating the education courses. School districts cooperate with schools under its charge. In doing so, the board pro- the Community College district of St. Louis County posed and began implementing its first desegregation to offer education services to adults. In the Clayton plan in 1976, the centerpiece of which was the crea- district, the countywide Community College district tion of 11 magnet schools in the City of St. Louis. The 1976 plan, however, did not settle the case. among all the communities of interest-that is, the The original plaintiffs continued to press, and even 25 school districts in St. Louis City and County. expanded, their case against the city board, finding The agreement explicitly established the Volun- support from both the Missouri Department of Ele- tary Interdistrict Coordinating Council (VICC) to mentary and Secondary Education and the U.S. De- monitor implementation of transfers of students and partment of Justice. Finally, in 1980, eight years af- staff between the St. Louis City district and those ter the original suit was filed, the U.S. Eighth Circuit districts in St. Louis County whose enrollments are Court of Appeals upheld the plaintiffs claims and less than 25 percent black. The agreement also re- ordered the city board to launch a desegregation quires the council's executive director, who is an ex plan more expansive than its 1976 effort. Besides officio member of the council, to report at least an- creating and allowing for voluntary student transfers nually to the federal district Court of Appeals on the to the magnet schools, the Intracity Plan called for plan's performance. The council is composed of 27 the introduction of a new system of middle schools, members selected as follows: one school board mandatory reassignment of students for more uni- member or administrator from the St. Louis City dis- form racial balance throughout the district, and the trict and from each of the 23 districts in the county, enrichment of compensatory and remedial programs one member representing the Liddell Plaintiffs, one in those schools having a predominantly black en- member from the NAACP, and one member from rollment. However, because St. Louis City schools the Missouri Department of Elementary and Secon- have so few white students compared to the number dary Education. VICC is organizationally partitioned of black students, it was impossible to integrate into one standing committee-a magnet review com- every school in the district. In fact, 60 schools in the mittee-and four subcommittees-policies and pro- district still have enrollments that are wholly or pre- cedures, staff development, staff exchange and dominantly black. transfer, and student transportation. In terms of performance, the interdistrict agree- The lnterdistrict Plan ment set a five-year time frame during which up to 15,000 black students in St. Louis City public Because the Intracity Plan left nearly 30,000 schools may opt to attend any of 16 predominantly black students in St. Louis City enrolled in schools white school districts in St. Louis County. The whose student populations remain wholly or pre- agreement states that predominantly white school dominantly black, the Court of Appeals informally districts are those whose enrollments are less than proffered, but did not formally mandate, a more 25 percent black. In accordance with the agree- ambitious desegregation program involving voluntary ment, the seven county districts with black enroll- transfers of students and teachers between the city ments of 25 percent or more-Ferguson-Florissant, and county school districts. When an early voluntary Jennings, Maplewood-Richmond Heights, Nor- transfer plan was implemented in 198 1, only five mandy, Riverview Gardens, University City, and districts in St. Louis County participated. Subse- Wellston-cannot accept students transferring from quent to this initial effort at interdistrict desegrega- St. Louis City. The agreement also provides for vol- tion, the case took an interesting turn when the St. untary transfer of a comparable number of white Louis city district joined with its adversaries, the students in St. Louis County public schools to any Liddell Plaintiffs and their supporters who had origi- St. Louis City schools, including the magnet schools. nally sued the city district, to bring suit against the Altogether, during the 1984-85 school year, 5,3 17 school districts in St. Louis County. After a series of students in the metropolitan area transferred across negotiations among the Liddell Plaintiffs, the school districts, 447 from county to city schools and NAACP, the Missouri Department of Education, 4,870 from city to county. During 1985-86, the the St. Louis city school district, and representatives number of students transferring to different schools of the 23 regular school districts in St. Louis County, in the metropolitan area rose to 7,528-542 trans- a voluntary agreement for interdistrict desegregation ferring from county to city, and 6,986 transferring was finally reached in 1983, precluding a court-im- from city to county.22 posed mandate. This landmark settlement, first in- This arrangement, by allowing for voluntary stu- stituted during the 1983-84 school year, is voluntary dent transfers across district boundaries, has some in the sense that no overarching authority unilater- attributes of a voucher system, while at the same ally imposed a desegregation plan on the St. Louis time avoiding an externally imposed system of bus- metropolitan area. Nevertheless, the expressed willing as a remedy to desegregate schools in the area. ingness on the part of the Court of Appeals to design The limiting factor in this arrangement is that not all and impose its own interdistrict arrangement pro- students in the area are permitted to choose what vided strong impetus for good faith negotiations school they will attend. Black students in county school districts (whether the district is Ladue, with are scheduled to run exactly and only in accordance about a 3 percent resident population of black stu- with the regular school day, possibilities for some dents, or Wellston, with virtually 100 percent) can- students to participate in extracurricular activities at not transfer under the interdistrict plan. This is also their new schools are foreclosed. If a transfer stu- the case for white students in St. Louis City who dent cannot arrange for a ride home after the Ger- might wish to transfer to a particular school in the man club or student council meeting or after band, county. The opportunity to choose is limited to basketball, or drama club practice, then the student black students in St. Louis City and white students has less opportunity to become fully integrated into in St. Louis County. the activities of the new school and to take advan- Several aspects of the plan are quite innovative. tage of all that is offered. Included are the development and extension of a These sorts of limiting circumstances were rec- network of magnet schools, a transportation ar- ognized as problematic during the negotiations over rangement that permits students who are allowed to interdistrict desegregation. This recognition led to transfer to participate in extracurricular activities, further negotiations between VICC and the Missouri and an arrangement authorizing voluntary teacher Department of Elementary and Secondary Educa- transfers and exchanges among any of the 24 dis- tion, resulting in a transportation arrangement that is tricts in the metropolitan area. intended to encourage transfer students to engage in Although most of its work is directed toward the extracurricular activities. Each year since 1983, the school district in St. Louis City, since this is where state education department has allotted sufficient the preponderance of magnet schools is located, the funds to provide for transportation of interdistrict magnet review committee is charged with evaluating transfers students who participate in extracurricular programs in a developing chain of magnet schools activities scheduled before or after regular school (27 city and 1 county). In 1984-85, the committee hours. evaluated 12 magnet programs in the city district The organization of bus routes and schedules for and one in the Kirkwood school district in the transfer students is quite complex, especially since county. Students interested in specialized instruction these routes and schedules are designed to facilitate may now choose from magnet programs that include both student choice among schools and student par- a Center for Management, Law, and Public Policy; ticipation in extracurricular activities. During the Junior and Senior Classical Academies; Centers for 1983-84 school year, 46 percent of all transfer stu- Visual and Performing Arts; a Naval Junior ROTC dents reported participating in an extracurricular ac- Academy; and Centers for Individually Guided Edu- tivity, and 27 percent reported participating in three cation. The magnet schools play a critical role in or more activities. During 1984-85, the total number maintaining the reciprocal nature of the plan be- of transfer students reporting extracurricular partici- tween city and county, that is, in maintaining a two- pation rose to 54 percent, with 34 percent participat- way flow of transfer students. More than 90 percent ing in three or more activities. The total cost of of all county-to-city transfers are transfers to magnet transportation for interdistrict transfer students, in- schools. Without the magnet program, it is doubtful cluding transportation to part-time programs and ex- that voluntary county-to-city transfers would have tracurricular activities, amounted to approximately reached current levels. $8 million during the 1984-85 school year, all of Student transportation, or busing, remains the which has come from the budget of the state educa- most complicated and controversial issue associated tion department. with programs for school desegregation. In some ar- The State of Missouri also provides compensa- eas, absolute mileage of student transportation need tion to school districts that lose students by transfer. not necessarily increase as a result of desegregative Districts that receive students by transfer also re- transfers. The complexity, though perhaps not the ceive full state aid, distributed by enrollment, for controversy, can be alleviated to a large extent un- that student. At the same time, the student's district der desegregation plans where students are unilater- of residence continues to receive half the amount of ally reassigned, and where bus schedules can be de- state aid to which the district would have been enti- signed to transport students to school in time for the tled had the student not transferred. These tuition beginning of the regular school day and pick them incentives amounted to $6.4 million for 1983-84 up when the final bell has rung. One of the major and, as the number of transfers increased, $14.9 problems associated with busing students, especially million for 1984-85. when buses carry students across district boundaries, One simple gauge of the effectiveness of the de- is to give students who transfer to another school segregation plan is the degree of conflict or coopera- opportunities to participate in activities that do not tion that has surrounded its implementation. As coincide with the normal school day. When buses noted before, most of the school districts in St. Louis County did not join the program until there as well as how a district is treated by state aid formu- was a strong likelihood that they would be subject to las. a court-ordered desegregation plan. Nevertheless, It should be noted that fiscal disparities also ex- once it became apparent that the U.S. Court of Ap- ist among schools within large, consolidated jurisdic- peals would mandate school desegregation for the tions. Absolute equality in the distribution of re- area, and probably design and implement its own sources is unattainable. One possible standard of plan, the districts in the area voluntarily entered ne- equity is some limited degree of dispersion about a gotiations to seek a mutual accommodation in de- mean. In this sense, the fiscal equity performance of signing a plan for the metropolitan area. VICC be- the St. Louis County system may be quite good, came the vehicle for working through conflicts to a though it could be made even better with a small, point of resolution. That the 24 districts in the area highly selective program of support for distressed were able, as independent jurisdictions, to negotiate a mutually agreeable settlement, and then to coordi- districts. One possibility would be to reserve a por- nate a highly complex and open system of school tion of the county railroad and utilities tax, already desegregation in the metropolitan area, would per- earmarked for public schools, to aid the most fiscally haps seem impossible to those who argue that in- distressed districts in the county, or to convert the terdistrict desegregation must be imposed from county tax distribution formula from an aliocative above-by state or national government-to be effec- (student population) basis to a redistributive basis. tive .23 The county school fund in 1984-85 totaled almost $10.5 million. just under $90 per student countywide. Concentrated on a few distressed districts, CONCLUSION however, this fund could be a much more significant It is not unreasonable to refer to the school dis- sum of money. The greatest difficulty is in defining tricts in St. Louis City and County as forming a mul- distressed districts. More than the level of spending tijurisdictional system of public education, especially per student must be taken into account in order to so with respect to the 24 districts in St. Louis avoid treating districts inequitably. Local tax effort County. The distinction between provision and pro- must also be a positive factor affecting redistribu- duction is helpful in understanding how this system tion. When this is done, some districts may still be works, as well as for examining the implications of left with significantly less to spend per student than the system for both efficiency and equity. Each others. school district functions both as a provision unit and as a production unit for direct services to students. Students in wealthy districts do not necessarily The use of state government as a basic unit of gov- cause students in poor districts to receive an inferior ernance, as well as a supplementary provision unit education. If there were a countywide school dis- for special education in the county, help to tie the trict, resources drawn from a few wealthy communi- districts together as an interdependent system. Per- ties would be spread over the entire county student haps more importantly, joint production units for in- population, albeit unevenly. It is not clear that the direct or auxiliary services have been created to op- resulting marginal increase in resources would be erate in addition to the separate school districts. The worth the loss of community control and account- result is a system that is fragmented in a jurisdic- ability for even the poorest communities. This is a tional sense, but not uncoordinated in a functional trade-off that local citizens must make when decid- sense. ing whether to consolidate districts. The state Foun- dation program operates, implicitly, on the assump- Fiscal Relationships tion that external resources are best combined with The system of schools in St. Louis County is local community preferences and local accountabil- characterized by sharp fiscal differences. Still, most ity. Given its level of funding, the program appears students are not extremely far from the mean level to operate with substantial success. Increasing efforts of district expenditure per student in the county. A such as these, and perhaps supplementing the state small minority (roughly 14 percent) is extremely well program with a county program, can increase equity off by this measure; fewer still (perhaps 6.4 percent) without undermining the accountability of produc- can be viewed as extremely disadvantaged. It is easy ers-education professionals-to the intended bene- to discern how those who are extremely well off got ficiaries of both redistribution and the ultimate edu- that way-the cause is high assessed valuations of location product. cal property per student. How the least well off came to be so is a somewhat more complicated question. Patterns of Performance Low assessed valuations per student are only part of Using the student-educator ratios predicted from the answer. District tax effort (presumably reflecting each district's assessed valuation per student as indi- local demand for public education) also plays a role, cators of performance in elementary schools, St. Louis County school districts are remarkably consis- regular classrooms. The case of indirect or auxiliary tent performers. Size of district, measured by num- services, or the case of services for special students ber of students in attendance, is unrelated to per- in special classroom situations, however, present dif- formance on this indicator. Districts from small to ferent situations. In these circumstances, the St. large perform much the same in translating fiscal re- Louis County system features an organizational sources into student-educator ratios. Neither econo- overlay of joint administrative units, maintained on a mies nor diseconomies of scale are much in evi- cooperative basis, and a special district that provides dence in elementary education. education for the handicapped (delivered to stu- The picture changes somewhat for secondary dents in part on a cooperative basis) and vocational- education. Very small districts tend to perform better than normal with respect to student-educator ra- technical training. These arrangements enable the tios, but the relationship between expenditure per separate school districts to pool their resources for student and students per educator is much weaker those functions in which economies of scale are than for elementary education. Such ratios are most likely to be present. therefore less likely to be an accurate performance It is especially significant that a countywide dis- indicator in high schools. It may be the case that trict demand for special education can be combined some of those schools exhibiting superior perform- with the resources of a regular school program in ance on this indicator are relatively inefficient, un- order to avoid duplication of facilities and give stu- able to capture potential economies of scale. dents the benefits of both experiences. Sharing of Some consolidation of small high schools in the school facilities with municipalities and with the county, therefore, may be worth considering. This Community College district also achieves economies would not necessarily entail a consolidation of for citizen taxpayers. Even private and parochial school districts. One district could pay tuition for its schools have been able to benefit from some of the students to attend high school in another district, or service arrangements initially developed for public two or more districts could consolidate their high schools. Functional coordination between overlap- schools without merging the districts. Another possi- ping jurisdictions occurs in St. Louis County, not bility is for two or more districts to agree on a recip- through central direction, but on the basis of two rocal open enrollment plan, allowing each district to develop different secondary school specialties. Of principal factors: (1) mutual need and (2) service to course, factors other than those considered in this a common constituency. As James Madison wrote analysis-such as the contribution of a high school to about federal and state governments, "officials in a sense of community in a residential area-may overlapping jurisdictions are but different agents and modify these conclusions. trustees of the [same] people, constituted with dif- Turning to levels of student achievement, it is ferent powers, and designed for different pur- clear that small district size is no bar to superior per- poses."24 Different agents who faithfully seek to rep- formance. Neither, of course, is small size a guaran- resent the same people in different, but related, tee. Variation in performance among a diverse set of matters have strong incentives to cooperate. small, homogeneous districts is to be expected be- Faced with the prospect of a highly divisive legal cause student achievement is highly correlated with battle over school desegregation, and the possibility social background characteristics. Another variable, of a conflict-laden implementation, the St. Louis unmeasured and unexamined in this analysis, is the system of public education responded with one of level of student achievement that a community finds the more creative and cooperative metropolitan ef- acceptable. If it were possible to compare district forts in the nation. The number of school districts in performance in light of community expectations and St. Louis County was not a deterrent to this arrange- standards, somewhat different results could emerge. ment. The most telling criticism one might advance A large number of small districts seems to generate of the school transfer program now in effect is that what one would expect: diversity. transfer to St. Louis County schools is limited to black students from St. Louis City. The potential is Indirect Services and Special Education there to do more, increasing opportunities for all Despite the substantial variation in district size disadvantaged students, not simply those now in- across St. Louis County, size is not associated with cluded in the program. Gradually increasing the school performance on the indicator of student-edu- ability of students to choose to attend schools out- cator ratios in elementary schools. This finding indi- side their district has the potential to slowly trans- cates few remaining economies of scale in the pro- form public education, increasing both school reduction of direct services to regular students in sponsiveness and equity. ENDNOTES sum of (1) the number of resident full-time students and the full-time equivalent number of part-time students who were 'After the field work for this study was completed, a citizens enrolled in the public schools of the district during the previ- movement was initiated to merge the Valley Park district- ous year, (2) the number of full-time resident students and the smallest district in the county-with an adjoining dis- the full-time equivalent number of part-time students who trict. As of this writing, the matter is unresolved. were enrolled in the public schools of the district on the last district superintendent interviewed for this study ob- Wednesday in January of the previous year and who were in served that many parents have deliberately relocated to St. attendance one day or more during the preceding ten school Louis County in order to take advantage of its high-quality days, and (3) the full-time equivalent number of summer education service to handicapped students. school students (Section 163.011 Missouri Revised Stat- =Mobility does not pose a problem from an economic effi- utes). ciency standpoint as long as it does not cause officials and OThe state board and commissioner of education recognized citizens to reduce their provision effort in order to discour- as early as 1984 that there were potential problems with the age migration. Foundation Program. In the annual state report on Missouri 41nterestingly, the percentage of district students attending public schools for that year the board skirted this issue by private or parochial schools is positively related to student arguing that "cutbacks in state funding have further weak- achievement, measured by the percentage of students who ened the equalizing effect of the Foundation Program-a pass an eighth grade test of basic skills. Private school at- major purpose of which is to reduce the disparity in spend- tendance, measured at the district level, is also correlated ing among school districts. " Nonetheless, the board con- positively, however, with those social background charac- cluded that "if these disparities are not corrected, our entire teristics that contribute to student achievement. There are school funding system is subject to legal challenge." The too few cases in St. Louis County to determine whether the commissioner and the board thus far have avoided such a effect of private school attendance on student achievement legal challenge by continuing to make adjustments in this in public schools continues to be positive after controlling formula for distributing the largest portion of state aid to lo- for the effect of student social background. The results ob- cal districts. tained from analyzing the 23 districts in St. Louis County lost. Louis City supplied 42 percent of its own funding for are not consistent in this regard. 1984-85, while spending a total of $4,164 per student from =Whether there is a competitive effect depends on whether all sources. The city district exceeds all county districts in state aid per student exceeds the marginal cost of education the amount of support it obtains from the state Foundation per student for a particular district. The proportion of state program, $1,314 per pupil in 1984-85. Comparisons be- aid varies roughly from 15 percent to 50 percent or more tween St. Louis City and the county districts are compli- across St. Louis City and County school districts. Whether cated, however, by the differences in functional responsibil- state aid is high enough and marginal costs low enough to ity between them. The city district collects state aid for engender significant competition is a question that lies be- handicapped students, for example, that, in the county, yond the scope of this report. goes to the special district. Because of these functional differences between the city and county districts, the city is ex- BSource: Cooperating School Districts of the St. Louis Sub- cluded from analysis of fiscal patterns among districts in the urban Area, St. Louis County Annual School Report, No- county. vember 1985. Also, Missouri State Board of Education, 1984-85 Report of the Public Schools of Missouri. Data "The distribution formula is identical to that used for the from the CSD report and the state report exhibit some dif- state Textbook Fund, described in Endnote 8. ferences, and adjustments were made to tabulated data for 120ne possible explanation for Jennings district's ill treatment purposes of consistency. centers on the mechanism which the Missouri Department 'Districts also vary in their reliance on deficit financing. The of Education uses to equalize assessed valuation per pupil mean debt service per student in 1984-85 ranges from $0 to across local school districts. The Foundation formula may $656 per student; the mean is $286. Districts that spend wrongly inflate some district tax bases, while it deflates oth- more per student from all current sources also tend to spend ers. The Jennings superintendent argues that in equalizing more per student on debt service. On the other hand, there assessed valuations across districts, the formula tends to is a slight tendency for districts that spend more per student overstate the assessed valuation of commercial property in to spend less per student on debt service relative to the his district and, therefore, decreases the amount of Founda- amount spent from district sources. tion revenue which he feels his district should actually receive. %The only fund from which revenues are distributed on a simple average daily attendance basis is the Fair Share '=The Clayton district, it might be noted here, also excels in Fund. Each district's appropriation from the Free Textbook parent and citizen participation in district planning. In 1986 Fund is determined by dividing its "membership" for the Clayton district formed a citizens committee, including not month of September by the sum of the September member- only parents but other citizens as well, divided into subcom- ship for all districts in the state. September "membership" mittees, to evaluate the adequacy of school facilities in a is half the sum of the number of resident full-time students building-by-building review. Citizens in the district were so and the full-time equivalent number of part-time students enthusiastic in undertaking this review that the work was who were enrolled in !he public schools of the district on the completed well ahead of schedule. In addition, parent par- last Wednesday in September of the previous year and who ticipation is required in the process of curriculum review. were in attendance one day or more during the preceding ten Each curriculum (e.g., science, language arts) is reviewed school days. This distribution criterion is identical to that for a two-year period on a five-year cycle by committees of for distribution of funds derived from county railroad and parents, teachers, and administrators. The strategy for utilities taxes. Money derived from the School District Trust community involvement followed in the Clayton district does Fund is distributed on the basis of a slightly more complex not involve standing committees. Instead, special commit- criterion. It is distributed monthly to each local school dis- tees are charged with addressing a particular problem in re- trict in the same ratio that the number of "eligible pupils" in lation to a specific deadline. This strategy may allow citi- the district bears to the number of eligible pupils throughout zens to focus their efforts and keep enthusiasm high. the state. "Eligible pupils" is determined by adding the 14The reduction in fiscal variation among the districts is also "membership" in a district to its average daily attendance evident from a comparison of the coefficient of variation and dividing by two. A district's "membership" is half the (CV) for district spending per student and total spending per student. (The coefficient of variation gives the standard de- ling, serially, for the district poverty rate (percentage poor), viation as a percent of the mean.) For the distribution of median household income, and percent white. district spending per student (i.e., spending from district o audiovisual Education Department, Annual Report (Creve sources), CV = 42.69%. For the distribution of total spend- Coeur, MO: Cooperating School Districts of the St. Louis ing per student (i.e, spending from all sources), CV = Suburban Area, June 1987). 27.85%. (The mean, standard deviation, and CV are here "Researchers in this study focused on elementary and secon- computed with respect to districts rather than students.) dary education and thus are not certain to what extent the '=One might expect Wellston to be the most favorably treated sort of relationship described here between the Community district considering its seriously disadvantaged fiscal posi- College district and the Clayton district extends throughout tion and high tax effort. Wellston ranks third, however, in the county. Foundation funding per student, although it does rank first =The source of these and subsequent figures is the Second in federal funding per student. Report to the Federal District Court by the Voluntary In- 'OThe correlation coefficient between the two variables in Fig- terdistrict Coordinating Council (VICC) . ure 6.4 is -0.70. ~2 = 0.49. 23This is essentially the point of view taken by Gary Orfield, who argues for a unified, national desegregation policy in "The correlation coefficient is -0.47. R~ = 0.22. Must We Bus? (Washington, DC: The Brookings Institu- '@Thecorrelation coefficient is -0.75. R2 = 0.57. tion, 1978). Jennifer Hochschild concurs in The New lgThe correlation coefficient between total spending per pupil American Dilemma (Princeton, NJ: Princeton University (from all sources, excluding debt service) and the percent- Press), 1984. age of students passing the eighth grade basic skills test is 24James Madison in The Federalist 46. See Alexander Hamil- t0.54. Using a multiple regression procedure, this relation- ton, John Jay, and Madison, The Federalist (New York: ship continues to be positive and significant when control- The Modern Library, nd.), p. 305. Chapter Eight The Political Economy of Metropolitan St. Louis

INTRODUCTION their efforts and too small to capture economies of This chapter turns from a detailed account of scale. 1 four specific service areas to an analysis of the gen- Economy of scale is a simple economic notion. eral fiscal and economic relationships among gov- For many industrial production processes, the per ernmental units in St. Louis City and County. Three unit cost of output declines as the number of units of broad questions are examined: output increases. This decline commonly reverses itself at some level of output, and one refers to dis- 1. To what extent are there uncaptured econo- economy of scale beyond that level, as per unit cost mies of scale among the small local govern- increases with more output. Scale economies are ments of St. Louis County? often found in the production of capital intensive 2. To what extent does the fiscal variation goods and services where a large fixed cost must be among county municipalities represent an incurred before any output can be obtained-this inequitable distribution of resources? fixed cost can be spread over larger numbers of output units as scale of production increases. Common 3. To what extent has governmental fragmen- public sector examples include water supply and tation in the St. Louis area retarded eco- sewage treatment facilities. nomic growth and development? Most local public goods and services are not of Even though definitive answers cannot be provided this nature, however. They tend to be quite labor to any of these questions on the basis of revenue and intensive. Most of the cost of production of police, expenditure data alone, patterns of fiscal relation- fire, refuse collection, and similar services consists ships can provide clues that point to potential effi- of personnel-related expenditures, and the level of ciency and equity problems. output is roughly proportional to the labor force employed. Empirical analyses of scale economies in the production of most local public services tend to find ECONOMIES OF SCALE few, if any, economies of scale over a range of small One of the evaluative criteria for systems of local and medium-sized production organizations. Dis- government discussed in Chapter One is economies economies, however, tend to be found among the of scale-whether services are produced by units that larger producers.2 The belief in the existence of are of sufficient size to produce them efficiently scale economies ,is, nevertheless, tenacious-such without being so big as to be inefficient. Fragmented economies are frequently assumed in the absence of local government systems are frequently thought to empirical evidence for their presence.3 be inefficient on account of having too many gov- Even labor intensive services, however, may ernments that are too small-too many to coordinate yield limited economies of scale with respect to certain service-components. Police protection, for ex- positive intercept term, and a relatively low coeffi- ample, is highly labor intensive, but certain compo- cient for resident population.6 In the per capita ex- nents of policing are more capital intensive, such as penditure equations, the presence of size economies radio dispatch, information systems, forensics, and would be indicated by a significant, negative coeffi- even training. If a fragmented system of local gov- cient for resident population. ernment is able to achieve coordinated production In neither of the total expenditure nor the ex- of certain service components, while retaining sepa- penditure per capita equations is there any signifi- rate production of other, more labor intensive com- cant evidence of economy or diseconomy effects of ponents, it can theoretically attain a higher level of size on expenditures, once service conditions have efficiency than a large consolidated system that or- been taken into account. This finding is consistent ganizes all of its production on a single large scale. with much of the empirical literature on size effects The presumption that service production and in recent years, which likewise finds little evidence delivery in a highly fragmented metropolitan area of size economies or diseconomies among small to will be uncoordinated is not borne out in this study medium-sized communities.' Rather than size of St. Louis County. The presumption derives economies or diseconomies affecting expenditure largely from the image of a jurisdictional patchwork, variations, the regressions indicate that, as with reinforced, perhaps, by looking at a map. What a revenues, the strongest predictor of these variations map cannot easily show, however, is the organiza- is the extent of commercial and industrial activity in tional overlay created by the multiplicity of local ju- each jurisdiction.8 risdictions for the purpose of coordinating certain Of historical interest, these findings of no signifi- service components. Similar patterns of coordination cant size economies and of expenditures related were found in all four of the service areas studied. strongly to property valuation, are quite similar to In policing, joint production arrangements have the findings of the Metropolitan St. Louis Survey been established for precisely those service components, noted above, that are expected to generate conducted in the late 1950s. The survey was a major economies of scale beyond the scope of a small mu- collaborative effort of scholars at several universities nicipality. In public education, similar arrangements in the St. Louis area, and was intended to provide are found, thus extending the capabilities of small evidence supporting reform of the local government school districts. In fire protection, special districts structure in the city and county. The survey found have been established in those incorporated portions no conclusive evidence with respect to scale econo- of the county where municipalities are too small to mies or diseconomies. The survey concluded its capture potential economies of scale. With regard to analysis by stating: "If economies (of scale) could streets, the use of county government to provide an have been shown, support is implied for consolida- arterial street network, combined with the extensive tion. However, since no major economies or dis- use of private (and to some extent intergovernmen- economies are apparent, this factor could not be tal) contracting by subdivisions and smaller munici- used as an argument for or against consolidation." palities, allows the production of street services to be As in the present analysis, the Survey found that carried out in a manner more likely to be efficient "per capita assessed valuation seemingly is the most from a scale-economy standpoint. important determinant of per capita spending."Q Previous chapters have explored the possibilities of uncaptured economies of scale in the specific ar- FISCAL VARIATION AND eas of police, fire, streets, and education. This sec- FISCAL EQUITY tion investigates the general effect of population size Efficiency grounds alone are generally consid- on total local service expenditures in the municipali- ered to be insufficient as a criterion for judging the ties of St. Louis County, employing regression analy- performance of public institutions. In Chapter One, ses of total expenditures and total expenditures per capita.4 The results are shown in Table 8.1. These equity considerations are also given attention. Does analyses include population size as a predictor vari- fragmentation create a fiscal pattern that signifi- able, along with service condition variables repre- cantly advantages some communities at the expense senting socioeconomic characteristics of the popula- of others? Perhaps more seriously, are certain tion, the presence of commercial and industrial groups, such as racial minorities, easily discrimi- activity in each jurisdiction, and an indicator for nated against in a fragmented system of local gov- whether the community was a point-of-sale city with ernment? These questions are difficult to address respect to the sales tax distribution formula. The because inequality in revenues or expenditures per equations are estimated both with and without trash capita cannot simply be assumed to be inequitable. collection costs paid directly by households.5 One of the virtues of fragmentation is that it allows In the total expenditure equations, the presence different communities to make different fiscal of size economies would be reflected in a significant, choices. The mere presence of high-taxlhigh-service alongside low-taxllow-service municipalities is evi- vantages or disadvantages? If, for example, poor dence only of diversity, not inequity. citizens reside predominantly in communities with While some communities raise small amounts of limited resources for the delivery of public services, revenue because they have small resources, others this would be viewed as greater evidence of inequity do so because they choose not to invest heavily in than if wealthy citizens resided in such communities. public services. Some communities raise large reve- In the latter case, one might assume that the wealthy nues from commercial or industrial activities that had chosen to spend little on public services, where often involve large influxes of nonresident popula- in the former the poor would be assumed unable to tions not reflected in per capita revenue compari- spend more for these services.11 Variations pat- sons and that entail extra service costs. By this argu- terned along racial lines would be similarly suspect ment, to show that variations in the amounts of on equity grounds. revenues available to the different communities in a To investigate the patterning of total per capita local political economy constitute inequity, one must revenues in St. Louis County municipalities, con- demonstrate not simply that variations exist, but that sider the statistics presented in Table 8.2. These sta- the variations are patterned along "suspect" lines.10 tistics are simple correlations (Pearson's r) between total revenues (municipal and fire service) per capita Patterned Inequalities? and indicators of socioeconomic status and other To what extent are variations in revenues avail- community features, and regression coefficients able for local public services patterned in the same from equations including all of the factors simultane- way as indicators of individual citizens' relative ad- ously. Coefficients were computed treating each gov-

Table 8.1 Regression Coefficients for Local Service Expenditures' In County Munlclpalities- 1985 (N =87, population weighted) Total Local Expenditures Total Expenditures per Capita Wlthout Trash2 With Trash Without Trash With' Trash Resident Population

Median Household Income

Percent Poor - 1979

Percent Nonwhite

Percent Over 65 Years Old

Density (number per square mile)

Percent Owner Occupied

Percent Housing Built 1970-80

Value of Commercial Property4

Village

Point-of-Sale

Intercept

R2 Includes expenditures by municipalities, fire protection districts, and (in two of the equations) households. See Table 8.3 *"without trash" and 'with trash" are defined in note 1 on Table 8.3 3Standard error of regression coefficients 4Assessed value in $1,000'~(total or per capita) *-Significant at p < ,001. + -Significant at p < .05. Table8.2 Correlation and Regression Analysis for Total Per Capita Revenues and Community Characteristics in County Municipalities- 1985 (N = 87) Correlation Coefficients Regression Coefficients (Unwelghted) (Weighted1) (Unweighted) (Weighted1) Median Household Income 0.10 0.35* -0.001 0.002 + (.002)2 (.001) Percent Poor - 1979 -0.03 -0.14 7.47 -2.82 (4.40) (2.23) Percent Nonwhite -0.17 -0.16 -0.76 1.28 + (0.98) (0.50) Percent over 65 Years of Age -0.26+ -0.10 -3.48 2.61 (3.22) (1.45) Density (number per square mile) -0.46* -0.49* -0.009 -0.001 (0.009) (.004) Percent Owner Occupied -0.02 -0.17 2.22 + -l.49+ (0.99) (0.55) Percent Housing Built 1970-80 0.26 + 0.44* -0.63 1.75 + (1.09) (0.62) Value of Commercial Property3 0.77* 0.88* 38.12* 39.04* (4.38) (2.79) Village 18.00 -56.61 * (35.57) (25.71) Point-of-Sale 125.64* 44.79 + (38.02) (14.97) Intercept 100.05 235.46* (144.0) (67.12) R2 0.71 0.87 Weighted by municipal population * -Significant at p < .001. 2Standard enor of regression coefficient 3Assessed value per capita in $1.000'~ ernmental unit as a single, equivalent unit of analysis of the housing stock built between 1970 and 1980, and, to compensate for the gross differences in and assessed valuations for commercial-industrial population among these units, recomputed with each property. In the unweighted analysis, two coeffi- governmental unit weighted by its resident popula- cients are relatively strong in absolute value and station.12 tistical significance. These are coefficients for popu- Evidence of patterned inequalities in revenue lation density (negatively related to total per capita availability and income or racial distributions would revenues) and commercial-industrial property valu- consist of strong positive coefficients for median ation (strongly positive in its relationship with total household income (showing that those with higher per capita revenues). Both of these coefficients re- incomes lived in communities with greater availabil- lect the effect of commercial-industrial activity on ity of revenues for public services), and strong nega- total revenues, as such activity is much less prevalent tive coefficients for the percentage of the population in densely populated communities which have less below the poverty level in 1979 and for the percent- land available to accommodate it. age nonwhite. In the weighted correlational analyses reported In the unweighted correlational analyses reported in Table 8.2, no such strong coefficients were in the table, the results are virtually identical for found. The signs of the coefficients are consistent these two factors, and additional ones are signifi- with an hypothesis of patterned inequalities, but cant. Total per capita revenues are higher in com- none of the coefficients approach statistical signifi- munities whose residents have higher household in- cance. Other community characteristics considered comes, and in communities with more recently built are the percentage of the population over 65 years housing. These coefficients reflect the fact that those of age in 1980, population density, percentage of with higher incomes tend to live in newer communi- housing which was owner-occupied, the percentage ties's which tend to raise greater per capita revenues.14 Low income and racial coefficients remain ties results when revenues raised directly from resi- insignificant, however. dential households are considered. These revenues The regression coefficients provide some addi- include municipal real and personal property taxes, tional insights. As in the weighted correlational a portion of the county road and bridge tax that is analysis, median household income and proportion returned to the municipalities, municipal taxes on of newer housing are positively and significantly re- the gross receipts of public utilities, and taxes on real lated to total revenues. So too are assessed valuation and personal property imposed by fire protection of commercial-industrial property and the communi- districts in those municipalities that do not provide ty's status as a point-of-sale city. Village status is this service.le An additional factor to be considered negatively related to total revenues. Somewhat sur- in these computations is the cost per household for prisingly, the regression equation reveals that com- trash collection, and whether this cost is paid by the munities with higher percentages of nonwhite and municipality or directly by the householder.17 elderly residents tend to have higher per capita revenues after adjustment for other predictors of those Residential or household tax burden can be con- revenues. However, this finding, rather than reveal- ceptualized in two different ways: (1) as the dollar ing an advantage for these citizens, has a different amount of taxes paid by the average household in a interpretation (see the discussion of residential tax community and (2) as the percent of average house- burden below). hold income which those taxes represent. Table 8.3 These analyses, examining patterns with respect provides an initial view of these burdens. The left to total per capita revenues, do not reveal strong two columns display average dollar burdens for all patterned inequalities, although there are some ten- municipalities, first without, then with the inclusion dencies in this direction. The strongest predictor of of trash collection costs.18 The right two columns variations in total per capita revenues is not house- display percentage burdens, without, then with, hold income, racial composition, or other socioeco- trash collection costs. With trash collection costs in- nomic characteristics of the population, but rather, cluded, average per household burdens measured in the presence of commercial and industrial activity. dollars ranged from about $60 to $760 in 1985. In Communities with greater amounts of such activity, percentage of average household income, the range particularly if they are point-of-sale communities, was from 0.14 percent to 1.8 percent. The average have substantially higher per capita revenues. These dollar burden was $263, and the average percentage communities tend to be somewhat newer and less burden was 1.O3 percent. densely populated. Their residents have higher The table shows how these burdens vary by type household incomes and tend to be predominantly white. Of these characteristics, population density is of local government and by status with respect to the by far the strongest (negative) predictor of commer- sales tax distribution. Residents of villages in the cial and industrial activity.ls However, these data county have, on average, the lowest tax burdens by speak to total revenues, and not to revenues raised either dollar or percentage measures. This is true directly from residents. whether trash collection costs are included or not. Tax burdens are highest in home rule cities, again by Residential Tax Burden either measure and with or without trash costs. Bur- A somewhat different view of revenue variations dens in fourth class cities are higher in dollar terms and patterned inequalities across county municipali- than in third class cities, but lower in percentage

Table 8.3 Tax Burdens on Residential Property- County Municipalities Average Dollar Burden on an Burden as Percent of Occupied Home Average income Without1 With Without With Trash Costs Trash Costs Trash Costs Trash Costs All Municipalities $21 5 $263 0.84% 1.03% Villages 136 167 0.62 0.76 Fourth Class Cities 21 4 25 1 0.77 0.91 Third Class Cities 187 196 0.92 0.96 Home Rule Cities 229 294 0.92 1 .18 Point-of-Sale Cities 22 1 249 0.80 0.90 Pool Cities 206 280 0.91 1.23 'Without trash" burdens include utility taxes and property taxes of municipalities and fire districts, and do not distinguish whether trash collection is covered by general fund expenditures or by households themselves. 'Wlth trash" burdens include these taxes plus the estimated cost to the average household for trash collection in those municipalities where it is not paid from general revenues. Table 8.4 Regression Coefficients for Residential Tax Burdens in County Munlcipalities- 1985 (N = 87, population weighted) Average Burden Per Burden as a Percent of Occupled Household Household income Without Trash With Trash Without Trash With Trash Collection1 Collection Collection Collection Median Household Income

Percent Poor - 1979

Percent Nonwhite

Percent over 65 Years of Age

Density (number per square mile)

Percent Owner Occupied

Percent Housing Built 1970-80

Value of Commercial Property3

Village

Point of Sale

Intercept

R2 'For the distinction between 'without trash" and *with trash" burdens, see note 1 on the preceding table. 2Standard error of regression coefficients 3Assessed value per capita in $1.000'~ *-Significant at p < .001. + -Significant at p c .05. terms, as average incomes are higher in fourth class Table 8.4 repeats the regression analyses of pat- cities. These correlations should not be interpreted terned inequalities, but with household tax burdens as showing that governmental form affects residen- as the dependent variables. Income effects are not tial tax burden. More likely, citizens choose a gov- strong-households in wealthier communities pay ernmental form in accordance with their preferences more in dollar terms, but there is no effect when for local services. Thus villages are chosen as a mu- measured as a percentage of income. Burdens are nicipal form that facilitates the provision of relatively somewhat higher in newer communities when the ef- low service levels, while home-rule cities are consid- fects of the other variables are controlled. Tax bur- ered more appropriate for the provision of high serv- dens are lower in villages, principally because of ice levels. these communities' limited use of the utilities tax. The last two rows of Table 8.3 contrast burdens Tax burdens are also lower in communities with for point-of-sale and pool cities. If trash costs are higher percentages of owner-occupied housing units. not included, households in point-of-sale cities pay Homeowners are likely to be more sensitive to slightly higher dollar burdens but lower percentage household tax burdens than are renters, as the lat- burdens as average household incomes are higher ter's burdens are generally subsumed within the there. With trash collection included, burdens are monthly rent payment, and thus homeowners may higher on residents of pool cities by either measure. exert more pressure on officials to hold down tax Point-of-sale cities are more likely to finance trash burdens. collection out of general revenues, while pool cities Unlike the equations predicting total revenues are more likely to require households to pay directly (Table 8.2), these equations for household tax bur- for this service.18 dens show relatively little effect from commercial and industrial property valuation. The coefficients case targeting additional revenues on municipalities for this variable are negative, indicating lower bur- that are relatively disadvantaged. Proposals involving dens as the value of such property in a jurisdiction both of these methods are currently under discus- increases, but the coefficients are weak, and lacking sion in St. Louis County. If St. Louis County resi- in statistical significance in all but one equation. dents were to pursue such redistribution, the dollar Looked at in this fashion, it appears that tax bur- amount required to do so would not be large. The dens on residential property are relatively unaffected average household burden, with trash collection in- by the presence or absence of commercial and included, as a percentage of average household industrial activities in a jurisdiction. Although this is come in all municipalities is 1.03 percent (see Table clearly untrue in specific cases,20 the countywide 8.3). If citizens in the county agreed to supplement pattern shows no strong relationship. The influence the revenues of all communities where the house- of point-of-sale status is insignificant if trash collec- hold percentage burden exceeded this average tion arrangements are not considered, but is strong value, providing those communities with additional when trash collection costs are included, again re- revenue that would enable them to reduce their av- flecting the propensity of point-of-sale cities to fi- erage burdens to 1.03 percent, the total additional nance this service out of general revenue^.^' revenue required would be roughly $8 to $ 10 million In terms of patterned inequalities that would annually.22 At the high end, this estimate represents point to inequities in the local political economy, the less than 12 percent of the sales tax revenues col- coefficients for the percentage of a community's lected in the county in 1985. Thus, a small addition population that is nonwhite and that which is over to that tax, or a change in its distribution formula to 65 years of age require attention. Both coefficients, target a portion of existing revenues to disadvan- in both forms of analysis, are positive and significant taged communities, could easily and significantly re- in the statistical sense. This means that, after adjust- duce this inequality in household tax burdens across ment for income differences and other community the municipalities.23 characteristics, residents of communities with higher proportions of nonwhite and elderly citizens pay The Case of School Districts higher taxes in dollar terms and higher proportions Revenues of the regular school districts in the of their incomes in taxes than do residents of com- City and County of St. Louis during 1985 totaled munities with fewer nonwhite and elderly citizens. $649 million, more than the total revenues of gen- Their communities, then, may find themselves in fis- eral purpose governments.24 These revenues repre- cal difficulties similar to those faced by some older, sented about $3,850 per student in average daily at- central cities-an aging and nonwhite population tendance. The Special School District in St. Louis with public service needs that can be met only by County had revenues of $62 million. relatively higher tax burdens directly on themselves The range of variation in the revenues of regular and their residential property. school districts is much less than that of general pur- Among the cities with relatively high household pose governments. The least funded school district burdens are some whose residents are quite poor- raised $2,5 14 per pupil in 1984-85 and the highest Kinloch and Wellston, for example. Others, such as ftinded raised $6,889. This yields a variation of less Olivette, are comprised of more well-to-do residents than 3 to 1, contrasting with the 17-to-1 variation in who apparently are willing to pay higher burdens for general purpose revenues per capita. Although the the services they receive. Some, such as University City of St. Louis was among the most advantaged City, Normandy, or Northwoods, combine a tradi- general purpose governments in revenue per capita tion of high service levels with populations whose av- terms, its school district is only somewhat above average incomes have not kept pace with those in erage in revenues per student in average daily atten- other parts of the county. Among the cities with dance. Eight of the 23 county regular school districts relatively low tax burdens are those with well-to-do had revenues higher than those of the city district. residents-Frontenac, Huntleigh, and Westwood- Table 8.5 displays the variation in per pupil and others whose residents have incomes well below revenues found among the school districts in St. the county average, such as Breckenridge Hills or Louis City and County. There is no particular pat- Hillsdale. There is no single, simple explanation for terning of per pupil revenues by type of district. As the complex variations in tax burdens found in the the lower portion of the Table shows, the seven dis- county. tricts in St. Louis County with the lowest revenues Ways to reduce the inequalities in tax burdens per pupil receive, on average, about 68 percent of across the municipalities in the county include rais- the revenues of the eight most advantaged county ing new revenues countywide for this purpose, or re- districts. School district revenues in Missouri have distributing revenues from existing sources, in either not been equalized; there is still significant variation Table 8.5 School Revenue Variations in St. Louis Percent of Average Percent of Average Range of Dally Attendance Total Revenues Revenues (ADA) Revenues Per ADA Per ADA City of St. Louis 30.4% 31.8% $4,030 St. Louis County Districts: Common (15) 23.2 22.7 3,757 Reorganized (7) 35.0 33.8 3,720 Consolidated (1) 11.4 11.7 3,936 Total County Districts:' 69.6 68.2 3,767 County Special School District - - 15,022 School Revenue Variations in County Districts Percent of Percent of Average Range of Number of Average Daily Total Revenues Revenues Districts1 Attendance Revenues Per ADA Per ADA Seven Lowest in Revenues Per ADA 21.3 17.5 3,097 Next Four 30.3 28.2 3,506 Next Four 24.9 26.1 3,949 Eight Highest in Revenues Per ADA 23.5 28.2 4,521 Not including Special School District of St. Louis County.

from district to district. School revenues, however, examine the extent of patterned inequalities in edu- exhibit much less variation than do revenues of gen- cation funding. eral purpose local governments.25 In Table 8.6 the question of patterned inequali- As argued above with respect to variations in ties is pursued with respect to low income and mi- general purpose revenues, variations in school reve- nority populations of school districts, as discussed nues are not prima facie evidence of inequities. above with respect to municipalities and fire dis- Residents of school districts, like residents of mu- tricts. Are poor andlor minority students more likely nicipalities (and fire districts), can and do make to reside in school districts with lower revenues? In- choices about how much educational services they equalities are measured by simple correlations want for their children and how much they are will- (Pearson's r) between total and district revenues per ing to pay. In education, however, there are greater student in average daily attendance and the follow- sources of extrajurisdictional funding, particularly ing variables: median household income, percentage from the state and, to a lesser extent, from the fed- of the population below the poverty level in 1979, eral government. These extrajurisdictional sources and percentage nonwhite. attempt, at least in part, to reduce variations in reve- Somewhat surprising is the fact that the data in- nues across districts.2'3 Given this, it is interesting to dicate greater patterned inequality among school

Table 8.6 Correlation Analysis for Total and District Revenues Per Student in Average Daily Attendance - 1984-85 Total Revenue Per ADA District Revenuer Per ADA City and County County City and County County Districts Districts Districts Districts Weighted by ADA? No Yes Yes* No Yes Yes2 Median Household Income 0.29 0.04 0.35 0.42+ 0.61 * 0.56 + Percent Poor - 1979 -0.16 0.13 -0.19 -0.33 -0.48+ -0.41 Percent Nonwhite -0.29 -0.12 -0.28 -0.43+ -0.57* -0.48+ Per Student Assessed Value 0.89* 0.62* 0.90* 0.94* 0.92* 0.92* 'District revenues without Proposition C funds. * -Significant at p < ,001. 2Unweighted coefficients for County Districts are virtually identical . + -Significant at p c .05 districts than was found among general purpose gov- 2) Most of the municipalities that have large ernments. That is, in both the unweighted and amounts of commercial and industrial prop- weighted analyses shown in the Table, the total reve- erty are able to retain the revenues gener- nue per student coefficients for income, poverty, ated within their borders from the sales tax, and race are larger than those found for total reve- because they are point-of-sale cities;31 and nue per capita among general purpose governments 3) The presence of significant commercial and (compare Table 8.2). District revenue per capita co- industrial activity generates service demands efficients are even stronger. Both sets indicate ten- over and above those found in predomi- dencies for greater revenues per student in districts nantly residential communities, and servic- with higher median household incomes and lower ing these demands requires higher expendi- proportions of poor and nonwhite residents. These tures. results are surprising because state and federal governments have actively redistributed funds to reduce The sales tax controversy in St. Louis County inequalities in school revenues, but have not, for the revolves around the interplay of factors 2 and 3. Are most part, tried to do so with municipal or fire dis- the higher sales tax (and other) revenues that accrue trict revenues.27 to point-of-sale cities windfall profits, attributable to For comparison with findings for general pur- locational happenstance, or are they equitable compose governments, correlations were also computed pensation for the added costs of providing services between revenues per student in average daily atten- to support commercial and industrial activity, and dance and property valuation per student in the dis- for willingness to accept the loss of amenities that trict~.*~These coefficients are very strong, indeed results from accommodating such activity within a even stronger than those computed for general pur- jurisdiction? The answer seems to be a little of both. pose governments. School districts have become less Many of the point-of-sale municipalities have dependent on local property taxes as state aid has locational or other advantages-proximity to the air- increased, but they remain much more reliant on port, access to key traffic arteries, availability of un- these taxes than do general purpose governments. developed land, status as the county seat-that are This is reflected in the strong coefficients for prop- not shared by pool municipalities. These advantages erty valuation. Variations in property valuation per make them attractive locations for commercial and district are strongly associated with revenue vari- industrial activity. As such activity locates within ations per district because a considerable portion of their borders, multiplier effects may occur. Many extrajurisdictional revenues is distributed on a per types of businesses prefer to locate near similar or pupil basis rather than in a more redistributive fash- related businesses, and the revenues available to ion. point of sale cities may allow them to offer attractive When total and district revenues are regressed service-tax-cost packages to new or relocating firms. on median household income, percentage poor, per- On the other hand, there are clearly costs in- centage nonwhite, and property value per student, volved in providing services to commercial and in- only property value per student has a significant and dustrial activity. A municipality that is the locus for very strong positive relationship with revenues.29 such activity will often have a large nonresident The patterned inequalities shown result from the lo- population requiring services. Point-of-sale cities on cation of poor and nonwhite populations in school average have 279 employees in the retail, wholesale, districts that have relatively lower assessed property and service industry sectors for every 1,000 resi- valuations. State and federal aid, as currently dis- dents. Pool cities on average have 74 such employ- tributed, reduces these inequalities-compare the ees per 1,000 residents.32 No data on numbers of district revenue coefficients to total revenue coeffi- nonresident customers are available, but it seems cients-but does not eliminate them. obvious that there are many more in point-of-sale cities than in pool cities. In addition to the services The Sales Tax Controversy Revisited required by this extra population, negative conges- The presence of and tion effects can occur. One indicator of relative dis- property in a municipality, indexed by its assessed amenity for point-of-sale cities is their crime rate. valuation, is by far the strongest predictor of munici- ~h,average crime rate in point-of-sale cities in 1985 pal expenditures in St. Louis County* This results was 51 per 1,000 residents, nearly 40 percent higher from a combination of three factors: than the rate in pool cities of 37 per 1,000.s A re- 1) Commercial and industrial property gener- luctance to accept such disamenities may help to ex- ates significant revenues in the form of sales plain why some pool cities that would otherwise be taxes, property taxes, licenses and permits, attractive locations for commercial and industrial ac- and utilities gross receipts taxes? tivity have been unwilling to accept it.34 The most difficult aspect of the sales tax contro- Per capita revenues from property taxes im- versy from an analytic perspective is disentangling posed by municipalities and fire districts, the revenue-expenditure effects of commercial and 1985. industrial activity. That is, to what extent do point- of-sale cities spend more for public services because Per capita revenues from the sales tax, such activity requires higher expenditures, and to 1985. what extent do they spend more because such activ- Per capita revenues from gross receipts taxes ity provides higher revenues, which are then avail- on the revenues of public utilities, 1985. able to be spent? This disentanglement is particularly difficult because of the near equality of revenues Other Factors: and expenditures for most municipal governments. Average household income, 1980. Still, an attempt can be made. Percentage of the resident population over The method of disentanglement used is one that 65 years of age, 1980. was developed as part of a program of state aid to offset fiscal disparities among municipalities in Mas- Rate of population change (estimated 1984 sachusetts.35 Per capita expenditures in each juris- population divided by 1980 population). diction are regressed on three sets of factors. The Rate of population change squared. sets represent: The cost factors and other factors were chosen for Environmental costs-factors largely outside their similarity to those used in Massachusetts. The the control of public officials, including such revenue factors are different from the Massachusetts things as large, nonresident daytime popula- analysis because of the difference in revenue struc- tions, levels of crime, age of a jurisdiction's tures between Massachusetts and Missouri.38 infrastructure, and population characteris- Table 8.7 displays the mean values of each of tics; the factors in all municipalities, in pool cities and in Community resources-the type and size of point-of-sale cities. Point-of-sale cities, on average, revenue pools that can be tapped by a juris- have higher expenditures, raise more revenues from diction; and each source, have residents with higher average in- Other factors-including resident character- comes, and have experienced greater population istics that serve as surrogates for their public growth than have pool cities. In terms of the cost service preferences, and variables measuring factors considered, point-of-sale cities have more population change in recent years. employees and more crime, lower population density, newer housing, and somewhat fewer poor resi- For each community, a predicted cost is computed dents than do pool cities. using the regression results, where this cost is "what Table 8.8 presents the estimated cost model for each community would have spent if it had average county municipalities.39 It is no surprise that total resources, average demand (preferences for public expenditures per capita, including those by munici- services), and average population change, but re- pal government, fire district, and householder tained its own values for the cost variables."36 The (where paying directly for trash collection) are predicted cost is then converted "into a cost index strongly influenced by available revenues. Among by dividing each prediction by the mean per capita the other, noncost factors, average household in- expenditure for all . . . cities."S7 come is positively related to total expenditures, and The variables used for computation of a cost in- the percentage of population over 65 years of age dex for St. Louis County municipalities are: shows a negative relationship to expenditures40 Cost Factors: Population growth is positively related to increased Number of employees in retail, wholesale, per capita expenditures, but at a decreasing rate as and service industries per capita, 1982. growth is larger. Among the cost factors, the number of retail, Number of Pan 1 crimes per 1,000 resi- wholesale, and service industry employees per 1,000 dents, 1985. residents of a jurisdiction is strongly related to local Population density (residents per square service costs. This variable measures the effect of mile). such employees plus the effect of customers who fre- Percentage of housing built before 1940. quent the enterprises where they are ernpl~yed.~' Inexplicably, the number of crimes per 1,000 resi- Percentage of the population with incomes dents is negatively related to service costs while, as below the poverty level, 1979. expected, the percentage of residents with incomes Revenue Factors: below the poverty level has a positive relationship Table 8.7 Variables Used for Computing Cost Index of County Municipalities- 1985 (N = 71, population weighted) Mean Value In: All Pool Point of Cities Cities Sale Cities Variable (71) (41) (30) Total Expenditure Per Capita $338 $254 $406 Cost Factors: Employees Per 1,000 Residents Part I Crimes Per 1,000 Residents Residents Per Square Mile Percent of Housing Built Before 1940 Percent Below Poverty - 1979 Revenue Factors: Property Taxes Per Capita Sales Taxes Per Capita , Utilities Taxes Per Capita Other Factors Average Household income- 1980 Percent over 65 Years of Age - 1980 Population Change 1980-84 Population Change Squared with costs. Neither reaches statistical significance. creased, cost. In the county municipalities, however, Population density and percentage of housing built these two factors pick up additional effects of com- before 1940 have the opposite sign, negative, from mercial-industrial costs not captured by the employ- what one would ordinarily expect. Both are usually ment factor. That is, commercial and industrial ac- found to be associated with increased, not de- tivity is less prevalent in older, densely populated

Table 8.8 Total Local Cost1 Model for County Municipalities- 1985 (N = 71, population weighted) Cost Factors Estimated Coefficient Employees Per 1,000 Residents 239 + Part I Crimes Per 1,000 Residents -21 1 Residents Per Square Mile -.003 Percent of Housing Built before 1940 -.287 Percent below Poverty - 1979 1.70 Revenue Factors Property Taxes Per Capita 1.lo* Sales Taxes Per Capita 0.58 + Utilities Taxes Per Capita 1.43* Other Factors Average Household Income - 1980 BOO3 Percent over 65 Years of Age - 1980 -1.45 Population Change 1980-84 995 Population Change Squared -345 'Including municipal, fire protection district, and trash collection expenditures. -Significant at p < ,001. + -Significant at p < .05. Table 8.9 Actual and Predicted Service Costs In County Municipalities by Sales Tax Status- 1985 (N = 71, population weighted) Mean Value in: All Pool Point-of-Sale Cities Cities Cities (71) (41) (30) Employment Cost Per Capita $ 45 $ 18 $ 67 Sum of Cost Factors 26 -2 49 Predicted Cost of Services Per Capita 338 31 0 361 Cost Index 1.00 0.92 1.O7 Actual Expenditure Per Capita 338 254 406 Indexed Actual Expenditure Per Capita 338 277 375

cities, and the negative influence of these factors on Based on predicted costs related to employment expenditures is a result of this fact. within municipal borders, the per capita service costs A predicted cost was computed for each munici- of the point-of-sale cities are estimated to be $49 pality using the coefficients in Table 8.8 multiplied higher than those in pool cities.45 Other cost factors by the average countywide values for the revenue add slightly to this difference, yielding a total differ- and other factors, and by the municipality's own val- ence of $5 1 per capita in predicted per capita serv- ues for the cost factors. This cost is what the munici- ice costs between pool and point-of-sale cities. The pality's per capita expenditure would be if it had av- average cost index for point-of-sale cities is 1.07, in- erage per capita revenues and average values of dicating 7 percent higher costs than the average for household income, older residents, and population all 7 1 cities, while for pool cities the average index is change, but retained its present values for the cost 0.92, indicating about 8 percent lower than overall factors. average costs. Several point-of-sale cities which are Commercial and industrial activity is a net reve- significant employment centers46 have cost indices nue generator in St. Louis County. For the 71 mu- well above the 1.07 average-Clayton is the highest nicipalities considered in this section, revenues from at 1.37, followed by Creve Coeur, 1.31; Frontenac, the sales tax, property taxes and gross receipts taxes 1.29; Brentwood, 1.22; and Des Peres, 1.20. Pool on commercial property, and from licenses and per- cities with low cost indices are predominantly small, mits totaled nearly $94 million in 1985. The esti- residential communities-Glen Echo Park with an mated cost of providing services to this activity was index of 0.79 is the lowest, and Bella Villa, Mack- $24.5 million.42 On average, each dollar spent pro- enzie, Pasadena Hills and Park, St. George, Syca- viding service was replaced by nearly $4.00 of reve- more Hills, Uplands Park, and Wilbur Park all have nue.43 The rate of return, the ratio of revenue from indices below 0.85. Three of the largest pool cities- commercial sources to the predicted cost of provid- Ferguson, Florissant, and Webster Groves-have ing services to them, is higher in pool cities than in cost indices equal to or slightly above the average for point-of-sale cities. Many pool cities incur little or all pool cities, but all have indices below 1.00, indi- no cost, having little or no commercial-industrial ac- cating that their predicted costs of providing services tivity, but obtain commercial source revenue from are slightly below the county average. University the sales tax pool-revenue collected in other mu- City's cost index is 0.89, below the pool average.47 nicipalities in the pool and in unincorporated parts Thus, as representatives of point-of-sale cities have of the county.44 However, the net revenue gain per argued, their service costs are higher than those of capita from commercial-industrial activity is substan- pool cities. tially higher in most point-of-sale cities because of On the other hand, point-of-sale cities obtain the much higher volume of such activity in them. revenues from their accommodation of commercial Table 8.9 displays the employment cost factors, and industrial activity which appear to significantly the sum of cost factors, predicted costs and cost in- outweigh the extra costs they incur. While predicted dices, actual total expenditures, and expenditures service costs in point-of-sale cities are $5 1 per capita deflated by the cost index for the average municipal- higher than in pool cities, actual per capita expendi- ity and the average pool and point-of-sale city in the tures are $152 higher. After discounting actual ex- county. These data provide a basis for commentary penditures by the cost index, adjusting them for dif- on the sales tax controversy in St. Louis County. ferences in service conditions, the indexed total per capita expenditures in point-of-sale cities are still 1985-a total expenditure per household of $3.60 $98 greater than in pool cities. for every dollar the household paid in direct burden. A significant portion of this positive difference In 38 of the 71 cities, the ratio fell below this figure. between actual and predicted expenditures is the To augment their revenues so as to reach this ratio larger per capita sales tax receipts in point-of-sale would require an additional $34 million over and cities, although differences in property taxes, gross above their revenues in 1985. If one discounts the receipts revenues, and other revenue sources con- revenue data by the cost index developed above, re- tribute to0.~8These estimates suggest rather strongly ducing the amount of subsidy required for munici- that residents of point-of-sale cities benefit from palities with low values on the index and increasing higher expenditures for public services in their juris- it for those with a high index, the average ratio dictions as a result of these higher receipts, and do drops to 3.5, the number of disadvantaged cities to so even after adjustment for the extra service costs 34, and the total additional revenue required to $28 incurred by those jurisdictions.49 Since, as noted million.51 earlier, per household tax burdens of point-of-sale The second definition of disadvantage draws city residents tend to be lower than those of pool city more directly on the cost model developed above, residents, there is a lack of full fiscal equivalence. and finds a city to be disadvantaged if its actual total Point-of-sale cities are able to export a portion of revenues fall short of the cost of providing se~ces their local taxes. They are able to spend more for predicted by the model.52 By this definition, more local services while maintaining lower per household cities are disadvantaged-45 of the 71 instead of 34 tax burdens because of taxes imposed on nonresi- with the first definition (and cost discounting)-but dents directly by the sales tax and indirectly through the total amount required to remove the disadvan- property and gross receipts taxes on commercial-in- tage is less, approximately $19.5 million.53 The pri- dustrial enterprises, whose owners and managers mary reason for a reduction in total amount re- may be able to pass those taxes along to nonresi- quired is that this second definition does not dents as higher prices and/or lower wages.50 recognize as disadvantaged several cities that have relatively high household tax burdens as a percent- Reducing Fiscal Disparities age of average incomes, but that also spend more As noted earlier, revenue and expenditure vari- than their predicted costs. These municipalities, in- ations per se need not demonstrate inequities, at cluding Clayton, Ladue, University City, and Web- least not to those who expect citizens to choose dif- ster Groves among the larger ones, can be viewed as ferent tax and service bundles if permitted to do so. those where residents prefer higher service levels However, the analysis of variations in household tax and have been willing to incur higher tax burdens in burdens, together with that of the distinction be- order to pay for them. tween point-of-sale and pool cities and the effect of The additional revenues that would be needed this difference on revenues and costs, suggest that a to reduce fiscal disparities by supplementing the part of the fiscal variation in St. Louis County is pat- revenues of disadvantaged communities represent terned along suspect lines. This part of the variation between 10 percent (by definition two) and 15 per- reflects more than simply different choices by the cent (by definition one) of total revenues in the 71 different collectivities found there. Thus, one might cities in 1985. The amounts required are roughly wish to reduce these disparities, either by raising comparable to what could be raised by a 0.22 to new revenues and targeting them to disadvantaged 0.32 cent addition to the countywide sales tax, a communities, or by redistributing existing revenues range that encompasses recent proposals for an in- in favor of such communities. Assuming that it is crease in that tax. If that increase were earmarked likely to be easier to target new revenues than to for reducing fiscal inequities, most of those inequi- redistribute existing ones, two estimates of the cost ties could be eliminated. of doing so are provided. The estimates differ in the way that disadvantage ECONOMIC DEVELOPMENT is defined. One definition is that a community is dis- "Metropolitan problem solving" is another eval- advantaged if its total revenues per household, di- uative criterion for a system of local government in- vided by its average household tax burden, yields a troduced in Chapter One. Assessing performance on ratio below the county average. This definition as- this criterion requires investigating whether problems sumes that the household tax burden ought to yield that are metropolitanwide can be addressed on a revenues per household in each municipality at least metropolitan basis. Regional economic development as large as it does for the average householder frequently is cited as such a problem, and is said to countywide. The average ratio for the 71 municipali- require decisionmaking arrangements that can act ties considered in the previous section was 3.6 in decisively and comprehensively for a metropolitan Table 8.10 Employment Growth in St. Louis and Comparison Metropolitan City-Counties - 1970-85 Employment Change: Employment in: 1970-85 1970 1985 Jobs Percent St. Louis City-County 613,540 734,883 + 121,343 + 19.8% Atlanta-Fulton County 372,692 459,524 + 86,832 + 23.3 Baltimore City-County 518,187 576,935 + 58,748 +11.3 Dallas-Ft. Worth-Dallas-Tarrant Counties 817,307 1,496,215 + 678,908 + 83.0 Denver-Denver County 253,696 352,714 + 99,018 + 39.0 Indianapolis-Marion County 31 8,683 380,398 +61,715 + 19.4 Kansas CityJackson County 307,836 328,760 + 20,924 + 6.8 Memphis-Shelby County 221,767 320,563 + 98,796 + 44.6 Pittsburgh-Allegheny County 554,125 572,764 + 18,639 + 3.4 SOURCE: U.S. Department af Commerce. Bureau of the Census. County Business Patterns. CBP-8527 and previous year's reports. region if the region is to be competitive economically Area (MSA). Defining the comparison regions in with other metropolitan areas.54 this fashion affords the best equivalence to the St. Louis City-County portion of the St. Louis MSA. A Comparative Perspective on the The comparison regions in the South and St. Louis Development Experience West-Atlanta, Dallas-Fort Worth, Denver, and A common indicator of economic growth or de- Memphis-had larger percentage gains in employ- cline in an area is the change over time of total em- ment from 1970 to 1985 than did St. Louis. At- ployment. Job opportunities increase as an area's lanta's gain was slightly larger; gains in Dallas, Den- economy grows, and decrease as it declines. New ver, and Memphis were much larger. The job creation in St. Louis has lagged behind the na- comparison regions in the Northeast and Midwest- tional average during the past 15 years. That aver- Baltimore, Indianapolis, Kansas City, and Pitts- age, however, includes boom areas of the South and burgh-added jobs at a slower or, at best, the same West, stagnant or declining areas in other parts of rate as St. Louis, and added many fewer jobs. St. the United States, and central city, suburban, and Louis did not keep pace with regions in the South rural employment changes. A better comparison of and West, but neither did any of the so-called rust St. Louis' experience is with that of other metropoli- belt comparison regions (or most others in the tan regions. Northeast and Midwest, for that matter). How employment change in St. Louis has com- Memphis-Shelby County, which has entrepre- pared with changes in other metropolitan areas de- neurial, countywide government leadership, out-per- pends on which other metropolitan areas one formed St. Louis City and County in employment chooses for comparison purposes, and how those ar- increase from 1970 to 1985. Indianapolis-Marion eas are defined. Table 8.10 provides a comparison County, which has a consolidated city-county Coun- of combined St. Louis City-County employment cil, a mayor elected countywide, a Department of change with that in six metropolitan city-counties Metropolitan Development with countywide powers, identified by the St. Louis Regional Commerce and and a nationally acclaimed public-private partner- Growth Association55 as significant competitors for ship for economic development, performed about St. Louis-Atlanta, Dallas-Ft. Worth, Denver, Indi- the same as St. Louis City and County. Baltimore- anapolis, Kansas City, and Memphis-and two addi- Baltimore County has had entrepreneurial govern- tional city-counties-Baltimore and Pittsburgh-that mental leadership (one of the most successful in the have the same population size as St. Louis and St. nation in securing UDAG funds) and has only two Louis County, but are polar cases with respect to local governments; yet it performed less well than St. governmental fragmentation. Baltimore City-County Louis City and County. Pittsburgh-Allegheny County has 2 general purpose local governments; Pittsburgh- has had entrepreneurial public and private sector Allegheny County has 131. The comparisons are leadership, has restructured its economy from steel made for the central city (s) and county (s) of each of production to service industries, has more local gov- the regions only, not including surrounding counties ernments than St. Louis, and performed the least that are part of each region's Metropolitan Statistical well of the nine regions. Perhaps most interesting is Table 8.11 Patterns of Employment in St. Louis City and County- 1970-85 0State of City County Combined Missouri 1970 376,113 237,427 61 3,540 1,393,751 1975 303,250 302,048 605,298 1,391,857 1980 286,896 408,986 695,882 1,668,510 1985 263,798 471,085 734,883 1,772,618 Change 1970-85 -1 12,315 + 233,658 + 121,343 + 378,867 Percent Increase or Decrease 1970-85 -29.9 + 98.4 + 19.8 + 27.2 SOURCE: US. Department of Commerce. Bureau of the Census, County Business Patterns. CBP-85-27 and previous year's reports. the fact that all four-Memphis, Indianapolis, Balti- percent. The city's job loss was about half of the more, and Pittsburgh-have national images of eco- county's job gain.58 During this period, the city's nomic vitality, while St. Louis' image is generally not population declined by nearly one-third, a loss of one of great vitality. Image, apparently, is not always some 193,000 residents. Employment totals in the correlated with reality .56 city and county were approximately equal in 1975, Table 8.11 tracks total employment from 1970 but 10 years later the county had nearly 80 percent to 1985 for the City of St. Louis, St. Louis County, more jobs than the city. When national media pre- the combined city-county region, and the State of sent an image of St. Louis as an area in decline, that Missouri. The combined city-county region has image has reference to economic changes in the lagged behind the state in percentage of employment city.59 growth, adding about 20 percent to its employment What has occurred in St. Louis City and County base while the growth statewide over the 15-year pe- is a phenomenon repeated in a number of older, riod was 27 percent. On the other hand, the "winter cities" in America, particularly those, like 120,000 plus employment increase in the combined the City of St. Louis, which are prevented from in- city-county represented 32 percent of the total em- creasing their size through annexation.60 Most eco- ployment increase for the state, although the region nomic growth has occurred in their suburbs, with the has only 28 percent of the state's population. The central cities growing very slowly or declining. Be- city-county region has traditionally been a major tween 1970 and 1980, all of the larger central cities employment center in the state, accounting for 41 in the Northeast and Midwest (except Indianapolis) percent of all jobs statewide in 1985. Given its high lost population and jobs, while all of their suburban employment base in 1970, it would have been diffi- areas gained. However, St. Louis City's percentage cult for St. Louis to match the statewide percentage population decline was the greatest among these cit- increase. Compared to its cross-state competitor, ies, and its job loss among the worst. St. Louis Kansas City-Jackson County, whose employment in- County's population remained about the same, but crease from 1970-85 was only 7 percent, St. Louis its employment growth was among the very highest. City and County's combined growth is not unimpres- One reason that St. Louis has a poor image while sive. However, the data in Table 8.11 also docu- Baltimore, Indianapolis, and Pittsburgh do not, even ment the disparity in economic development be- though the economic performance of these three ar- tween the City of St. Louis and St. Louis County, eas' core city-counties is no better and, for two, and it is here that the region's image problem, to- worse than St. Louis City and County, may be that gether with real economic difficulties, are found. their central cities have experienced job growth, not Employment in St. Louis County almost doubled decline.61 from 1970 to 1985, while the county's population Employment change in St. Louis County has ex- grew only slightly, and the jobs added there repre- hibited a similar, although generally less dramatic sented 62 percent of all jobs added in Missouri dur- pattern. Newer cities, particularly those located out- ing the period. St. Louis County is recognized as one side of Interstate 270, and unincorporated portions of the major growth centers in the United States, of the county in that same area, accounted for the and its growth is projected to continue for at least bulk of the county's employment growth in retail the next 15 years.S7 Employment in the City of St. and wholesale trade between 1972 and 1982. Older, Louis, on the other hand, declined by almost 30 inner cities in the county generally declined in retail and wholesale employment in these years. Employ- have significantly fewer local governments. Very few ment in service industries increased in most munici- large urban counties in the Northeast and Midwest palities and in the unincorporated county, but per- even approached St. Louis County's employment centage increases were greatest in newer, outer cities growth, although virtually all have fewer local gov- and in the unincorporated county.@ Several older ernments. An argument that governmental fragmen- cities-Jennings, Northwoods, Pagedale, Pinelawn, tation limits (or, the other side of the coin, en- and Wellston-suffered percentage employment hances) a region's potential for overall growth in losses in retail, wholesale, and service industries at employment must be given the old Scottish verdict, higher rates than those of the City of St. Louis. "not proven. " Since 1970 some 24 million square feet of new One might still advance an argument that frag- office space has been built in the county. More than mentation affects the locational pattern of employ- one half of that new space was built in the uninc- ment growth and decline, but the argument must be orporated county, and most of the remainder in carefully drawn to be consistent with the data. newer cities or in Clayton, the county seat. Assessed Growth in St. Louis County has been stronger in the valuation in the county increased by 85 percent unincorporated, less fragmented part of the county from 1970 to 1984. Fifty-eight percent of the in- than it has been in the incorporated, more frag- crease occurred in unincorporated areas and 42 per- mented part. On the other hand, decline has been cent in incorporated parts of the c0unty.W Although stronger in the unfragmented City of St. Louis than nearly 75 percent of all jobs in the county are cur- it has been in all but a small portion of the frag- rently found in the incorporated areas inside of mented, incorporated part of the county. Major em- 1-270, local employment projections through 2000 ployment creation and major employment loss both show significant growth outside of 1270, with em- occurred in parts of the city-county region where ployment remaining roughly constant inside that large populations and land areas were under the ju- b0rder.w Economic growth in St. Louis, having risdiction of a single general purpose government, shifted from the city to St. Louis County, is shifting where fragmentation was not present. Employment within the county from the older, inner suburbs to creation was slower in the fragmented, incorporated newer, outer ones and to the unincorporated part of the county than in the less fragmented, un- county. incorporated part, although some cities grew quite significantly, and decline in the fragmented, incor- Economic Development and Fragmentation porated areas, where it occurred, was generally Does governmental fragmentation limit a metro- slower than in the unfragmented city. Centrifugal politan region's capacity to develop economically? movements of existing economic enterprises, to- Many of those active in economic development ef- gether with preferences by newly created enterprises forts in St. Louis argue in the affirmative.= They to locate in newer suburban areas, appear to be the believe that St. Louis, as a region, has not been able major factors explaining the pattern of employment to make its case effectively to relocating firms be- change in St. Louis, not governmental fragmenta- cause the area has no single spokesperson to present tion. that case. Further, they believe that new companies Where fragmentation does have an effect on have been deterred from locating in the region, es- economic development and patterns of employment pecially in the incorporated parts of the county, by location in St. Louis is in accelerating its centrifugal the difficulties that firms confront when having to flow. Fragmentation in St. Louis, particularly the deal with public officials from multiple jurisdictions, separation of city and county, prevented the city and with variations in tax rates, zoning and permit from expanding through annexation. When firms procedures, and local attitudes toward business ac- and population moved out of the city into the tivity. While the data on employment change in the county, the city could not recapture them by city-county region and comparable areas are not annexing their new locations. Neither can inner-ring congruent with these arguments in all of their de- suburbs recapture firms and population that move to tails, there are some points of agreement. outer-ring suburbs and unincorporated parts of the First, the incongruities. St. Louis City and county. Further, county government and county mu- County, as a combined region, either outperformed nicipalities that aggressively pursue new enterprises or performed as well as similar regions in the North- are likely to find an enterprise relocating from the east and Midwest. It seems unlikely, although possi- city or another municipality to be as attractive as one ble, that the region would have experienced even newly created or relocating from outside the region, greater growth if it had fewer local governments. No and to encourage such relocations. In this way, frag- other large, core city-county region in the Northeast mentation may affect the location of economic de- and Midwest outperformed St. Louis, although most velopment activity in St. Louis, facilitating the centrifugal flow found in most metropolitan areas, and CONCLUSION perhaps exacerbating it. Another factor that must be considered is the This chapter has considered several issues re- earnings tax in the City of St. Louis. From economic lated to the political economy of jurisdictional frag- theory one can argue that employers in the city must mentation in the St. Louis City and County area. pay higher wages than comparable employers in the One issue is whether a fragmented area fails to cap- county in order to offset the effect of this tax on ture economies of scale in the production of public employees' take home pay. This provides city em- services. Chapters Four and Five reported limited ployers with an incentive to relocate outside of the economies that might be captured in the production city to avoid the tax. In an analysis of Philadelphia's of police and fire services respectively. In the pre- wage tax and its effects on employment there, sent chapter, considering whether such economies Robert Inman estimated that 60 percent of Philadel- existed for total local government expenditures, no phia's total employment loss over the past 20 years effects were found among the municipal govern- could be attributed to its tax on the earnings of per- ments in St. Louis County. Consistent with most of sons employed there.66 No attempt was made to esti- the research findings of the past 30 years, no rela- mate an effect of this nature on the City of St. tionships could be found between the size of popula- Louis' employment decline, but it is reasonable to tion served and local government expenditures. assume that part of the decline can be explained in The second issue addressed in this chapter is this way. that of distributional equity. Local government and The jurisdictional separation of city and county school district expenditures were related to indica- is an important aspect of the problem too. In most tors of local household income, poverty, and race to metropolitan regions, the core county contains the examine whether patterned inequalities existed core city, and county planners are likely to consider among county municipalities and school districts. the economic well being of the central city crucial to Little such patterning was found for total municipal the county's development and its revenue base. expenditures. That is, relationships between local Where core city and county are separated as in St. per capita expenditures and local incomes, presence Louis, the city's economic decline has no direct, of poverty populations, and presence of minority negative effect on county government revenues-in- populations were weak or nonexistent. Some pat- deed county government benefits directly when terning was found across the municipalities when firms relocate there from the city.67 household burdens were examined, however. The One of the significant initiatives of the city- data indicate that communities with higher propor- county government of Indianapolis, operating tions of minority and elderly populations tend to im- through its countywide Department of Metropolitan pose higher burdens on their households, whether Development, has been to combat this centrifugal measured in absolute dollars or by a percentage of flow by pressuring developers to locate new activities household income. An estimate is provided of the within the central city's downtown area. These ef- additional revenue required to reduce these bur- forts have had substantial success,ea thus preventing dens. The amount required is relatively small com- the development of an alternative downtown, such pared to total local revenues, indicating that this in- as the Clayton area in St. Louis County. Pittsburgh's equity would not be difficult to erase if the citizens Renaissance I and Renaissance I1 partnerships both of St. Louis County chose to do so. involved Allegheny County government along with Patterned inequalities were more evident across that of the city and private entrepreneurs, and school districts than across municipalities. This re- Pittsburgh's location within Allegheny County made sults from the significantly greater reliance of school its economic redevelopment of crucial concern to districts on the property tax, greater than by munici- county government there.69 The focus on central palities, which have a variety of other revenue city revitalization that occurred in both of these rust sources. Because property tax revenues are closely belt cities, which in each case involved close col- linked to district property valuations, and because laboration between the core city and county govern- those valuations are in turn linked to district income ments, has not been characteristic of St. Louis.70 Al- and racial characteristics, school finance is less equi- though some cooperative city-county ventures exist, table than municipal finance in the county. This is the incentive for the county to join in these ventures true in spite of the efforts by the State of Missouri to is weaker than in Indianapolis or Pittsburgh. As crit- reduce inequalities in district per pupil revenues ics of St. Louis area fragmentation note, the region through a variety of distributive and redistributive has no effective governmental mechanism that in- programs. Reduction of the patterned inequalities in volves both city and county in joint economic devel- school finance would require a reduction in the reli- opment efforts. ance on property taxes by the districts, together with increased school aid from the state and/or adjust- county. Interorganizational linkages among service ments in the formulas by which state aid is distrib- producers have emerged for most service compo- uted. nents where such economies might be expected. One further aspect of distributional equity-the Second, fragmentation is related to limited in- sales tax distribution process in the county-is exam- equities in local finance. Older citizens and minority ined by comparing expenditures of point-of-sale citizens tend to reside in communities with relatively cities to those of the pool cities. Even after adjust- higher burdens on households. The economic geog- ment for service conditions that clearly demand raphy of the county is such that merging these rela- higher expenditures in the point-of-sale cities-espe- tively disadvantaged communities with equally disad- cially the much larger numbers of persons employed vantaged neighboring communities would do little to within their borders-the point-of-sale cities were reduce these inequities. Some revenue augmentation found to be relatively advantaged by the current for disadvantaged communities would be necessary sales tax distribution system. Estimates are provided to accomplish this. The amount required would be of the additional revenues needed to reduce this dis- relatively small. The other local finance inequity tributional inequity. These revenues, like those that found in the county is related not to fragmentation would be needed to eliminate inequities in house- but to the sales tax distribution formulas. Here too, hold burdens, are relatively small.71 the amount of revenue augmentation for disadvan- Lastly, the chapter examines metropolitan prob- taged communities needed to reduce the inequity is lem solving in the context of economic development small. in the city and county. On an aggregate basis the city Finally, fragmentation does not appear to have and county area has performed well at economic de- slowed economic development in St. Louis. No rela- velopment. Its growth in total employment has tionship could be found between the number of lo- equaled or exceeded that of comparable areas in the cal governments and economic development when Northeast and Midwest, including a number of areas St. Louis City and County were compared to similar that have metropolitan economic development city-counties. In fact, St. Louis out-performed most authorities. One cannot find evidence that fragmen- comparable areas in its region. However, one aspect tation is related to slow economic growth in such of fragmentation does appear to have affected the comparisons, because comparable areas that are distribution of economic development significantly. much more consolidated as well as those that are Fragmentation, in the form of city-county separa- more fragmented have both lagged the city-county tion, appears related to an uneven pattern of signifi- area in new job creation. cant decline in the city and very significant growth in At the same time, St. Louis lacks authoritative the county. The extent of this uneven pattern ex- mechanisms for fostering economic development ceeds that found in most comparable areas. St. across the combined city-county area. Unlike most Louis, unlike most of those areas, has no authorita- metropolitan areas, where a central county includes tive body with competence in economic develop- the central city and, therefore, a positive relation- ment over the combined city and county. While the ship exists between central city development and existence of such a body would not, in all likelihood, county development, in St. Louis that relationship is have prevented uneven development, it might have often negative. The county benefits directly when slowed the centrifugal movement of employment in enterprises relocate from the city to the county. In St. Louis, as such bodies have been able to do in part because of this and, most likely, due also to the other similar areas. city's earnings tax, the pattern of economic develop- ENDNOTES ment in the city and county has been decidedly un- 'See, for example, Reshaping Government in Metropolitan even. The City of St. Louis has experienced a sig- Areas (New York: Committee for Economic Development, nificant decline in employment over the past 20 1970). Recent statements of this argument in the St. Louis years, while the county has experienced extraordi- context include the Report of the St. Louis County Annexa- tion Study Commission (Clayton, Missouri: St. Louis nary employment growth. This uneven pattern is County Department of Planning. July 1985), p. 27, and most likely a result of fragmentation, though not mu- Too Many Governments? A Report on Governmental Struc- nicipal fragmentation in the county. Rather, the ture in St. Louis City and County With Recommendations for Change (St. Louis, Missouri: Confluence St. Louis, relevant source of fragmentation is city-county sepa- February l987), p. 21. ration. 2Good summaries of research on scale economies can be Summarizing the linkages of fragmentation to found in William F. Fox, Jerome M. Stam, W. Maureen the political economy of St. Louis, the following Godsey, and Susan D. Brown, Economies of Size in Local Government: An Annotated Bibliography (Rural Develop- generalizations can be made. First, fragmentation ment Research Report No. 9, U.S. Department of Agricul- does not result in any significant failure to capture ture, Economics, Statistics, and Cooperatives Services, economies of scale across the municipalities in the April 1979): Robin Barlow, "City Population and Per Cap- ita City Government Expenditures," Research in Populution Louis) with many small governments, than in an area (Bal- Economics, Vol. 4, 1982, pp. 49-82, and in Werner Z. timore) with only two, relatively large governments is con- Hirsch, Urban Economics (New York: Macmillan Publish- sistent with the view that economies of size are not an im- ing Company, 1984). Interestingly, much of Hirsch's early portant factor in local government service delivery. empirical work was done in St. Louis. Q~ohnC. Bollens, ed., Exploring the Metropolitan Commu- A difficulty with most research on scale effects in local nity (Berkeley: University of California Press, 1964), government research is the lack of valid measures of output. Chapter 14, "Factors Affecting Local Governmental Expen- Economy of scale means decreasing per unit costs as output diture Levels. " levels increase, but most of the literature measures scale by size of population served rather than units of outputs pro- losee, for example, the argumentation in Hawkins v. Town duced. It is perhaps better to refer to such research as of Shaw, and other cases before the U.S. Supreme Court measuring size, rather than scale effects. In addition, sim- where violation of "equal protection" was alleged on the ba- ply relating per capita service costs to size of population sis of revenue and expenditure variations. served does not account for differences in service conditions "Whether such attribution of motivation and constraint is ac- among the units studied-differences that are likely to be re- curate cannot be ascertained through statistical analysis of lated to the level of service demanded and the cost of service revenue and wealth patterns. However, such analyses are production. frequently made, with accompanying discussion which suggests these factors as providing explanation. 3Those who argue the need for larger municipal jurisdictions to capture presumed economies of scale or size frequently 12~reatercredence should be afforded the "weighted" coeffi- base their arguments on standards of "good business prac- cients. Without such weighting, Champ (population 31) tice" which could be employed by municipalities if only they and Peerless Park (67), are treated as the equivalent of were larger. The irony of this argument is that many quite University City (42,929) and Florissant (55,949). Weight- competitive business firms are smaller than the government ing by relative populations provides a more accurate picture of a town of 5,000 people. See George J. Stigler, "The Ten- of revenue patterns by adjusting for the gross population dif- able Range of Functions of Local Government," in Edmund ferences among these and other municipalities. S. Phelps, ed., (New Private Wants and Public Needs 13The weighted correlation between median household in- York: W.W. Norton, 1962). come and ~ercentof housing stock built between 1970 and 4Total expenditures include municipal and fire protection 1980 is a iositive 0.38 and Ys statistically significant at p < district expenditures, where relevant and, in two of the four .001. equations, expenditures by householders for residential 14Those with higher incomes also tend to pay higher direct tax trash collection. burdens in the county (the correlation between household rash collection is an important factor because in some mu- tax burden and median household income is 0.61-see the nicipalities its cost is roughly comparable to the tax burden following section). The positive relationship between total for all other local services-municipal and fire protection. per capita revenues and median income is in part a reflec- Some county municipalities collect residential trash with tion of this fact. their own sanitation departments, others contract with trash 15significant correlates of the commercial-industrial assessed hauling companies, and still others franchise one or more valuation per capita are median household income (+0.31), haulers. In some municipalities trash collection is not billed percent nonwhite (-0.22), percent of housing built between to a householder (i.e., it is paid for by the municipality out 1970 and 1980 (+0.33), and population density (-0.45). In of general revenues), in others the municipality pays for a a regression analysis predicting assessed value of commer- portion of the cost out of general revenues and the house- cial-industrial activity per capita, only population density is holder pays the remainder directly to the trash collector statistically significant [data not shown]. (whether municipal, contract, or franchise), and in still others, the householder pays the full cost of collection di- 16~stimatesof household tax burdens for each of the county rectly to the collector. The method of estimating the per municipalities are contained in the Appendix to this Chap- household cost of trash collection in the municipalities is ter. The method for developing these estimates is presented described in Roger B. Parks, "Revenue and Expenditure Es- ,in Roger B. Parks, op. cit. timation for the ACIR St. Louis Study," Working Paper 17see the previous discussion of the alternative arrangements CPPM 88-1, Bloomington, Indiana: Center for Policy and for provision of trash collection across the municipalities Public Management, School of Public and Environmental and how this affects the comparability of household burden Affairs, Indiana University, 1988. estimates. sThis would be the linear approximation to a cost-size rela- 18"~ithouttrash collection costs" in this context means that tionship that exhibited decreasing average costs as popula- the burdens include utility and property taxes, but not direct tion increased-or economies of size. payments for trash collection where these are required. 7Fo~,et a]., Economies of Size . . . ; Hirsch, Urban Eco- "With trash collection costs" means that direct trash collec- nomics. tion payments have been added to tax burdens in those communities where trash collection is not supported from *A further bit of evidence on the economy-diseconomy ques- general revenues. tion as it relates to fragmentation can be offered. Using lQ~wenty-fourof the point of sale cities paid for this collec- data published by the U.S. Bureau of the Census, total local government expenditures for current operations (exclud- tion with general revenues. These cities contained about 60 ing those for education) reported in 1984-85 for the frag- percent of the point of sale population. Twenty-one of the mented St. Louis City and County area were $548 per pool cities also paid for trash collection with general reve- capita. Comparable data for Baltimore City and County nues, but their population represented only 22 percent of the area (which has only two general purpose local govern- pool population. Thus, households in pool cities are likely ments, separate from each other like St. Louis City and to have higher dollar and percentage burdens than those in County, and a population virtually identical to that of St. point-of-sale cities-in large part because of the cost of Louis and St. Louis County) were expenditures of $601 per trash collection to them. capita. This comparison, showing slightly lower per capita 200ne has only to cite cities such as Fenton, Frontenac, expenditures for local services in a fragmented area (St. Hazelwood, or Peerless Park, where residential tax burdens are quite low as a result of large revenues from commercial 31~uiteobviously, the municipalities that chose point-of-sale and industrial activity. status in 1977 or 1980 were those which had significant 21While commercial and industrial activity has only a small sales tax generating activity. See James M. Brasfield, "The effect on household tax burdens, it does affect expenditures Great Land Rush: Annexation and Sales Tax in St. Louis for local services quite significantly (see "The Sales Tax County, " presented at the Annual Meeting of the American Controversy Revisited" below). Political Science Association, New Orleans, Louisiana, August 29-September 1, 1985, and Andrew D. Glassberg, 22~hisestimate is reached by computing the excess percent- "The Management of Uncertainty: The St. Louis County age burden in those communities where it is found, multi- Sales Tax," in Carol W. Lewis and A. Grayson Walker, plying that excess percent by average household income eds., Casebook in Public Budgeting and Finance there and then by the number of households, and summing (Englewood Cliffs, New Jersey: Prentice-Hall, 1984), pp. the results across those communities that have percentage 73-76. burdens above the 1.03 percent average. The actual estimate is slightly below $8 million, with the $8-10 million fig- %.s. Bureau of the Census, Census of Retail Trade, Census ure in the text allowing for as much as a 25 percent under- of Wholesale Trade, Selected Service Industries, 1982 and estimate of the amount required. prior years. Data in this section refer to 71 municipalities for which retail, wholesale, and service industry employ- 230ne reason why the amount required is so small is that the ment figures were available or which had less than five per- State of Missouri and its local governments raise quite small cent of their assessed valuation in commercial and indus- revenues from their citizens when compared to governments trial property (employment was set to zero for the latter). in other states. In 1985, Missouri's combined state-local These 71 municipalities contain approximately 97 percent tax collections per capita were $1,091. Only seven states of the municipal population in 1984. All averages and coef- collected less per capita. Missouri's state-local tax revenue ficients are computed on a population-weighted basis so as as a percent of personal income was 9.01 percent in 1985. to more accurately reflect the average values in the county. Only New Hampshire had a lower percentage. See U.S. Advisory Commission on Intergovernmental Relations, Sig- 33Data for Part I (serious) crimes taken from 1985 Fact nificant Features of Fiscal Federalism, 1987 Edition, pp. Sheet: Police Services in St. Louis County (Clayton, Mis- 49-50. souri: Bureau of Management Systems, St. Louis County Police Department, l985), Table 3. 24~evenuedata for county school districts is taken from the St. Louis County Annual School Report, November 1985, 34~eeJames M. Brasfield, "Follow the Money: Fiscal Struc- published by the Cooperating School Districts of the St. ture and Metropolitan Reorganization in St. Louis County," Louis Suburban Area. Revenue data for the city district is presented at the Annual Meeting of the American Political taken in part from St. Louis Public Schools: A Winning Science Association, Chicago, Illinois, September 3-6, Year 84-85, published by the St. Louis (City) Public 1987. Schools and in part from Table 4 of the 1984-85 Report of 35~atharineL. Bradbury, Helen F. Ladd, Mark Perrault, the Public Schools of Missouri, published by the Missouri Andrew Reschovsky, and John Yinger, "State Aid to Offset State Board of Education. Data from the Cooperating Fiscal Disparities Across Communities, " National Tax School Districts Report and the Missouri State Board of Journal 37 (June 1984): 151-170. The method is developed Education Report exhibit some differences. Adjustments from economic theory and adapted for practical decision- were made to tabulated data for purposes of consistency, making by Massachusetts' policymakers. The reader is re- but comparisons of the city and county school districts must ferred to the article for the theoretical development. be considered tentative. 25~ecallfrom Table 2.8 that the least advantaged municipali- 3e~bid,p. 158, parenthetical phrase added for clarification. ties (and fire districts) raised, on average, about 38 percent of the revenues of the most advantaged. 381n Massachusetts, municipalities were limited almost exclu- 2e~xtrajurisdictiona1revenues are partially per-capita, dis- sively to the property tax as an own-source resource. There, tributed on the basis of enrollment, and partially compensa- the equalized property tax base was used to index this re- tory, distributed according to a "need" formula. Both forms source. Missouri municipalities and fire protection districts of distribution act to reduce revenue variations that result which serve them rely on property taxes, but the municipali- from differences in district tax bases and tax efforts, al- ties also rely on the sales and gross receipts taxes for major though the compensatory distribution has a stronger effect. shares of their revenues. 27~ederalfunds, which in the past were distributed in part to 3Q~hemodel is a powerful predictor of per capita expendi- general purpose local governments in a compensatory man- tures, explaining 86 percent of the variance in these expen- ner, have shrunk to a very small proportion of municipal ditures across the 71 municipalities. revenues. 401.e., cities whose residents are wealthier spend more for 28~hesecoefficients were computed using total property valu- public services. In national studies the elderly generally are ation, unlike Table 8.2 computations, as the valuation of found to support lower levels of public expenditures. See commercial and industrial property by school district was Bradbury, et al., "State Aid . . . ." unavailable. 41~recent analysis of expenditures by nearly 1,900 suburban 2Q~heregressions are not shown because only assessed valu- municipalities found that the "strongest determinant," the ation has any effect. The coefficient for assessed value per variable which best predicted their expenditures, was the ra- student in both total and district regressions, weighted or tio of the number of persons employed in a community to unweighted, is 0.02, indicating an average revenue per stu- the community's resident population, with a strong, positive dent of $20 for each $1,000 of assessed valuation. The re- effect. See John R. Logan and Mark Schneider (1982) gressions explain 90 percent of the variation in total and "The Effects of Business Activity, Functional Responsibil- district revenues. ity, and Regional Context on Suburban Municipal Expendi- 30~11municipalities retain the revenues generated within their tures," in Richard C. Rich, ed., Analyzing Urban-Service borders from nonsales taxes and fees on commercial and Distributions (Lexington, Massachusetts: Lexington industrial property. Books, 1982). pp. 235-250. 42Computed as each city's per capita cost of retail, whole- nition, 29 have tax burdens as a percent of household in- sale, and service industry employment multiplied by city come below the county average-indicating that a portion of population and summed across the 71 cities. their disadvantage is the result of choosing lower burdens. If they were required to bring their own burdens up to the 43~heSt. Louis County Annexation Study Commission re- county average before receiving additional funds, however, ported a similar ratio using 1982 data. See its report at page the reduction in additional funds need would be only $1.8 17. million, as most of these 29 cities are quite Small. 441t was not possible to estimate the amount, but it is clear 54This argument has been advanced by critics of local govem- that the unincorporated county and Maryland Heights pro- ment fragmentation in St. Louis, who find that the large vide revenue redistribution to other cities in the pool by number of local governments there has prevented the area sharing their higher per capita sales tax collections. from resolving regional economic development problems, 45~67in the average point-of-sale city minus $18 in the aver- causing a relative economic decline. For example, Dana L. age pool city. Cost factors and indices for each of the 71 Spitzer, Fostering Development in Metropolitan St. Louis cities are listed in Parks, "Revenue and Expenditure Esti- (St. Louis, Missouri: City-County Task Force of Civic Pro- mation . . . ." gress, December 1982) : . . . it (the Advisory Committee which prepared this 46~ignificanthere means employment per 1,000 residents well report) does see [governmental] proliferation and above the average, not necessarily high total employment. fragmentation creating major problems related to 47Two pool cities, Normandy and Wellston, have cost indices area-wide issues, services and functions. Economic above the average of the 71 municipalities. Four point- development is included among this latter realm of of-sale cities, Dellwood, Jennings, Kirkwood, and concerns. Shrewsbury, have cost indices below the 71 city average. In attempting to address issues of a metropolitan Other than these, all pool cities have below average indices, nature, St. Louis is seriously incapacitated. Many while all point-of-sale cities have above average indices. independent governmental units exist in an environ- 48~ecallthat point-of-sale cities raise more revenues from all ment in which they (sic) are few clear, immediate sources than do pool cities. As a rough estimate, about 60 inducements to cooperation and collaboration. percent of the point-of-sale cities' advantage is a direct ef- There is a frequent perception that local jurisdic- fect of the sales tax distribution, while the remainder is a tions are in competition. No strong, general purpose cumulative effect of property taxes, gross receipts, and mechanisms for representing areawide interests ex- other sources. ist, and the general orientation to addressing most issues is for a maze of separate jurisdictions to vie 49~neconomist would question how much of the extra expen- for what they perceive to be thei; individual advan- diture represents added service to residents and how much tage. (pp. 21-22). might be attributed to possible inefficiencies encouraged by This finding was repeated in Confluence St. Louis, Too excess revenues. Absolutely no data were available to ad- Many Governments?: dress this question. Multiplicity affects economic development in three =OThis result suggests that a change in the sales tax distribu- major ways: (a) indirectly, by inhibiting the re- tion formula would be in order if achieving fiscal equiva- gion's ability to solve metropolitan problems that lence were to be pursued. Moving to a countywide pool dis- have a clear impact on economic development; (b) tribution formula would not achieve this, however, as by fostering competition, which inhibits the develop- point-of-sale cities do incur higher service costs than do ment of a regional consensus on strategies for eco- pool cities as a result of the commercial-industrial activities nomic development; and (c) by complicating the within their borders. A mixed formula, allowing all cities to process for those interested in relocating to St. Louis retain sufficient revenues from commercial activities to (P. 25). cover the added cost of servicing them, with the remainder distributed on a per capita basis would be one way of mov- 5% Spitzer, Fostering Development in Metropolitan St. ing toward fiscal equivalence. If redistributive goals were to Louis, p. 18. be pursued as well, the distribution of net revenues from 56~eitherdo these comparisons reveal any particular correla- commercial-industrial activities might be tied to a formula tion between number of local governments in the regions which indexed each jurisdiction's "needw-perhaps using a and their relative successes in adding employment. cost index as developed here-and it's own tax effort as 571n a recent article, U.S. & identified measured by household tax burdens. The State of Missouri News World Report uses such a mixed formula to allocate redistributive funds to St. Louis County as one of 46 "boom" counties in America, local school districts under its Foundation Program. projecting a further increase of 126,000 jobs in the county through the year 2000 ("The Boom Counties," November 5'~nthe latter case, with discounting for the service cost in- 9, 1987, p. 108). dex, approximately $20 million would be allocated to 26 5BUndoubtedly many of the new jobs in St. Louis County were pool cities that are disadvantaged, while $8 million would jobs lost to the city, as firms moved out from the city to the be allocated to eight disadvantaged point-of-sale cities. county. Similar movements have been reported in many 52~itiesthat are disadvantaged by this definition are not un- metropolitan areas, particularly older ones. No data were able to pay for local services. Rather, current revenue bases found by which to quantify this movement in St. Louis, and fiscal choices are such that they spend fewer dollars per however. If one were to assume that all of the city's loss capita for local services in relation to their service condi- was simply relocation to the county and, therefore, should tions than do other municipalities with similar conditions. not be counted in the county's growth, the county's employ- They are disadvantaged in the sense that, while facing simi- ment increase from 1970 to 1985 would drop to 51 percent, lar conditions, they are less able to raise revenues to fund still a very significant increase. local services. 59For a compendium of negative national images of St. 53~hirty-sevenof the 45 are pool cities, needing $14.5 million Louis, see Spitzer, Fostering Development in Metropolitan to eliminate their disadvantage by this definition. Eight are St. Louis, pp. A-1 to A-3. All those cited referred to the point-of-sale cities which, in total, would need about $5 city as the locus of very serious urban problems, including million. Of the 45 cities that are disadvantaged by this defi- loss of jobs. 60"Winter cities," contrasted with those in the sun belt, have mate conducive to economic growth." She cites as specific recently arrived as a subject of study. See, for example, difficulties those of ". . . presenting a united front to the Gary Gappert, ed., The Future of Winter Cities (Beverly state [of Missouri], . . . reaching agreement on regional pri- Hills, California: Sage Publications, 1987). In an article in orities and mustering the forces to carry them out, ... this volume (pp. 13-34), "The Cold City: The Winter of [inlability to make commitments and close deals when try- Discontent?" Seymour Sacks, George Palumbo, and Robert ing to attract industries, and . . . [inlability to present a Ross identify the linkage between fixed central city borders united front to outsiders . . .". Dee Joyner, December 23, and relative population and job losses. Cities without fixed 1987. borders (e.g., Memphis) have maintained or increased BeRobert P. Inman, "Regression Analysis of the Wage Tax population and employment to a significant degree by an- and Employment Trends," in Thomas F. Luce and Anita nexing suburban areas to which population and employment A. Summers, Local Fiscal Issues in the Philadelphia Met- has relocated. ropolitan Area (Philadelphia, Pennsylvania: University of elsee Sacks, et al., "The Cold City," pp. 30-31. Central city Pennsylvania Press, 1987), pp. 176-178. growth in employment has been particularly significant in 67~herevenue base of county government benefits especially Indianapolis-Marion County, even though the combined when firms locate in the unincorporated part of the county. city-county growth percentage there is equivalent to that in There, the county can collect utilities taxes and other fees St. Louis City and County. Local economic planners in St. from them in addition to property and sales taxes. For this Louis foresee a reversal of the city's employment decline. reason, county government benefits even from intracounty projecting an increase of nearly 31,000 jobs there by the relocations if firms move from incorporated to unincor- year 2,000. This projected increase will all occur within a porated areas. County government has had an incentive to few block area of the central business district, however. encourage business location in the unincorporated county- with the remainder of the city losing jobs or staying virtually whether new or relocating business-and this encourage- the same. See Yesterday, Today, and Tomorrow: A Profile ment no doubt helps to explain the rapid growth there in re- of Change for the St. Louis Metropolitan Area (St. Louis, cent years. Missouri: East-West Gateway Coordinating Council, No- vember 1986), p. 122. 68~amesC. Owen, "A Critique of Metropolitan Government: 62~eeU.S. Bureau of the Census, Census of Retail Trade, The Indianapolis Unigov Experience," in Roger C. Stough Census of Wholesale Trade, and Selected Service Industries and Theo Toonen, eds., Public Infrastructure Redefined for 1972 and 1982. (Rotterdam, The Netherlands: Department of Public Ad- ministration, Erasmus University), forthcoming 1988. B3Data from Roger Grow, St. Louis County Planning Depart- ment, and Dee Joyner, Executive Director, St. Louis egBrian J. L. Berry, Susan W. Sanderson, Shelby Stewman, County Economic Council. and Joel Tarr, "The Nation's Most Livable City: Pittsburgh's Transformation, " in Gary Gappert, ed., The 64~atafrom East-West Gateway, Yesterday, Today, and To- Future of Winter Cities, pp. 173-195. morrow, p. 122, project about one-third of the jobs in the county will be in the area outside of Interstate 270 by 2000. 70~uchfocusing is not without cost. Owen reports that Indi- East-West Gateway's projections are not comparable with anapolis, by focusing on its downtown, is alleged to have those cited earlier from U.S. News and World Report, nor slighted development in economically. depressed- areas sur- are their 1980 base data comparable with the Census Bu- rounding the downtown. Berry, et al., indicate similar reau's County Business Patterns data for that year. problems in the Monongahela Valley towns south of the City of Pittsburgh. @=seethe statements cited in Endnote 54 and additional ones in Civic Progress, Fostering Development in Metropolitan "voters in St. Louis County turned down a proposed 0.275 St. Louis, and Confluence St. Louis, Too Many Govern- percent increase in the county sales tax in June. The ments? In a personal communication, the Executive Direc- amount that would have been raised by this increase, if tar- tor of the St. Louis County Economic Council stated geted to relatively disadvantaged communities, could have that ". . . individuals who 'do' economic development on a substantially reduced the inequities in household burdens daily basis feel strongly that government multiplicity does and between point-of-sale and pool cities found in these impede their efforts to attract jobs or create the kind of cli- analyses. Appendix Table 8.1 Revenues and Expenditures in County Municipalities- 1985 Fire District Municipal Revenues from Total Total Municipality Revenues Municipality Revenues Expenditures' Ballwin $5,838,512 $667,568 $6,506,080 $5,040,125 Bel Nor 309,194 96,103 405,297 375,323 Bel Ridge 634,000 1 15,534 749,534 749,534 Bella Villa 147,793 25,447 173,240 142,920 Beilefontaine Nbrs 2,289,106 364,932 2,654,038 2,529,208 Bellerive Acres 61,000 22,006 83,006 83,006 Berkeley 7,000,000 0 7,~,~7,608,339 Beverly Hills 393,949 22,926 41 6,875 408,858 Black Jack 552,000 146,710 698,710 706,710 Breckenridge Hills 862,000 99,033 961,033 960,894 Brentwood 4,205,159 0 4,205,159 3,977,352 Bridgeton 7,250,170 1,283,205 8,533,375 8,313,837 Bridgeton Terrace 52,617 8,020 60,637 58,398 Calverton Park 207,464 43,738 251,202 221,559 Champ 6.501 6,602 13,103 1 0,417 Charlack 296,221 28,042 324,263 290,321 Clarkson Valley 336,219 228,272 564,491 561,491 Clayton 9,366,241 0 9,366,241 9,759,879 Cool Valley 462,900 100,262 563,162 563,162 Country Club Hills 304,748 0 304,748 296,791 Country Life Acres 18,288 0 18,288 36,358 Crestwood 4,344,101 0 4,344,101 4,022,928 Creve Coeur 6,658,290 1,378,691 8,036,981 8,548,043 Crystal Lake Park 114,255 0 1 14,255 93,247 Dellwood 1,435,377 138,110 1,573,487 1,548,997 DesPeres 3,463,088 0 3,463,088 3,404,692 Edmundson 440,068 84,792 524,860 495,945 Ellisville 2,845,918 440,248 3,286,166 3,549,512 Eureka 1,324,180 330,878 1,655,058 1,569,355 Fenton 2,549,707 505,065 3,054,772 2,424,480 Ferguson 5,609,421 0 5,609,421 5,619,812 Flordell Hills 159,389 0 159,389 147,424 Florissant 9,784,098 1,811,938 11,596,036 1 1 ,192,200 Frontenac 1,783,000 0 1,783,000 1,783,176 Glen Echo Park 30,440 7,388 37,828 38,941 Glendale 1,305,305 0 1,305,305 1,224,323 Grantwood Village 195,000 64,747 259,747 259,747 Greendale 1 15,000 38,162 153,162 141,752 Hanley Hills 31 5,930 75,422 391,352 41 3,324 Hazelwood 6,723,446 0 6,723,446 6,599,284 Hillsdale 375,882 64,667 440,549 443,825 Huntleigh 88,953 0 88,953 68,774 Jennings 4,970,964 0 4,970,964 4,688,674 Kinloch 866,517 46,670 913,187 1,010,251 Kirkwood 7,034,281 0 7,034,281 8,398,390 Municipal expenditures plus fire district revenues where applicable. Fire district expenditures in each mlunicipality could not be computed. NA - Data were not available. (continued on next page) Appendix Table 8.1 (cont.) Revenues and Expenditures in County Municipalities- 1985 Fire District Municipal Revenues from Total Total Municipality Revenues Municipality Revenues Expenditures1 Ladue $3,993,696 $0 $3,993,696 $4,086,817 Lakeshire 266,523 49,550 31 6,073 321,325 Mackenzie l9,OOO 4,215 23,215 8,835 Manchester 2,316,152 305,337 2,621,489 3,078,630 Maplewood 3,800,000 0 3,800,000 3,400,000 Marlborough 463,474 81.010 544,484 487,796 Maryland Heights NA NA NA Moline Acres 490,590 560,891 560,891 Normandy 1,337,922 1,557,765 1,489,345 Northwoods 1,481,775 1,663,622 1,671,911 Norwood Court 95,000 135,839 96,839 Oakland 242,553 242,553 21 1,994 0l ivette 3,485,647 3,485,647 3,347,070 Overland 4,948,355 5,499,716 5,452,179 Pagedale 1,590,885 1,778,519 1,584,667 Pasadena Hills 230,787 294,414 257,878 Pasadena Park 88,000 112,711 1 12,735 Peerless Park 1 12,094 127,390 131,921 Pine Lawn 1,132,087 1,247,749 1,159,841 Richmond Heights 3,684,144 3,684,144 3,620,560 Riverview 471,801 548,445 472,545 Rock Hill 1,373,609 1,373,609 1,369,767 Shrewsbury 1,817,882 1,817,882 1 ,772,414 St. Ann 5,250,000 5,741,530 5,734,230 St. George 256,378 296,568 321,042 St. John 1,277,000 1,436,500 Sunset Hills 1,906,673 2,358,140 Sycamore Hills 94,346 109,977 Town and Country 3,645,034 3,661,272 Twin Oaks 203,580 245,748 University City 13,216,828 13,216,828 Uplands Park 100,643 114,986 Valley Park l,l5O,O49 1,309,162 Velda Village 328,838 369,438 Velda Village Hills NA NA Vinita Park 685,298 902,400 Vinita Terrace 44,788 58,919 Warson Woods 700,000 700,000 Webster Groves 6,608,915 6,608,915 Wellston 1,137,OOO 1,255,910 Westwood 46,000 81,519 Wilbur Park 71,289 84,804 Winchester 406,138 467,787 Woodson Terrace 1,021,617 1,176,867

Municipal expenditures plus fire district revenues where applicable. Fire district expenditures in each municipality could not be computed. NA - Data were not available. Appendix Table 8.2 Service-Cost Burden on an Average Household- 1985 Property Property Average Taxes: Taxes: Gross Trash Total Household Municipality Municipal' Fire District Receipts Collection Burden Income Ballwin $1 13.97 $74.59 $384.61 $25,941 Bel Nor 0.00 0.00 195.71 27,157 Bel Ridge 0.00 0.00 91.47 18,754 Bella Villa 47.01 74.40 242.21 18,860 Bellefontaine Nbrs 114.86 97.20 31 1.97 22,682 Bellerive Acres 16.23 0.00 498.37 53,707 Berkeley 98.43 30.31 198.93 19,207 Beverly Hills 83.1 1 0.00 163.64 16,228 Black Jack 0.00 86.40 185.33 27,975 Breckenridge Hills 0.00 0.00 61.54 16,836 Brentwood 77.44 0.00 168.30 21,897 Bridgeton 86.66 0.00 231.37 24,623 Bridgeton Terrace 83.59 72.00 249.38 17,442 Calverton Park 5.72 86.40 198.64 25,530 Champ 0.00 72.00 176.04 40,000 Charlack 90.36 60.00 206.87 17,794 Clarkson Valley 30.47 0.00 761.41 63,559 Clayton 186.44 0.00 392.00 36,211 Cool Valley 80.21 0.00 181.61 18,233 Country Club Hills 124.49 52.80 275.61 16,841 Country Life Acres 0.00 0.00 520.02 89,372 Crestwood 87.12 87.12 234.67 28,710 Creve Coeur 124.55 0.00 435.39 49,212 Crystal Lake Park 70.57 0.00 294.28 32,333 Dellwood 87.69 0.00 165.26 23,676 Des Peres 83.55 0.00 185.42 39,856 Edmundson 0.00 0.00 71.44 19,735 Ellisville 139.36 47.95 41 7.69 27,287 Eureka 66.80 34.80 283.66 21,776 Fenton 0.00 0.00 111.99 26,604 Ferguson 91.60 115.20 260.72 21,867 Flordell Hills 73.25 52.80 176.24 16,053 Florissant 61.49 86.40 277.1 4 24,712 Frontenac 15.74 0.00 100.38 72,323 Glen Echo Park 12.39 0.00 140.48 24,462 Glendale 113.07 100.80 306.98 31,172 Grantwood Village 22.44 72.00 379.20 40,720 Greendale 77.85 60.00 244.33 22,505 Hanley Hills 0.00 60.00 165.71 18,573 Hazelwood 0.00 64.80 137.39 22,542 Hillsdale 0.00 0.00 62.29 15,038 Huntleigh 0.00 76.80 377.46 99,469 Jennings 96.09 0.00 172.77 17,689 Kinloch 59.02 60.00 169.55 11,007 Kirkwood 122.52 77.76 289.44 26,605 Including municipal real and personal property taxes and the municipality's share of the County Road and Bridge tax. Municipal and County Library taxes are excluded. NA-Data were not available. (continued on next page) Appendix Table 8.2 (cont.) Service-Cost Burden on an Average Household - 1985 Property Property Average Taxes: Taxes: Gross Trash Total Household Municipality Municipal1 FIre District Receipts Collection Burden Income Ladue $250.59 $1 18.80 $686.54 $80,135 Lakeshire 74.49 74.40 321.08 20,718 Mackenzie 0.00 72.00 146.19 25,559 Manchester 74.26 0.00 274.89 29,266 Maplewood 73.95 0.00 154.80 1 4,703 Marlborough 0.00 78.00 147.50 16,158 Maryland Heights NA NA NA NA Moline Acres 74.93 0.00 143.74 20,409 Normandy 158.99 0.00 278.87 18,875 Northwoods 161.09 0.00 291.01 22,243 Nowood Court 0.00 0.00 201.56 19,815 Oakland 93.09 78.00 229.42 29,696 Olivette 190.30 0.00 380.78 33,120 Overland 81.81 0.00 152.64 19,929 Pagedale 83.00 0.00 150.79 18,698 Pasadena Hills 71 .I7 0.00 290.17 28,912 Pasadena Park 0.00 0.00 171.57 22,580 Peerless Park 0.00 0.00 74.02 21,288 Pine Lawn 100.1 1 0.00 171.02 16,751 Richmond Heights 65.40 0.00 215.98 24,853 Rive~iew 19.25 74.40 181.43 16,584 Rock Hill 79.55 90.72 220.52 21,036 Shrewsbury 93.62 0.00 198.30 21,116 St. Ann 55.51 0.00 126.46 18,929 St. George 0.00 74.40 195.10 19,389 St. John 26.05 0.00 105.34 18,944 Sunset Hills 86.40 0.00 274.36 30,119 Sycamore Hills 17.10 60.00 175.53 21,152 Town and Country 90.54 0.00 41 1 .I5 64,157 Twin Oaks 0.00 0.00 302.08 36,190 University City 152.99 93.60 392.05 21,759 Uplands Park 15.96 48.00 161.87 20,614 Valley Park 60.00 74.40 270.57 16,435 Velda Village 87.63 0.00 182.98 17,514 Velda Village Hill NA NA NA 18,560 Vinita Park 72.74 0.00 171 .O3 17,955 Vinita Terrace 19.04 60.00 187.58 21,891 Warson Woods 176.54 100.80 436.27 3,435 Webster Groves 135.50 1 18.80 389.27 26,555 Wellston 89.54 0.00 136.13 1 1,628 Westwood 42.18 0.00 402.98 74,326 Wilbur Park 0.00 74.40 156.20 21,760 Winchester 95.52 74.40 304.75 22,790 Woodson Terrace 56.04 86.40 204.21 22,015

Including municipal real and personal property taxes and the municipality's share of the County Road and Bridge tax. Municipal and County Library taxes are excluded. NA- Data were not available. Chapter Nine Functional Dimensions of Metropolitan Organization

INTRODUCTION municipalities are able to perform a uniform set of St. Louis County is a classic case of the jurisdic- functions throughout the county. To many observ- tional "patchwork" long criticized by advocates of ers, this finding is a source of considerable distress. metropolitan reform. Neatness and uniformity in the Whatever the particular reform proposal, advocates distribution of responsibilities among local govern- of metropolitan reform generally seek to create a ments are not to be found here. Alongside 90 mu- pattern of organization that can apply uniformly nicipalities, 40 percent of the county's residents live throughout a metropolitan area. In this manner, in unincorporated areas. Municipalities range in size every county resident would receive police services, from less than a hundred to more than 50,000 resi- for example, in substantially the same manner, not dents. Some municipalities provide fire protection, in one of several different ways. but others depend on overlying fire protection dis- Local self-determination tends to foreclose the tricts, as do all those who live outside municipal development of uniform patterns of organization. A boundaries. Streets and street services are provided lack of uniformity, however, does not necessarily not only by the county government and 90 munici- imply a lack of organization. To discover how St. palities, but also by more than 400 organized subdi- Louis County is organized, one has to inquire into visions, many of which retain separate control over more than jurisdictional arrangements. Jurisdictional vehicular access. Twenty-three school districts, rang- fragmentation within the county is associated with a ing in size from a few hundred students to nearly productive political and economic process that gives 20,000, provide elementary and secondary educa- rise to an elaborate organizational overlay-volun- tion, supplemented by the Special School District to tary associations of governments, cooperative agree- provide education for handicapped students and vo- ments, and general rules of law-that links, and cational-technical education. Although a few school often integrates, separate jurisdictions in complex districts conform closely to municipal boundaries, patterns of functional relationships. Unlike the lines most do not. and boxes drawn on an organization chart, the func- The variation in municipal sizes makes it diffi- tional arrangements that link multiple jurisdictions cult to sort out functions neatly between county and tend to be dynamic, created in response to particu- municipal governments, or between municipalities lar problems and opportunities as they arise. Yet and fire districts, in the incorporated portion of the many of those arrangements have also acquired a county. The presence of a large unincorporated area considerable stability and regularity across jurisdic- requires county government to maintain direct serv- tional boundaries. ice capabilities that are not necessary in the incorpo- If the frame of reference is broadened to in- rated areas. Neither the county government nor the clude St. Louis City, the picture changes somewhat. The dominant feature of the governmental land- and provisions of the state constitution, depends scape is a jurisdictional boundary between two non- heavily on citizen initiative and consent to create, overlapping, general purpose units-St. Louis City expand, and modify units of local government. With and St. Louis County. The only overlying jurisdic- the exception of the county government, local juris- tions, aside from the State of Missouri, consist of dictions have been formed at local initiative, and special purpose districts. Yet the City of St. Louis their formation approved by local citizens, according and the city school district participate in many of the to various processes of petition and referendum. joint service delivery arrangements that connect the Even the county government has been substantially county municipalities, school districts, and fire dis- modified through local initiative by the adoption and tricts. The city-county border is clearly not irnper- amendment of a home rule charter. meable to functional linkages. The absence of a gen- It is incorrect to conclude that St. Louis County eral purpose overlying jurisdiction, however, may has no form of metropolitan governance (i.e., no inhibit conflict resolution between city and county. governance arrangements that embrace all units of The immediate objective of this study was to de- government in the county). Although lacking a sin- scribe the functional arrangements that coexist with gle metropolitan government, the county has a co- jurisdictional fragmentation in a single, highly frag- herent governance structure based on the principle mented metropolitan area. Drawing on process crite- of local self-determination. This structure consists of ria to evaluate these functional arrangements, the rules that allow citizens to make basic constitutional results of the study can also shed light on this addi- choices that create and govern their local govern- tional question: Can a functional metropolitan order ments through local charters. Although derived from successfully be built on a jurisdictional base of frag- numerous legal sources, together these rules create a mentation? functional arrangement, here called a local govern- Using the theory of fragmentation developed in ment constitution, that provides for metropolitan Chapter One, the discussion in this closing chapter self-governance. first reviews the functional dimensions of metropoli- As explained in Chapter Three, a local govern- tan organization found in St. Louis County and, to a ment constitution is built on two levels of constitu- somewhat lesser extent, between St. Louis City and tional choice: (1) a choice of enabling rules that County. The discussion then moves on to the task of provide the basis for organizing units of local govern- evaluation, requiring an assessment of functional ar- ment (e .g., procedures for municipal incorporation) rangements on the basis of the functional-process and (2) those choices involved in the actual creation criteria and fiscal criteria stipulated in Chapter One. of local units (e.g., municipal charter making). Both The assessments of St. Louis made as a result of this levels are constitutional in nature, insofar as both study are then compared to the evaluations of oth- levels define mechanisms for governing the proc- ers. Finally, the chapter takes up some of the prob- esses of government. It is at these constitutional lev- lems that are currently being addressed in St. Louis els that a framework of metropolitan governance is County, and the possible solutions available. created and sustained. The content of a local government constitution FUNCTIONAL DIMENSIONS: was analyzed, in Chapter Three, in four basic parts: HOW FRAGMENTATION WORKS (1) rules of association that allow local citizens to Certain generic processes and arrangements are create specific types of local units with general or characteristic of St. Louis County. Rather than being specific powers; (2) fiscal rules that empower local simply a collection of ad hoc accommodations, as units to raise specific types of revenue; (3) boundary the literature of intergovernmental cooperation often change rules, especially annexation rules that allow suggests, the processes and arrangements identified for adjustments in the boundaries of local units; and here are systemic: they fit together as the properties (4) interjurisdictional rules, especially contracting of a system that tends to function in a coherent and rules, that allow separate jurisdictions to enter into predictable manner. Each can be viewed as a work- various joint arrangements. ing part in the mechanics of metropolitan organiza- Rules of Association tion. These rules are the institutional foundation of A Local Government Constitution jurisdictional fragmentation, allowing citizens to cre- The jurisdictions found in St. Louis County are ate local governments, notwithstanding the legal fic- for the most part products of local self- tion that all local governments are creatures of the determination, as anticipated in the theory of frag- state. In fact, throughout most of the United States, mentation developed in Chapter One. The govern- municipalities and special districts tend to be crea- ance structure of the county, set out in state statutes tures not of state governments but of local communities. The constitution and laws of a state generally was the special position enjoyed for many years by provide legal tools to local citizens who use those the county government. Because annexation was tools to create local governments. treated as a judicial proceeding, the county govern- In St. Louis County, state law empowers local ment was able to enter the proceeding as an inter- citizens-and only local citizens-to create villages, ested party. The court was required, as a matter of third or fourth class cities, fire districts, school dis- judicial construction, to weigh the interest of the tricts, and a number of other special districts. The county against that of the annexing municipality. typical procedure involves citizen petition and refer- Usually, the county interest was deemed of greater endum. Cities over 5,000 in population may choose weight, enabling county government to exercise a de to incorporate under home rule charters, which en- facto veto over municipal annexation. Over the joy the protection of the state constitution. County same period, county government was able to de- government also operates under a home rule charter velop a legal and administrative capability to supply adopted by citizens. municipal-type services-especially police and A local government constitution that enables streets-in the unincorporated sections of the citizens to make constitutional choices by creating county. Potential competition from already existing (and occasionally abolishing) local units of govern- municipalities could be blocked in court. ment at their own initiative involves much more than In 1983, however, a decision of the Missouri obtaining the consent of the governed. This sort of Supreme Court deprived county government of its constitutional arrangement directly taps the creative special position in municipal annexation proceed- energies of citizens; it allows those most directly af- ings.' For the first time in decades, the county gov- fected by local public problems to take constitutive ernment had to take its case to local citizens. Thus action to address those problems. Instead of peti- far, the results have been mixed, but county govern- tioning the government to take some action, citizens ment has won more often than it has lost. In the can literally create their own government or govern- landmark case, however, the small City of Town and mental arrangement. Country more than doubled its size. Thus threatened, potentially, by a new wave of annexations and Fiscal Rules incorporations, county officials have called for a These rules specify the revenue sources avail- radical restructuring of local government throughout able to different types of local units and provide for the county. St. Louis County again faces a basic procedures to increase tax rates. The procedures re- constitutional choice, and the citizens of the county quire recourse to citizen approval-either by a sim- may have to decide. ple majority vote or in some cases by a two-thirds Rules for Constituting lnterjurisdictlonal majority-in a popular referendum. Having created Arrangements their units of local government, citizens continue to hold the purse strings. The local public economy is These rules are very permissive. The basic con- one where local officials, as political entrepreneurs, straint is that all jurisdictions that are party to some find it necessary to persuade the local electorate that arrangement must have separate legal authority to engage in the relevant activity. Although this would tax rate increases are desirable. This requires local officials to make a case that increasing costs and/or prevent county government or a municipality from potentially greater benefits justify greater expendi- providing public education, for example, it does not ture. seriously limit the development of both bilateral and multilateral arrangements among units of various Boundary Change Rules types. Generally, all that is required to undertake interjurisdictional activity is mutual consent between These rules are an especially sensitive part of the elected officials of the respective jurisdictions. the local government constitution. St. Louis County Contracting authority extends to agreements with was among the first in Missouri to adopt (by means private parties as well. of special state legislation) a municipal annexation rule requiring approval by concurrent majorities in The Use of Special Legislation the annexing city and the area to be annexed. This Important parts of the local government consti- rule has now been adopted statewide. A principle of tution in St. Louis County are unique to the county. self-determination thus came to govern municipal Although some aspects derive from provisions of the expansion as well as formation. Citizens can be ex- state constitution and others. from statewide provi- pected to approve a proposed annexation when. they sions of general law, other aspects derive from the believe they will be made better off. use of what might be considered special legislation The most controversial aspect of state annexa- that, in effect, applies only to St. Louis County. This tion law as it applied in St. Louis County, however, device allows the county delegation to the state legislature, composed of 7 senators and 31 representa- modification in the fundamental rules of local gov- tives, all elected from separate districts, to modify ernance. the local government enabling rules solely for St. In the realm of public education, the Cooperat- Louis County. Functionally, when substantial local ing School Districts of the St. Louis Suburban Area agreement regarding a change in rules emerges, the (CSD), an organization discussed below for its con- local delegation becomes a mechanism of constitu- tribution to coordinated service production, serves a tional decisionmaking for the county and all its juris- purpose similar to the municipal league with respect dictions. Yet the local delegation is confined to the to school districts. Professional associations in other first level of constitutional choice noted above-the functional areas, especially police and fire, also choice of enabling rules. The second level of consti- serve as forums for maintaining and adjusting basic tutional choice is reserved to citizens, acting with re- rules in their respective fields. spect to their particular communities. It is they who Public-private relationships are the focus of a decide whether to create a municipality or special number of groups, including the East-West Gateway district, increase a tax rate, or to consolidate two or Coordinating Council, Bi-State Development Corpo- more jurisdictions. ration, Civic Progress, and Confluence St. Louis, The ability to obtain special legislation for the which tend toward a broader city-county perspec- county allows local jurisdictions to resolve intergov- tive. These groups, however, have been less success- ernmental conflicts, or to attempt to do so, by ad- ful as sponsors of legislation than those associations justing the enabling rules. The creation of a county focused specifically on St. Louis County. The bilat- sales tax pool among participating municipalities and eral relationship between city and county is less for- county government, with exemption for those mu- mally organized than the multilateral relations nicipalities that chose to retain point-of-sale distribu- among local governments in the county. tion, for example, is an arrangement unique to St. Louis County. It was adopted pursuant to an agree- A Large investment in Representation ment worked out among the municipalities and the In addition to the direct participation of citizens county government. Subsequent efforts to change in metropolitan governance associated with incorpo- the distribution system, initiated by pool cities, have ration, annexation, charter revision, and tax refer- not been successful; however, county government enda, the majority of citizens in St. Louis County was successful in obtaining legislation that continues have invested in high levels of representation pool status for areas annexed by point-of-sale cities, through elected officials (see Chapter Two). The as well as for newly incorporated municipalities. county's 90 municipalities, containing some 600,000 residents, maintain a very low ratio of citizens to elected council representatives (aldermen or village MuRijurlsdictionai Forums for Discussion and Negotiation trustees), most of whom are elected from local wards. In all but two of the 26 villages and in many The local government constitution, despite being of the smaller fourth class cities, the ratio is less than formally rooted in state law, is primarily a product of 500 citizens per elected representative. In none of local decisionmaking processes. The local delegation these local government units does this ratio exceed to the state legislature provides a mechanism for im- 3,000. Of the larger third class and home rule cities, plementing agreements reached informally among only two have citizedelected-official ratios exceed- local actors. A number of multijurisdictional assoc- ing 5,000, and they only slightly exceed this figure. iations serve as forums both for the articulation of A total of 729 elected municipal officials (plus 144 diverse interests and for negotiations to arrive at lo- school board members) gives the county a sizable cal settlements of issues. Members of the local legis- infrastructure of local representation. lative delegation frequently appear as guests before Municipal representation contrasts sharply with these groups. The St. Louis County Municipal the seven members elected to the county council, League-representing nearly all cities and villages in each of whom must represent some 140,000 per- the county-is the broadest of these groups, supple- sons. Yet, for the incorporated area, members of the mented by an organization representing Mayors of county council afford additional representation be- Large Cities and a newly organized group for May- yond the municipal unit. State legislators, discussed ors of Small Cities, as well as by various subcounty above, provide still further representation through- forums both formal and informal. Proposed adjust- out the county, and in St. Louis City as well. In gen- ments in the local government constitution are in- eral, the system of representation is one that pro- tensely debated within these organizations. Substan- vides for redundant representation at differing tial consensus among affected groups-though not geographic scales. This redundancy enables various unanimity-is the prevailing norm for obtaining a communities of interest to be represented, allowing citizens with varying interests and opinions to find a small to employ a full-time executive officer are, public spokesperson, and facilitating the representa- well, too small. Professionalism then becomes an im- tion of interests that derive from interjurisdictional perative that drives local government reorganization spillovers. Together with the multijurisdictional fo- rather than a necessary tool to be employed as rums noted above, the system is one that seems to needed. encourage a vigorous public exchange of views with Part-time mayors are found not only in the respect to the basic issues of local governance. smallest municipalities but also in cities large enough to employ professional managers or administrators- Citizen Mayors as many of them do. Yet these same cities also elect A pervasive feature of local government in St. a mayor who takes an active leadership role in mu- Louis County is a reliance on part-time officials nicipal affairs. A citizen-mayor can serve a useful whose local government responsibilities are shoul- role even where professional administrators are also dered as a civic avocation rather than as a political employed, by holding professional employees ac- or professional career. Except for the county's larg- countable to the interests of ordinary citizens. est city, Florissant, which has a full-time mayor, and In an important sense, the citizen mayors inter- one community, Olivette, with a council-manager viewed in St. Louis County are themselves special- form of government, all of the county's municipali- ists. Aside from the fact that many of them are in ties have separately elected part-time mayors or, in fact career professionals in their full-time occupa- the case of villages, board of trustees chairmen who tions, citizen mayors are not so much functional spe- act as chief executive officer for the municipality.2 cialists as place specialists who specialize in knowl- The majority of these citizen executives hold full- edge of their own communities. Some serve their time jobs outside of local government. communities for a number of years. The focus of This report refers to these part-time chief execu- their specialization is more on specific time-and- tives as citizen mayors, stressing their nonprofes- place information than on abstract knowledge. They sional orientation in both a political and administra- tend to be well informed about service conditions tive sense.3 To be sure, the role of the mayor varies within their communities and about what is going on somewhat with the size of the municipality. The in adjacent communities and overlapping jurisdic- larger cities tend to employ a city manager or admin- tions, including school districts and county govern- istrator who has day-to-day responsibility for admini- ment. They also tend to exhibit a high level of stration. Smaller cities and villages tend to rely on a knowledge about the alternatives available within clerk for the same purpose. their local public economy, especially alternative A number of advantages accrue from the use of producers of services and possibilities for coopera- part-time officials. For the smaller cities and villages, tive efforts with other jurisdictions.4 Information is their use is, in part, a matter of economic necessity: an important and scarce resource in the provision the scale of operation does not warrant a full-time and production of public services. The close tie be- chief executive officer. The ability to rely on citizens tween information and responsibility that citizen- to do the job is essential to the economic viability mayors bring to their work can be viewed as a com- of the smallest units. The feasibility of using non- munity asset.5 specialists in this role is enhanced by the tendency Nevertheless, after a city reaches some mini- of the smallest units to contract out for the produc- mum size-perhaps around 50,000 people-part- tion of all, or nearly all, the services provided. The time mayors, able to function effectively as a chief chief executive officer, instead of functioning as a executive, become infeasible from the sheer scale of supervisor, is more nearly a procurement officer for the enterprise. The advantages of using a part-time what amounts to a neighborhood. The experience of mayor tend to decrease as a city increases in size many communities in St. Louis County is that highly and heterogeneity. Therefore, the use of citizen specialized professional skills in public administra- mayors in the position of chief executive is probably tion are not essential to this task. limited to relatively small-though not only to the Indeed, performance may be enhanced by the very smallest-municipalities. qualities that citizens bring to the office. This is an observation that runs counter to the conventional The Use of Overlapping Jurisdiction8 and wisdom in an age of professionalism. To be nonpro- Speclal Dlstrlcts fessional in the eyes of many observers is to be an Overlapping jurisdictions and, especially, special amateur, often assumed to be the equivalent of un- purpose governments are frequently viewed, at least professional. The mere presence of part-time offi- in the orthodox public administration literature, as a cials may often be treated as a deficiency to be cor- source of duplication, confusion, and inefficiency. rected. From this perspective, local governments too Overlap, nevertheless, is one of the major tools of metropolitan organization in a fragmented metro- tend to require quite different maintenance routines. politan area. Without overlap, the frequent charac- Planning is even more specialized by highway type. terization of highly fragmented metropolitan areas as Larger jurisdictions are able to provide services Balkanized would become a more accurate depic- that respond to a wider community of interest-such tion. With overlap, however, a metropolitan area is as arterial streets in the case of county government. able to address both common and diverse interests Smaller jurisdictions are able to address concerns simultaneously. that are smaller in scale, affecting a more limited The major overlapping jurisdiction in St. Louis community-such as the maintenance of and control County is county government. Two special districts- of access to neighborhood streets in the case of sub- one for sewerage, another for the zoo and muse- divisions and villages. ums-overlie both St. Louis City and County. The In education, the Special School District ad- countywide Special School District (overlying the 23 dresses a set of concerns that are more specialized regular school districts in the county) provides for than is the case with general education. Unlike pref- education of handicapped students and for voca- erences for general education, preferences for spe- tional-technical education. The most significant sub- cial education may not aggregate well on the basis of county districts are the 23 school districts and 25 fire residential location. Because of the considerably protection districts, many of which overlap smaller higher expenditure per pupil in special education, municipalities, but also cover all of the unincor- residents cannot expect to secure a response to their porated area of the county. A few miscellaneous dis- demands by relocating to particular school districts. tricts exist for such purposes as street lighting and No single district has much incentive to specialize in roads. Irl addition to formal units of local govern- the education of handicapped persons. Provision for ment, there are in excess of 400 organized subdivi- special education by a countywide district makes sions, both within municipalities and in unincor- sense in a way that countywide provision for general porated areas, which are especially active in the education does not. provision of residential streets and a wide array of The use of overlapping jurisdictions in fire serv- street services. ices is somewhat different still. In the incorporated Two basic types of overlap can be distinguished. area of the county, fire protection districts overlap One is overlap for the provision and/or production small municipalities which, separately, might not be of specific services. Special districts often contribute able to economically maintain a fire station. In the to this sort of overlap. The other is overlap for the unincorporated area, fire districts are a smaller scale purpose of governance and conflict resolution. The alternative to provision by county government. Fire county as a legal unit provides this latter sort of overlapping jurisdiction to arrange for the govern- districts thus augment both municipal and county ance of the multiplicity of local governments, includ- governments, though in different ways. ing county government, in St. Louis County. No In terms of metropolitan governance, one of the such local unit exists to arrange for conflict resolu- basic organizational features that distinguishes St. tion in the relationship between city and county. Louis County as a metropolitan county from St. Streets and highways provide a good illustration Louis City and County as a metropolitan area is the of the basic principles of overlap in service provi- fact that the county is a general purpose local juris- sion. Strictly residential streets are provided for in diction that overlies other general purpose jurisdic- most instances by subdivisions and small municipali- tions. No general purpose jurisdiction, other than ties; collector streets, by municipalities; arterial the state, overlies both St. Louis City and County. streets, by county government; and major thruways, The significance of the county as a jurisdiction goes by the state highway department. Specialization, beyond county government and is rooted in the fact rather than duplication, is the basic feature of this that a county is a basic legal subdivison of the state. arrangement. Different interests are represented in As such, it is a convenient unit for legislative pur- the process. Subdivisions tend to impose access re- poses. Yet the county, unlike the city, does not strictions that control the flow of traffic through resi- preempt the formation of general purpose subunits. dential communities. County government, on the The county provides a set of legal boundaries within other hand, is oriented toward facilitating traffic which a variety of interlocal relationships can be cre- flow, and preempts municipal regulation on a ated and adjusted. No similar unit exists to adjust countywide arterial road system. That municipal and the interests of city ahd county. City-county separa- county governments engage in some very similar option deprives the separated units of a common juris- erations, such as surface patching, does not imply diction in which to resolve their differences and ad- duplication. Different types of streets and highways dress interdependencies. Separation of Provision and Production unit, but to evaluate its performance and remain alert to alternatives.10 One cannot readily understand the functional Separation of provision and production is also arrangements for metropolitan organization in St. pronounced in street services, but the independent Louis County without reference to the distinction production units tend to be private producers. Virtu- (introduced in Chapter One) between provision and ally all subdivisions and most small municipalities production of local public servi~es.~Provision refers contract out privately for nearly all street services, in general to taxing and spending decisions, deter- although some subdivisions contract with an overlap- mining appropriate types of service and levels of ping municipality for a few services (e.g., snow re- supply, as well as arranging for and monitoring pro- moval or sweeping). While this study did not collect duction. Production, on the other hand, denotes the data on the private production side of the local process of transforming inputs into outputs. Provi- sion activities include raising revenue and deciding streets economy, the number of independent provi- how to spend it, while production is concerned with sion units and relatively limited economies of scale making a product or rendering a service in accor- in most street services would point to the likelihood dance with provision standards. of a highly competitive market. The separation of provision from production in Separation of provision and production is much St. Louis County can be illustrated with reference to less prevalent in fire services, although there are in- police services. Ninety jurisdictions, including the stances of contracts between municipalities and fire county government, have authority to provide police districts. The reason is that the fire district arrange- services.' Eighty-nine of these jurisdictions actually ment tends to substitute for a contracting relation- do make some provision for local po1ice.a The total ship. While a municipality may continue to function number of police departments-those units that pro- as a provision unit for police services, but contract duce police services-is 65 .@ Sixty-two municipalities for production, the same community may receive and the county government have organized full-time fire service from an overlying fire district, which police departments. Two more municipalities main- functions as both provision unit and production unit. tain part-time departments. Twenty-four municipali- This relationship indicates a trade-off between con- ties that provide for police protection have not or- tracting and the use of overlapping jurisdictions as ganized units in-house to produce police services. provision units. A third alternative consists of coor- Seventeen municipalities have entered into contracts dinated service production, discussed below. with adjoining municipalities, while seven contract One of the most interesting results of the partial with the county police. Ninety provision units are separation of provision from production that occurs served by 63 full-time and two part-time production in St. Louis County is the existence of a large num- units. ber of pure (or nearly pure) provision units-small Does this analysis indicate that the number of municipalities that organize little or none of their provision units might just as well be 63 as 907 Not at own production activities, but instead contract out all. Provision and production are functionally dis- for a wide array of local services. Street services may tinct. Different criteria apply to the organization of be contracted with a number of private suppliers, provision and the organization of production. Provi- police with an adjacent municipality, refuse collec- sion units are based on the way in which individual tion with a private firm, and so forth. The orthodox preferences for services cluster in neighborhoods. Production units depend on limited economies of public administration literature has often treated scale and a more specialized division of labor. The such units as nonperforming and nonviable." two sets of criteria need not coincide. Provision units This conventional view neglects the distinction that choose to contract out for production are not between provision and production. In an argument inactive. They may be very active-as providers. A advanced by Anthony Downs.12 a pure provision degree of competition on the production side gives unit can be viewed as an ideal, insofar as in-house provision units some capacity to choose a production production may tend to distort planning and pro- unit from among alternative producers. Evidence of curement decisions, introducing a producer's bias this sort of comparison shopping is found in St. that inhibits the representation of citizen-consumer Louis County policing, as municipalities occasionally interests. Our interviews with part-time mayors and shift contracts from one production unit to another. village chairmen of small municipalities in St. Louis Local officials in a provision-only unit (one that or- County indicate a generally high level of provisioning ganizes no production in-house) function mainly as activity-especially pronounced in those cases of procurement officers, representing citizen-consumer pure provision-including comparison shopping interests. Their job is not to supervise a production among alternative vendors. Yet even in the many provision units that have At less than an areawide scale, numerous joint organized their own departments for in-house serv- operations are undertaken for police dispatch and ice production, officials frequently indicated aware- sharing of investigative officers. The sharing of mu- ness of and appreciation for the alternatives af- tual aid among patrol officers crossing jurisdictional forded them by a possible separation of provision lines is found throughout the county. and production. Without using this particular lan- Fire protection also exhibits substantial coordi- guage, they indicated that the availability of alterna- nated service production: tive suppliers of local services-adjoining municipali- Mutual aid agreements link all of the mu- ties, county government, and, in some services, nicipal and fire district departments in the private firms-gave them additional leverage when county, and ensure needed backup capabil- negotiating budgets and service delivery expectations ity or redundancy to respond in high de- with their own bureau chiefs. As more than one mand circumstances, such as large or nu- mayor and bureau chief phrased it, the option of merous fires. contracting out local service production keeps local producers "on their toes, " and therefore increases Many of these mutual aid agreements con- service responsiveness. tain first response provisions that delimit areas within one jurisdiction where a fire com- Coordinated and Joint Service Production pany from an adjoining jurisdiction will In a fragmented system of local government, respond immediately to a fire call. First- many occasions arise when the coordinated efforts response agreements add considerable flexi- of organizational units in different political jurisdic- bility to service delivery boundaries, en- tions become advantageous. Economies of scale in abling fire provision units to minimize the production of particular components of a service response times. (e.g., communications or training) often make joint As in policing, county fire departments production arrangements economically attractive. l3 maintain a joint recruit training program. St. Louis County is the locus of many such coopera- Equipment sharing is an important element tive ventures organized in every sector of public of service coordination in fire services. The service. Greater St. Louis Fire Chiefs' Association In policing, a number of components in the sup- sponsors the preparation of an annual Cata- ply of police services draw on coordinated or joint log of Apparatus and Special Equipment production: held by each of the departments, allowing An areawide Major Case Squad draws in- quick identification of specialized equip- vestigators from many different police de- ment that may be needed in an emergency. partments to bring personnel and expertise Fire administrators in many jurisdictions to bear on serious crimes. have established joint fire and emergency A countywide "Code 1000" plan provides vehicle dispatch centers serving several de- for rapid mobilization and deployment of partments, enhancing mutual aid capabilities officers from multiple jurisdictions in the and increasing the deployment of personnel event of natural or man-made disasters, civil for fire prevention and suppression activi- or labor disturbances, or any other occur- ties. rence requiring a large number of officers. In education, a number of organizations have The St. Louis County Police and Fire Train- been created by consortia of elementary and secon- ing Academy supplies recruit training for all dary education producers: of the police departments in the county. The Cooperating School Districts of the St. The Regional Justice Information System Louis Suburban Area supplies its members maintains a computerized data base for po- with extensive audio-visual capabilities, data lice related matters, affording on-line access processing, and joint purchasing of supplies to police dispatchers and, through network and equipment, as well as a forum for unit- lines, access to State of Missouri and FBI ing to develop-and lobby for-educational data bases as well. programs that require state action. An areawide 9 11 system of call-for-service w The Regional Consortium for Education and routing and dispatching is operated as a Technology supplies its members with com- joint venture by municipal departments and puter technology, software, training, and the county police. maintenance. The Special School District of St. Louis production seeks to reduce duplication as it builds County engages in highly coordinated serv- redundancy. ice delivery for mainstream students who require special education. Special district Public Entrepreneurship teachers work in the classrooms of each of Coordinated and joint service production does the 23 regular public school districts in the not happen spontaneously, but emerges from proc- county. Coordination also occurs in the di- esses of discussion and negotiation-efforts to dis- agnosis and evaluation of students for spe- cern common interests among diverse communities. cial education programs. The development of joint ventures among the multiple jurisdictions of St. Louis County is the work of County school districts have joined with the public entrepreneurs-those who take the initiative St. Louis City district to create the Volun- to propose ideas and carry the burden of ensuring tary Inter-District Coordinating Council to discussion, compromise, and creative settlement. implement a desegregation plan linking the The ability and incentive to exercise initiative is the city and county schools. key to entrepreneurship. The potential for entrepreneurship increases with the number of possible Highly coordinated activity is less important in sources of initiative. Counting the number of elected some service areas than in others. Street services, officials, police chiefs, fire chiefs, school superinten- for example, exhibit more "alternation" than coor- dents, directors of public works, and city administra- dination. Alternation refers to a type of mutual ad- tors or managers yields a rough measure of the po- justment among service providers or producers in tential for public entrepreneurship in a local public which different organizations divide up service re- economy. The greater the number of possible sponsibility according to area, clientele, or time pe- sources of initiative, the more likely entrepreneur- riod. In this way, street responsibilities are divided ship becomes.14 Greater entrepreneurship is one of among the county government, municipalities, and the advantages associated with area size when cou- organized subdivisions. Private contracting tends to pled with complexity. be more important than joint production for street Individual entrepreneurship is frequently exer- service producers. Nonetheless, equipment sharing cised in the context of a professional association in does occur, especially among smaller municipalities. St. Louis County. Such organizations as the Board The local chapter of the American Public Works As- of Governors of the Law Enforcement Officials of sociation organizes an annual joint purchase of road the Greater St. Louis Area, the Greater St. Louis salt. Fire Chiefs' Association, the Cooperating School One of the most interesting facets of research on Districts of the St. Louis Suburban Area, and the a highly fragmented metropolitan area, such as St. local chapter of the American Public Works Asso- Louis County, is the relative paucity of what is called ciation facilitate the work of public entrepreneurs by duplication of effort. Duplication, according to dec- bringing relevant parties together on a regular basis. ades of reform and consolidation proposals, is the Each instance of coordinated service production bane of a fragmented existence. Nevertheless, in noted above is a product of individual initiative in one of the most highly fragmented urban counties in the context of a local service-specific association. the nation, obvious and significant duplication is Local elected officials, working bilaterally and hard to find. Coordination and alternation are much through the auspices of the multijurisdictional fo- more common. In fact, coordinated service produc- rums discussed above, have also engaged in tion tends to occur in precisely those facets of serv- entrepreneurial activities that create functional links ice production where one would expect duplication. among jurisdictions. From our examination of St. Louis County, we con- Local public entrepreneurship, exercised clude that jurisdictional fragmentation is not synony- through voluntary associations, is apparently pre- mous with duplication. Far from fostering duplica- ferred in a fragmented system to imposition by tion, a fragmented system can work-with substantial higher authority. Local agreement is the essential success-to reduce it. condition. Although sometimes local agreements re- Moreover, fragmentation minimizes duplication quire ratification by the state legislature or by voters while enhancing backup capabilities through redun- in countywide or even statewide referenda, to view dancy. Duplication is a characteristic of service de- this process as one where a higher authority pre- livery and is rightly considered wasteful and ineffi- scribes to subordinates is inaccurate. Codification in cient. Redundancy, on the other hand, is a state law, in a number of cases, serves as a partial characteristic of service capacity and, within limits, guarantor of agreements made locally, helping to en- serves to enhance efficiency. Coordinated service sure that parties to an agreement maintain their participation through changes in political administration and County, requiring approval by concurrent ma- and other changing circumstances. Agreements nev- jorities in both jurisdictions.15 ertheless remain flexible and, for the most part, sub- Self-determination is also the basic principle in- ject to renegotiation when needed, with state ratifi- volved in most of the coordinated service arrange- cation of changes proposed locally. Having recourse ments that link different municipalities, fire districts, to state law may also constrain the occasional hold- and school districts. Nearly all of these are coopera- out among local jurisdictions that might seek a spe- tive, not coercive, arrangements. Departures from cial advantage in relation to others; absolute una- self-determination tend to be quite limited in scope nimity is not required to obtain local consensus. and, in the instances found in this study, tend to support cooperation. Cooperative arrangements that FUNCTIONAL EVALUATION involve a large number of independent actors may at some point encounter freerider or holdout prob- The discussion in Chapter One stipulated five lems.18 When this happens in St. Louis County, lo- functional-process criteria for evaluating the frag- cal governments look to the county charter or to mented system of organization in the St. Louis area. A review of the findings from this study in view of state legislation sponsored by the local delegation for these criteria can provide measures of both the the legal tools to constrain potential holdouts. For achievements and shortcomings of metropolitan or- example, minimum training standards for fire fight- ganization in St. Louis. Somewhat different conclu- ers were adopted by amendment of the county char- sions apply depending on whether the frame of ref- ter at the initiative of the Fire Chiefs Association erence is St. Louis County or St. Louis City and (see Chapter Five). This requirement provides a ba- County taken together. The county, as shown above, sis for cooperation in the training of fire fighters. In provides some valuable instruction in how fragmen- another example, the state legislature implemented a tation works. Yet, as always, there is room for im- statute applying only to St. Louis County that re- provement, both in relationships within the county quires municipalities with a population of 400 or and between St. Louis City and County. more to make provision for full-time police protec- Process criteria allow for an evaluation of function (see Chapter Four). This requirement helps en- tional arrangements-the patterned activities through sure that police cooperation is reciprocal, instead of which specific functions are performed-apart from allowing a community to depend on county police an evaluation of the outcomes of performance. The without having to pay for the protection they re- focus of evaluation is process, not rules, even though ceive. rules provide the framework within which functional Although self-determination is important as a arrangements are developed. Instead of taking rules base rule for organizing local governments, it works at face value, the question is what sort of process the only within limits. An important issue is how far self- rules produce. Do the rules of municipal incorpora- determination should be carried. When should the tion and annexation enable people to practice effec- rule of self-determination for each local government tive self-determination? Do multiple local jurisdic- be relaxed and an enforceable rule substituted that tions allow for meaningful citizen choices? Are applies in common to all relevant jurisdictions? In services and service components appropriately coor- some circumstances, a strict reliance on self- dinated? To what extent does competition among determination exacerbates conflict. For example, St. service producers affect service delivery to citizens? Louis County extended self-determination to the Do multiple local jurisdictions inhibit metropolitan county fiscal structure by means of state legislation problem solving? that allowed municipalities to choose whether to participate in a county sales tax pool or to collect the Self-Determination and Citizen Choice tax on a point-of-sale basis. The sales tax pool is Local self-determination is the core principle of basically a cooperative arrangement. Those who the local government constitution in St. Louis would have the most to contribute to a pool are thus County. It comes as no surprise that St. Louis scores able to stay out. An alternative arrangement, but well on this criterion. As the rules now stand, citi- one that would depart from self-determination, zens decide for themselves whether to form, or to would provide for limited sharing of sales tax reve- join, a municipality or special district, or to merge nues among all municipalities and the unincor- their local government unit with another. Citizens in porated area. Each community would then be re- unincorporated areas decide for themselves whether quired to pool some proportion of its revenues to be to incorporate or to annex to an existing municipal- redistributed on a per capita basis. At the same ity. Self-determination also applies to potential ad- time, all communities could remain, to some extent, justments in the relationship between St. Louis City point-of-sale. Such an arrangement might provide for more agreeable relationships among county mu- separately, or to annex to an adjacent city or village. nicipalities in the long run. Initial efforts to incorporate Chesterfield barely It is insufficient, by way of evaluation, simply to failed to obtain the required majority because of describe the rules of self-determination; as noted large majorities against incorporation in communities above, a process criterion must examine process. on the fringe of the proposed city (see Chapter The functional arrangements that actually operate Three). When incorporators cast their net less may deviate from those intended in the design of a broadly in a second attempt, the incorporation met rule structure. Such difficulties may indicate a need with approval by 75 percent of the voters. Although to reexamine the rules and make, at least, some this result represents a substantial consensus, the ex- marginal changes. There is some evidence of diffi- isting voting rule for incorporation did not initially culty in the processes of self-determination with re- lead incorporators to seek this consensus. spect to annexation and municipal incorporation. In a closely related matter, the opportunity for a Annexation procedures, discussed in Chapter community to vote to annex to an adjacent munici- Three, have recently undergone considerable pality is foreclosed by a valid incorporation petition. change in St. Louis County as a result of state court This ability to preempt an annexation vote seems to decisions. Although a judicial proceeding is still re- be inconsistent with a principle of local self- quired to complete an annexation, the principal ob- determination. stacle to municipal expansion is now an election One way to address these difficulties is to in- held in the area to be annexed, with a majority of crease the percentage of voters required to approve those voting required to approve the annexation. a municipal incorporation. The larger the percent- This requirement is consistent with local self- age, the less the probability that an identifiable com- determination. County government is no longer munity will be included in a new city when a major- able, in effect, to veto proposed annexations by op- ity of that community is opposed. Another possibility posing them in court. This, too, is consistent with is to establish a public body with authority to ap- local self-determination. In one case, however, an prove an incorporation proposal, and to modify the annexation was approved on the basis of one vote proposed boundaries on the basis of public hearings cast by a single resident (see Chapter Three). The or counter-petitions, but not with authority to deny annexation split a housing development, taking only the proposal (for this would diminish, not the undeveloped portion. In this case, a municipality strengthen, self-determination). Such a public body, may have been able to use its legal power to propose if its members are selected in a way that reflects a an annexation strategically, in order to avoid genu- broad spectrum of interests, could serve as a useful ine self-determination on the pan of an inchoate forum for exploring the reasonableness of proposed community. boundaries and allowing affected individuals and To address this possibility, annexation law could communities to air their views. Perhaps on petition, be modified to provide standards for judging the ap- such a body could be allowed to waive the first-in- propriateness of an annexation proposal (e.g., does time rule preempting an annexation vote or a village the proposed tract divide a housing development or incorporation vote. Care would have to be taken in subdivision?), and allow the county government, or the design of this body, however, lest it acquire the any affected citizen for that matter, to challenge an ability to substitute its own preferences for those of annexation in the judicial proceeding already re- the citizens affected." quired by law. Without such standards, it is difficult A legal framework based on local self- to see the point of a judicial annexation proceeding. determination appears to be an essential mechanism The process that led to the incorporation of the of metropolitan organization in a fragmented system. City of Chesterfield early in 1988 (see Chapter The logic of the system derives from, and depends Three), increasing to 9 1 the number of municipali- on, citizen choice. St. Louis has such a basic frame- ties in St. Louis County, suggested a potential prob- work, but its maintenance in view of changing cir- lem with the procedures for incorporating a new cumstances is no simple task. If one believes that city. The purpose of local self-determination is to citizen choices tend to be irrational, uninformed, or obtain the consent of local communities; yet satisfy- otherwise perverse, then any pattern of local governing this criterion does not depend on the consent of ment that emerges from this process might justifiably each and every individual affected. Incorporation of be viewed with a critical eye. If, on the other hand, cities, but not villages, requires only a simple major- one believes that citizens are capable of self-govern- ity of those voting. A proposed city of some 50,000 ment-that if well informed about local matters that people, nevertheless, may include a number of dis- affect them directly they can make rational judg- tinct communities. Some of these communities may ments taking into account both benefits and costs- prefer to remain unincorporated, to incorporate then the jurisdictional and functional arrangements created by citizens to govern a metropolitan area are the city and county delegations to the state legisla- deserving of careful study. ture can do so to some extent. If either the city or county takes actions detrimental to the other, it is Representation and Accountability more difficult to find common ground for resolution, The system of representation in the St. Louis due to the lack of a common decisionmaking facility area, as discussed above, features low citizen-to- in which to talk and a common jurisdiction in which elected-official ratios in the incorporated portion of to take action-except for the state. The lack of a the county and, throughout the area, a reliance on general purpose jurisdiction overlying both city and redundant representation. Traditional metropolitan county, as discussed above, has the effect of dimin- theory tends to view redundancy in representation ishing the representation of metropolitan interests. among a variety of elected officials as a source of voter confusion. Yet the system is one that provides Coordination for an open public arena with multiple opportunities The traditional theory of metropolitan organiza- for the expression of diverse views. tion, reviewed in Chapter One, expects jurisdictional Citizens in the incorporated portion of St. Louis fragmentation to be associated with functional frag- County have chosen to maintain relatively small lo- mentation-a lack of necessary coordination in the cal governments as the basic units of representation production and delivery of related services and serv- and accountability, usually subdivided still further ice components. Instead, however, jurisdictional into wards. Although not the only units of represen- fragmentation in St. Louis County coexists with high tation, they are basic units in the sense of providing levels of functional coordination among jurisdic- citizens with an easily accessible first recourse in tions. Whether a desirable or sufficient level of coor- part-time local officials and a readily usable method dination has been attained is a more difficult ques- of accountability in local elections. Accountability tion. has costs, mainly consisting of time and effort, but, It is useful, in tackling this question, to consider in small jurisdictions, the ease of individual citizen the differences in coordination among functional access to part-time municipal officials, who often areas and types of jurisdictions. High levels of coor- schedule meetings in the evenings, together with the dination exist among police and fire departments relative ease of challenging a local official in an elec- both in the delivery of direct services to citizens and tion, tends to minimize those costs. From the stand- in the production of auxiliary services (see discus- point of an individual constituent, both representa- sion of coordinated service arrangements above). tion and accountability are more costly in larger Between the two, more police coordination is in jurisdictions with higher citizen-to-elected-official ra- auxiliary services and more fire coordination in di- tios, often requiring group organization to be effec- rect services. This difference can be explained by tive. In a redundant system of representation that differences in the nature of the two services. Some rests on a sizable infrastructure of elected officials in coordination is also observed between fire districts small jurisdictions, however, the more costly chan- and municipal police departments in the area of nels of representation associated with overlying units communications, but less so. A high level of coordi- can be reserved for a limited range of issues. nation is also found among regular school districts, Municipal officials also provide their citizens but it is almost entirely related to the production of with redundant representation in relation to county auxiliary services. Direct services to students are government, as well as other jurisdictions. The ad- handled separately by each district, except for the vantages of small municipal jurisdictions thus extend education of handicapped students. In this service to citizens' relationships with other units of govern- area, a high level of coordination is found between ment. To be represented by one's mayor in these the county-wide Special School District and each relationships can be a significant advantage. In this regular school district in the delivery of services di- sense, the primary unit of representation-a munici- rectly to students. pality-acquires significance within a secondary Less coordination is found in street services. In- unit-the county, and perhaps the state. stead of coordination, county and municipal juris- County government-including the county coun- dictions exhibit patterns of alternation, each taking cil, the county executive, and the county charter- separate responsibility for some separable piece of provide representation for countywide interests, as service responsibility. In the incorporated portion of does the county delegation to the state legislature. the county, county government is limited to the de- Similar mechanisms to represent interests that are livery of services on the countywide arterial street metropolitanwide or simply city-county in scope are system. Just like coordination, alternation inhibits not as readily available. City-county separation pre- duplication of effort. Horizontal coordination among vents the county council from filling this role, but municipalities in the production of street services is much less extensive than in police and fire, but the gram (see Chapter Seven for details) is built on co- need for coordination in direct services is also less. operative arrangements among county school dis- Coordination in auxiliary street services is only be- tricts and the city district. St. Louis City also ginning to emerge (see the discussion of cooperative coordinates bilaterally with some of the county mu- purchasing of road salt in Chapter Six). There is also nicipalities on its borders. For example, the City of a significant degree of coordination between munici- Wellston has developed cooperative arrangements palities and subdivisions that provide residential with St. Louis City's community development street services. agency. The least prevalent form of coordination is be- Less coordination occurs between the city and tween overlapping jurisdictions engaged in the pro- county with respect to economic development. Com- duction of different types of services, namely, be- petitive rivalry, rather than cooperation, seems to tween school districts (or fire districts) and dominate in this area. Clearly, the city and county municipalities. Coordination between school districts compete for the location of major commercial, in- and municipalities seems to be greater when the dustrial, and recreational facilities. As discussed in boundaries of the two jurisdictions are coterminous, Chapter Eight, the county has enjoyed impressive or nearly so. The most common types of coordina- growth while the city has experienced two decades tion are related to recreation services and use of of economic decline. school facilities. Even less apparent is coordination An explanation of the varying degrees of coordi- between fire districts and municipalities. Although nation found in different functional areas and there may be little actual need for coordination among different jurisdictions must first take account (aside from street level coordination between police of the nature of the service supplied. Police services and fire units), there is still a possibility of much relate to phenomena that are mobile, while street greater coordination-illustrated by the use of con- repair relates to facilities that are fixed in place. Fire solidated public safety departments in some munici- services have unique peak load problems. Size of palities. In fact, one of only two public safety depart- jurisdiction is also important. Smaller jurisdictions ments in the county was dismantled when the City of have a greater need to engage in coordinated activi- Town and Country doubled its area by annexation ties of all types in order to secure economies of and elected to obtain fire protection by contract with scale. A jurisdiction the size of St. Louis City or an adjacent fire district. County, on the other hand, has exhausted most of Other types of coordination also exist, but lie its potential economies of scale. outside the four service areas studied for this report. The history of St. Louis County indicates a fairly Coordination occurs between county and municipal continuous search for new areas of coordination. governments in relation to tax collection, assess- This implies that, at any point in time, the full po- ment, and licensure, for example. Other types of co- tential for beneficial coordination has not been real- ordination may occur that the study did not un- ized. There is no reason to believe that coordination cover. among jurisdictions in the area has reached a satura- The discussion thus far has taken St. Louis tion point. Purchasing is an area of potential coordi- County as the frame of reference, but expanding the nation for nearly all services, but is most extensively discussion to include St. Louis City changes the pic- practiced by school districts. Municipalities could ture only in its details. Although the city is function- perhaps explore the possibility of greater coordina- ally more independent than any of the county mu- tion in this area; the county government could facili- nicipalities and school districts, it is still an active tate such an effort. Street services generally seem participant in many of the functional arrangements ripe for greater cooperation; perhaps, again, the for coordinated and joint service production organ- county government could assume a leadership role. ized in the county. Thus, St. Louis City is a member It is also important to realize, however, that if of the Major Case Squad that investigates serious coordination among jurisdictions organized by place crimes and is a party to mutual aid agreements with is imperfect, so also is coordination among adminis- county fire departments. Police training is currently trative agencies organized by function within a single handled by a city-county academy, but this will soon large jurisdiction. A large-scale metropolitan govern- change when the county opens its own police-fire ment substitutes fragmentation among functional training academy. The city's school district is an as- agencies for fragmentation by place. Instead of frag- sociate member of the Cooperating School Districts menting the responsibility for each function among a of the St. Louis Suburban Area (CSD), allowing multitude of place-specific jurisdictions, a metropoli- them to participate selectively in the CSD's coopera- tan government fragments the responsibility for spe- tive purchasing program, but not to receive its other cific places or communities among a multitude of services. The St. Louis school desegregation pro- functional agencies.18 Coordination among large functional agencies with respect to small communi- private vendors. Another service area with consider- ties is not necessarily better than coordination able competition is trash collection, in which a num- among small local governments with respect to spe- ber of private producers operate. cific functions. Market competition among producers is gener- In this context, the capacity to continue to seek ally thought to be beneficial. Fragmentation does out new and better mechanisms of coordination, not automatically engender this sort of competition, and to adjust existing mechanisms to changing cir- however. Most of the competitive arrangements in cumstances, becomes perhaps the critical coordinat- St. Louis County are found in those areas where ing capability in systems of public organization. In provision units are too small to produce a service these terms, St. Louis County (and to a lesser extent in-house and capture sufficient economies of scale. the city and county together) performs well. The Most of the services delivered to most of the coun- process of public entrepreneurship, discussed above, ty's population are not directly affected by such is one that works in a more or less continuous man- competition. ner to improve coordination where it is perceived to Two explanations for the lack of competitive ar- be advantageous. It should also be noted that every- rangements can be advanced. thing need not be coordinated with everything else. The problem of coordination is first to identify those (1) Municipalities tend to move to in-house aspects of otherwise separable activities that can organization of service production as soon as benefit from coordination, and then to seek the economies of scale permit. Contracting out means to achieve it. This sort of process is ongoing seems to be viewed as a second best alternative. in the St. Louis area. (2) Unlike Los Angeles County,'g for example, St. Louis County government has not ac- Competition tively sought out service contracts with munici- The principal evidence of competition among palities except in relatively limited functional the producers of public services in St. Louis County areas. consists of (1) widespread use of private contracting by small municipalities and subdivisions and (2) a Interviews with municipal executives indicate that degree of shopping for public producers by small the county government is not considered to be a municipalities. In the service areas studied, the strong potential competitor with municipal depart- greatest degree of competition is found in the pro- ments. In part, this lack of interest on the part of duction of street services, but it is limited mainly to county government may be due to its having concen- very small municipalities and street providing subdi- trated on its role as a de facto municipality in the visions-those provision units that contract out pri- unincorporated areas. Now that the county govern- vately for street repair and other street services (see ment's municipal service domain is no longer legally Chapter Six). Private schools clearly compete with protected from municipal annexation, county offi- public schools for students, but it is unclear to what cials may perceive a greater advantage in pursuing extent this competition, in the absence of a voucher municipal contracts.20 If so, the competitive envi- system, constrains public school districts. Consider- ronment of local service production could improve. able state aid, however, is apportioned to districts by Another sort of interlocal competition often student enrollment; thus, the loss of students does thought to characterize fragmented metropolitan ar- affect school revenues and, at the margin, may or eas is competition among municipalities (and among may not expose public schools to meaningful compe- school districts) for residents and businesses. To tition (see Chapter Seven). Some competition exists some extent, competition for commercial and indus- in the area of police protection, but only in relation trial taxpayers is found among county municipalities to those municipalities too small to produce their (see Chapter Three). Although some villages and own police services. The available contractors are cities seek to keep commercial and industrial uses of other governments-adjacent municipalities and the property out of their jurisdictions, others actively en- county government. Contract municipalities show courage them to locate there. Competition of this evidence of shopping around-by occasionally sort has been especially prominent between the unchanging suppliers (see Chapter Four). Fire services incorporated county and St. Louis City. Competition appear to be the least subject to competition, with for residents, however, is more difficult to docu- only a few fire service contracts among municipali- ment. Even if such competition occurs, it is unclear ties and fire districts (see Chapter Five). However, to what extent it affects the tax price of services and, there is greater competition in emergency medical therefore, the efficiency of municipal service pro- services, for which a number of municipalities use ducers.2' Metropolitan Problem Solving ever, can be viewed as a metropolitan problem that has not been effectively addressed on a metropolitan If the frame of reference is St. Louis County, basis. To some extent, economic growth in St. Louis metropolitan problem solving is active and ongoing. County has occurred at the expense of St. Louis City The existence of the county as a legal subdivision of (see Chapter Eight). the state that may embrace autonomous subunits, including both general purpose and special purpose FISCAL EVALUATION governments, is associated with the formation of a In addition to process criteria, this study em- wide range of countywide organizations that seek in ployed two fiscal criteria, that is, criteria that can be various ways to combine the efforts of multiple juris- used in an evaluation based on fiscal relationships: dictions and solve problems of interdependency. economies of scale and equity. Neither attribute can The countywide organizations include county gov- be directly measured with fiscal data alone, but ernment, the County Municipal League, Cooperat- would require performance data as well. Fiscal rela- ing School Districts, and numerous professional as- tionships do, however, offer indications of perform- sociations (e.g., police chiefs and fire chiefs), as ance capacity. The analysis of these fiscal data was discussed above. The usual process is for one or an- described mainly in Chapter Eight. The results are other countywide organization to raise an issue and briefly summarized below. provide a forum for discussion and resolution; formal enactment of rule changes needed to implement Economies of Scale a local consensus is then obtained through some The basic question here is the extent to which combination of changes in state law (sought by the there are remaining or uncaptured economies of county delegation), the county charter (amendments scale in the production of services in the county. approved by the voters), and municipal (or fire dis- Chapters Four through Seven each took up the trict or school district) charters or policies. question of uncaptured economies of scale with re- Not all problems addressed are fully resolved. spect to the four service areas studied. Evidence that The resolution of problems frequently requires a relatively slight economies of scale remain to be cap- considerable investment of time and effort on the tured by the smaller jurisdictions that provide police part of public entrepreneurs. The greatest degree of services is presented in Chapter Four. Similar evi- success in countywide problem solving has been with dence is found with respect to fire services in Chap- respect to service delivery by established local gov- ter Five. No evidence of remaining economies of ernments. Once agreeable arrangements are worked scale is found in street services (Chapter Six) or in out, interlocal service coordination is for the most public education, except perhaps for the smallest part stable. Less stability is apparent with respect to high schools (Chapter Seven). In Chapter Eight, a local revenue-raising arrangements. The settlement search for evidence of remaining economies of scale that divided the county into point-of-sale and pool when all services are considered together turned up municipalities for the purpose of sales taxation, nothing. while still in effect, has not quieted those pool cities The absence of remaining economies of scale, that oppose point-of-sale distribution (see Chapters except for some slight economies in police and fire Three and Eight). Even less stability has been protection, can be explained by the widespread use achieved with respect to municipal annexation and of coordinated service arrangements, special disincorporation, but this area is one in which the rele- tricts, and both public and private contracting, vant parties are presently hard at work (see discus- created as an organizational overlay to supplement sion of current problems below). One area in which the basic set of municipal and school district juris- there has been little countywide effort is fiscal redis- dictions. Intergovernmental cooperation has appar- tribution aimed especially at the county's distressed ently left little additional cost savings to be realized communities (see discussion of fiscal evaluation on from municipal consolidation. This is a general con- equity criteria below). clusion with respect to the county as a whole, and Shifting the frame of reference to St. Louis City should not be interpreted to mean that further cost and County, the record of metropolitan problem savings could not be gleaned from additional coop- solving is somewhat less impressive, but still notable. eration, contracting, or consolidation in particular The creation of city-county special districts has been instances. The method used to search for potential the principal vehicle of metropolitan problem solving economies of scale is to compare service costs in at this level. These efforts include a metropolitan small municipalities and special districts with those district organized to support the St. Louis Zoo and in larger jurisdictions. It still may be possible to im- other cultural activities and another to provide sew- prove efficiency and productivity in specific jurisdic- erage. Economic development within the city, how- tions. Widespread jurisdictional consolidation alone, however, would be unlikely to result in substantial Note, however, that this approach does not directly efficiency gains overall. address the fiscal disparities among jurisdictions that are the focus of concern in this report. It is also sig- Equity nificant that the desegregation plan was developed in Evidence of equity problems in St. Louis County the shadow of a federal court suit and the strong is most compelling in the circumstances of a few possibility of a judicially imposed solution. small, poor jurisdictions, including both municipalities and school districts (see Chapter Eight). Fiscal OTHERS' EVALUATIONS disparity among school districts, while exhibiting less Although no new proposals for areawide reform range than among municipalities, is more likely to be have been presented to voters in St. Louis since patterned along lines of race or wealth. The greater 1971, reform has remained on the agenda of county dependence of school districts on the property tax is government and of interest groups, principally the one source of this greater inequity. Despite the seri- business community through its organization, Civic ousness of the fiscal stress in particular jurisdictions, Progress. In 1982 a report to the City-County Task the magnitude of acute fiscal problems measured on Force of Civic Progress stated that: a countywide basis is not overwhelming in its propor- Governmental structures and civic leader- tions. The financial wherewithal to address the ship in the St. Louis area are severely frag- county's most serious fiscal disparities would seem to mented. lie within the county's means. Fragmentation results in an inability to act An interesting question-one that this study can- decisively on major problems facing the re- not fully answer-is why local governments in the gion. county have not explicitly addressed and solved the There is no coordinated system for financ- equity problem. Thus far, the issue has been raised ing areawide needs and functions, particu- most prominently in an indirect manner through the larly those related to economic develop- efforts of pool cities to obtain sales tax revenue from ment. point-of-sale cities. Disparities among school districts have been addressed entirely through statewide leg- Many local governments are facing fiscal islation. County government has no locally funded crisis and may not be able to maintain es- programs of targeted assistance for fiscally distressed sential services. communities. The report went on to outline a variety of op- Inequities between St. Louis City and County tions to alleviate what it viewed as serious metropoli- may present a greater challenge. One difficulty is tan problems resulting from governmental fragmen- that differences among communities in their eco- tation, recognizing that "consolidation-merger and nomic conditions seem to be as great within the city annexation do not currently appear to offer appro- as within the county. These differences, however, priate, politically feasible approaches to addressing are not reflected in fiscal data aggregated for the city the area's problems," but that additional special dis- as a whole. To compare the city as a whole with the tricts, reentry of the city into the county, and "any county as a whole is somewhat misleading. In reve- other plan" might be worth consideration.22 nue terms, the city as a whole is not especially disad- In 1985 the St. Louis County Annexation Study vantaged. In economic development terms, the city Commission issued a report on local government in as a whole is greatly disadvantaged. Yet, as commu- St. Louis. In a finding regarding "problems of munities, both the city and county are too heterogene- nicipal government disparity," the commission cited ous for meaningful comparison. The equity issue both positive and negative aspects of the system. On cannot be satisfactorily addressed on a city-county the positive side, the commission said: basis without disaggregating revenue and expendi- The 90 municipalities, 24 fire districts and ture data within the city, a task that lies beyond the various other special purpose service districts, scope of this study. Targeted assistance to specific provide a great diversity of living environments communities is as essential within the city as it is in St. Louis County. The variety in size, devel- within the county. Within the county, however, opment character, and local government organi- there are potential recipient jurisdictions that corre- zation is viewed by some as a positive attribute. spond closely to those communities in need. Citizens have greater choice in the selection of a The one major effort that has been made to ad- community which reflects their personal views dress inequities on a city-county basis is the school and needs. They can find full-service cities and desegregation plan (see Chapter Seven). By giving "no service" municipalities; small, neighbor- students opportunities to transfer between school hood-type governments or larger, more central- districts, greater equity of treatment can result. ized governments; municipalities with extensive park and recreation programs and others that tion." The commission could exercise a veto over offer free, twice-a-week, backyard trash pick- any proposal it found unwarranted, preventing the up .23 proposal or similar proposals from being considered Two major negative aspects were cited: inequi- again for two years. The commission could also "in- ties in access to revenues caused principally by the itiate proposals for incorporation, annexation, sales tax distribution formula and the resulting reve- merger, or consolidation, and modify proposals sub- nue differences between pool and point-of-sale mu- mitted to it." The legislature declined to approve nicipalities, and inefficiencies in service delivery this proposal in 1985. caused by many small service providers. While the Another group active in proposing changes in lo- commission acknowledged that "point-of-sale cities cal government organization has been Confluence deserve some special considerations due to the St. Louis. Confluence emerged from activities of the added service costs and other disadvantages associ- local Danforth Foundation and its Leadership St. ated with the commercial development," and that Louis program, an organization similar to the Citi- "incentives are needed so that commercial and em- zens League in the Twin Cities Area. Confluence ployment generating development will occur," it task forces have studied a number of issues of re- found the spread between revenues accruing to the gional concern since its founding in 1983. One of average pool city ($49 per capita in 1983) and the the concerns addressed by Confluence is the num- average point-of-sale city ($1 15 per capita) to be ber of governments in the area. overly large and to represent revenue importation by The "Too Many Governments?" Task Force of some point-of-sale cities. Confluence St. Louis recently issued its final report, reiterating many of the findings of the 1982 Civic With respect to inefficiencies, the commission found Progress Task Force, supporting full incorporation that and governmental consolidation in the county and recommending reentry of the City of St. Louis into the problems resulting from the disparity in St. Louis County.24 The Confluence plan proposes revenue sources are magnified by the costly du- the following structural changes: plication of services and inefficiencies resulting Incorporate all of St. Louis County and from many small service providers. Proponents clearly divide responsibilities between mu- of the current local government structure have nicipalities and county government. argued that citizens have the right to those inefficiencies so long as they are willing to pay for Enlarge, merge, and form municipalities in them. St. Louis County until each has 25,000 to 75,000 residents. The annexation study commission went on to state, however, that "voters in some of the less 8 Reform revenue structures so that each unit wealthy municipalities may ultimately reject assum- of government has enough revenues to pro- ing even greater tax burdens" as "inflation driven vide needed services. increases in municipal expenditures force repeated Plan for the reentry of St. Louis City into returns to the voters for tax increases." This voter the county. rejection would, in the commission's view, lead to a One proposed change, creating a two-tier struc- reduction of basic service levels in some municipali- ture of local government, would significantly reduce ties. This would be a problem of countywide import county government responsibilities, eliminating its because "crime, traffic congestion, the blighting in- role as an urban service provider for unincorporated fluence of deteriorating buildings, and health haz- parts of the county but retaining a number of ards are but a few of the potential municipal prob- countywide service responsibilities. At the same lems that do not recognize municipal borders." time, municipal consolidation would, in Con- Based on its findings regarding municipal dis- fluence's view, ensure that municipalities would be parities and the threat posed to the county by new of sufficient size to promote efficiency through annexation activities, the study commission recom- economies of scale. The revenue restructuring, in- mended legislation to the Missouri legislature that cluding the possibility of new revenues through a would have established a boundary review commis- countywide earnings tax andlor revenue sharing sion for St. Louis County. The legislation envisioned from the state, together with the possibility of redis- a boundary commission composed of county resi- tributing revenues from existing sources, would re- dents appointed by the governor that would review duce disparities among revenues currently available all proposals for "formation, consolidation, merger, to municipalities as well as provide additional reve- or dissolution of a municipality or service district, as nues for the newly created and reorganized units. well as proposals for annexation and disconnec- Reentry of St. Louis City into the county would "promote the resolution of city-county issues and while zoning, local taxes, and street and road main- create a strong sense of regional responsibility." tenance fell short of 'a majority with favorable rat- Confluence considered three mechanisms for ings.a Citizens also were asked for their ratings of instituting the recommended changes: (1) an the importance of local control over these services. intracounty board of freeholders, (2) a city-county For all services with the exception of libraries, the board of freeholders, and (3) a boundary commis- majority believing that direct local control was very sion. The first would require an amendment to the important was 60 percent or more, and was above Missouri Constitution and a statewide vote for adop- 80 percent for police and fire protection services. tion. The second is currently allowed under the state An advisory committee report to Civic Progress, constitution. The third could be created by an act of based partially on the survey findings, discussed lo- the state legislature. cal service delivery in these terms: The Confluence report found local government In the field of municipal service delivery, multiplicity in the St. Louis area to have both posi- the existence of a large number of jurisdictions tive and negative effects, but found the balance to increases potential access to local governments be negative: and can expand the market basket of service Although there are some advantages to mul- packages and taxing levels available to residents. tiplicity, the St. Louis metropolitan area has too In such a setting, some duplication and overlap many units of government, many of which are may occur, minor efficiencies may exist, and too small, and the advantages are completely economies of scale may not always be realized; outweighed by the disadvantages. Each of the but the evidence suggests that the cost to taxpay- facets of multiplicity (city-county separation, in- ers from these occurrences is likely to be rela- tracounty multiplicity, and special districts) pro- tively minimal. At the municipal level, deficien- duces negative effects. cies in the system of service delivery may only The advantages cited by Confluence included be technical or aesthetic. Costs to the taxpayers the accessibility of small governments to their citi- are likely to increase minimally or not at all, service quality is likely to be adequate, and citizens and the ensuing responsiveness of officials in zens may, in fact, feel better about their access such governments. Confluence considered the enhanced citizen choice afforded by a system of many to government. From a structural and organiza- small governments, but concluded that few citizens tional standpoint, the advisory committee did not find major problems related to most munici- actually exercise such choice. Disadvantages reported by Confluence included service difficulties pal service levels caused by the government proliferation and fragmentation found in St. Louis. caused by municipal boundaries, a perceived inability of the smallest local governments to provide serv- At the level of municipal service delivery- ices, the potential erosion of county revenues from provided local governments can finance them- incorporation and annexation, and an inability for selves and their services-no overwhelming, ma- the area to address regional service needs. Particular jor problems were found to exist, and it can be

areas of concern to Confluence were regional plan- , demonstrated that some advantages may accrue. ning and economic development (see Chapter In certain instances greater cooperation (and Eight). even consolidation) might result in improved service quality or greater efficiency, but no cur- CITIZENS' EVALUATIONS rent crisis was found. It would be difficult to In support of its 1982 report, Civic Progress build a compelling case for governmental reform commissioned a survey of citizens in St. Louis City at the current time based on a need for general and County.25 Along with many other questions, improvement of municipal service delivery capa- citizens were asked about their satisfaction with serv- bility through elimination of governmental frag- ices provided by local government. On a scale rang- mentation.27 ing from zero (extremely dissatisfied) to 10 (ex- The advisory committee found what others who tremely satisfied), survey respondents offered have studied service delivery in St. Louis have generally high ratings of local services in the area. found: local service delivery there appears respon- Fire protection, libraries, parks and recreation, and sive to local citizens, and citizens are quite pleased police services received ratings of extremely or with the services they receive. In particular, given somewhat satisfied from 80 percent or more of those the strong desire evidenced by citizens for retaining having an opinion, with ratings slightly higher in the local control of service delivery, the committee rec- county than in the city. Garbage collection and the ognized that citizens were unlikely to accept govern- sewer system received majority favorable ratings, mental reform proposals that would reduce local control, and that arguments for reform based on lo- Most recently a poll was conducted for the local cal service delivery weaknesses would not be com- group, Citizens to Streamline Government-a group pelling to citizens. The committee instead found that that initiated the petitions enabling the present the problems created by governmental proliferation Board of Freeholders to be formed (see the follow- and fragmentation were linked to areawide issues, ing section). The 2,500 county residents polled once particularly economic development. again gave very favorable ratings to local service de- In 1985 a local television station, KTVI, con- livery there.20 More than 90 percent reported fire, ducted a poll of St. Louisans that sought to assess police, and trash collection services to be good or public opinion on governmental reform.28 In re- excellent. Other services received at least majority sponse to the general question "Would you favor the favorable ratings, most nearly 70 percent or above. metropolitan consolidation idea for St. Louis Thus, when surveyed, citizens of the area con- County-St. Louis City?" 47 percent of the 647 adults sistently give their local governments high ratings on interviewed answered yes, 24 percent answered no, many aspects of local service delivery. They appear, and 29 percent were undecided. Opinions differed both from these surveys and from the reports of significantly between city and county residents. their elected officials, to be strongly wedded to the While 58 percent of city residents favored some system of numerous, generally small, local govern- form of consolidation, only 37 percent of county ments that they have created over the years. Yet re- residents indicated their support. Even though con- spected local bodies continue to argue to the consolidation was not favored by county residents, they, trary, finding the local government system to be a and residents of the city, were supportive of "the source of weakness, even in the area of local service establishment of special metropolitan districts to delivery where citizens assert that they are well provide public services common to the St. Louis served. County-St. Louis City area." The city and county On the service delivery issue, these findings are are currently experimenting with just this approach particularly difficult to reconcile with public opinion. for public hospital services. One might expect fur- How can a system that provides services rated so ther initiatives along this line in the future. highly by citizens be a source of serious concern for St. Louisans' unwillingness to embrace proposals local opinion leaders? Why argue for radical change for significant changes in the structure of local gov- of a system of service delivery that appears to be ernment appears to be based on their assessments working quite well in the eyes of service recipients? that the local government system, particularly in St. This study found no ready answers to these ques- Louis County, works well. An interesting feature of tions. It would seem that concerns for regional or the KTVI survey's findings was substantial agree- areawide issues, especially economic development, ment among city and county residents that "public have outweighed the evidence with respect to service services were 'better' to 'much better' in St. Louis delivery in the minds of reform advocates. These County, except for public transportation. " Eighty- concerns, and a perceived inability to address them five percent of county residents felt this to be true, given the current system of local government, have as did 76 percent of city residents. Empirical support led these leaders to call for major changes in govern- for this common assessment by citizens can be found mental arrangements. Regional issues, especially in studies of police service delivery in St. Louis in economic development, are important of course. 1972 and 1977. In both of these studies, residents of The proposed solutions, which would significantly al- small and medium sized communities in St. Louis ter an apparently effective system of local govern- County reported lower levels of criminal victimiza- ment service delivery, might enhance regional plan- tion, more rapid police response when called, higher ning and, perhaps, economic development (see the levels of assistance from local police, and more fa- discussion of this latter issue in Chapter Eight). vorable assessments of local police on a series of ad- Whether citizens are willing to engage in such a ditional indicators than did residents of comparable trade-off, assuming the trade-off to be real, remains areas in the city and unincorporated county (see to be tested at the polls. Perhaps a stronger possibil- Chapter Four). While police services are only one ity is that St. Louisans will find alternative ways of component of the service packages provided by local addressing areawide problems-ways that are more governments in St. Louis, they are a component that consistent with their prevailing system of metropoli- is highly valued by citizens. In the KTVI survey, 87 tan governance. percent of the respondents rated police and fire services in the county as better or much better than CURRENT REFORM ACTIVITY similar services in the city, a percentage equaled by In the fall of 1986, St. Louis County Executive public health services and exceeded only by educa- Gene McNary launched a new effort to reform local tional services. government in the county. His proposal, developed by the county planning department and quite similar which they can agree. Such a plan would then be to a plan proposed by Confluence St. Louis, sought presented to voters in the city and county, with con- to create new municipalities in the unincorporated current majorities in both required for adoption. portion of the county (thus incorporating the entire The required majority in the county would be a county), reduce the total number of municipalities countywide majority and, therefore, could allow to 21, and consolidate fire services in four fire dis- merger of smaller municipalities even in the face of tricts that together would cover the entire county.30 strong opposition from their citizens. Although prompted by the uncertainties created for In response to this initiative, the St. Louis county government by the prospect of repeated an- County Municipal League appointed a 19-member nexation and incorporation efforts that, if success- committee to develop alternative proposals. This ful, would erode the county's de facto municipal committee has developed a policy statement on gov- service area, the county government's proposal goes ernmental organization in the county that differs in much further. By creating municipalities of similar significant respects from the county and Confluence size throughout the county, local government could plans.31 The municipal league plan embodies some achieve greater uniformity in service delivery ar- differences in functional assignments as between rangements. Police services, for example, could be municipalities and county after full incorporation. It produced uniformly in-house by municipal police differs most significantly, however, in its stance to- departments, instead of relying on a mixture of in- ward existing municipalities and the mechanism by house and contract production as at present. Fire which their boundaries could be changed. The services could be produced uniformly through large league supports a phased annexation and incorpora- fire districts, rather than through the mixture of mu- tion of unincorporated portions of the county: nicipal and fire district production found at present. The St. Louis County Municipal League County government would be largely taken out of endorses the concept of phased universal the production of direct services, except for arterial incorporation of all of St. Louis County. A streets, but might increase its role in the production phased approach over the next several years will of auxiliary services. Whatever the eventual role of allow existing municipalities to develop plans for county government, it would be uniform throughout orderly and systematic expansion through an- the county. nexation. Groups of citizens may also wish to This proposal was a topic of interest to many of propose the establishment of newly incorporated the municipal officials in the county who were inter- cities with reasonable borders and sufficient size viewed for this study. Opinions among these officials to provide services. The League concludes that were mixed, with no clear pattern of support or op- universal incorporation does not require the dis- position. They did consistently indicate, however, mantling of existing municipal governments. that they believed their constituents generally would be opposed to a widespread merger of existing mu- Because the Municipal League finds that nicipalities. citizens of St. Louis County have a strong at- The vehicle chosen to advance this reform pro- tachment to the municipalities they have created posal, and to consider alternatives, was the metro- over the last century, we oppose a reorganiza- politan board of freeholders provided by the Mis- tion plan that would change the boundaries of souri Constitution as a mechanism for altering the an existing municipality without an affirmative arrangement between St. Louis City and County, vote of the residents or the governing body. separated in 1876 (see Chapter Three). One poten- where appropriate, of that municipality. tial difficulty with this institutional vehicle was the mandatory participation of St. Louis City in its for- The league has proposed, as an alternative, the mation. The city would be allowed to name as many establishment of a commission representative of members of the board as the county. Mayor Vincent presently incorporated and unincorporated parts of C. Schoemehl, Jr., however, announced his inten- the county to insure the orderly development of mu- tion to cooperate with the county in its reform ef- nicipal government in the unincorporated areas.32 fort. Late in 1987, pursuant to citizen petitions, the As the freeholders began consideration of alterna- board was selected and began operation. tive plans, county government substituted a 42-city The board of freeholders is composed of nine plan for its original 21-city proposal. The freeholders members appointed by the county executive, nine by also gave attention to issues of fiscal disparity among the mayor of St. Louis, and one by the governor. municipalities. One of the principal arguments made The freeholders have potentially a very broad man- in behalf of the county government's proposals was date under the constitution, enabling them to pro- the reduction of fiscal disparity (see the discussion in pose any plan for metropolitan reorganization on Chapter Eight). The freeholders were unable to reach a settlement of these issues by their original While recognizing that this requirement makes con- deadline in time for the May 1988 primary. solidation more difficult, "the commission believes The municipal league also opposed the use of that concurrent majority approval gives added assur- the board of freeholders mechanism on the grounds ance to the residents of each city, town, or township of the latter's constitutional mandate "to facilitate that they will not be unwillingly included in a con- some type of city-county consolidation." It saw the solidated government and provides a salutary politi- board as cal basis for launching a new municipality."38 It should be noted that the board of freeholders proc- an inappropriate vehicle for the resolution of ess would not even require approval from a majority service delivery problems in unincorporated St. of those voters within the boundaries of a proposed Louis County. We believe that the residents of new municipality. In the ACIR's words, this proce- St. Louis County will not support a reorganiza- dure may not provide a salutary political basis for tion plan developed by a Board of Freeholders the creation of a new political community. with a majority of members who reside outside The proposals for revamping county governance of St. Louis County. have also found life in a more traditional forum-the The contending approaches to reform provide a county delegation to the state legislature. A version contrast in attitudes with respect to the role of citi- of the St. Louis County Municipal League's proposal zen choice in local government, and the appropriate was introduced in the recent session of the legisla- communities of interest to make such choices. Un- ture, but failed to pass without county government der the county and Confluence plans, the appropri- support. The basic process of metropolitan govern- ate community of interest is a countywide constitu- ance in St. Louis County continues. ency.33 Viewing countywide issues as of overriding importance, advocates of these plans would be will- CONCLUSION ing to have municipalities dismantled in the face of St. Louis County has developed an elaborate strong opposition from their residents.34 The St. system of metropolitan organization. This organiza- Louis County Municipal League's plan embodies a tional system includes a basic governance structure different perspective, one based on incremental (embodied for the most part in state law), a diverse changes with the requirement of majority acceptance array of jurisdictions, and an overlay of multi- by those most directly affected. This perspective organizational arrangements associated with ongoing gives added attention to the smaller communities of processes of intergovernmental relations. The system interest represented by existing municipalities, by is a product of several decades of organizational de- groups of citizens in areas that might be annexed by velopment, a process that, given current problems existing municipalities, and by citizens who might and circumstances, is almost certainly not finished. wish to incorporate new communities in the pres- As a system of metropolitan governance, the or- ently unincorporated areas in the county. One can ganization of St. Louis County is based on arrange- anticipate that these competing views will receive a ments that allow citizens to define and organize their full airing in the debates that will undoubtedly ac- local communities much as they see fit. If one asks company the current reform efforts in St. Louis. the question, "Who governs?" in St. Louis County, The process for obtaining citizen consent associ- the best answer is, "Citizens govern." State legisla- ated with the St. Louis board of freeholders-con- tors-the local delegation-are merely the keepers of current majorities in city and county-is consistent the rules that provide a framework within which citi- with local self-determination when this process is ap- zens make the basic structural choices of local gov- plied to changes in city-county relations. As dis- ernment. Those who might conclude that no one is cussed above, the principle of self-determination un- in charge in this fragmented metropolis are looking derlies metropolitan organization in the St. Louis past the basic processes of governance. In St. Louis area. Applied to reorganization within the county, County, citizens are clearly in charge. however, the citizen consent procedures associated Citizen choice leads to diversity, not to the uni- with the board of freeholders are not consistent with formity desired by advocates of metropolitan re- self-determination. Half or more of the county's mu- form. The formal jurisdictions that comprise St. nicipalities could be eliminated, under this arrange- Louis County, however, provide simply a point of ment, without the consent of a majority of their citi- departure for patterns of organization that span ju- zens. The Advisory Commission on Intergovern- risdictional boundaries. This multiorganizational mental Relations (ACIR) has recommended that overlay, if it could be superimposed on a map of St. municipal consolidations be approved by citizens Louis County, would radically alter the popular im- voting in a referendum "by simple concurrent ma- age of fragmentation. To describe the organization jorities in the governmental jurisdictions involved. "35 of St. Louis County without reference to voluntary associations of municipalities, fire districts, and as a jurisdictional overlay in addition to city and school districts is to leave out half of the picture. To county governments. discuss service production and delivery without giv- Even city-county separation, however, has not ing attention to cooperative agreements and joint proved to be an insuperable barrier to cooperation. production arrangements among all types of local St. Louis City participates with county municipalities units is to tell half a story. Focusing exclusively on and school districts in a number of cooperative ven- formal jurisdictions leads to a conclusion that citi- tures. Two special districts have been organized that zens are hopelessly lost in a maze. Yet this conclu- embrace both city and county, and more are in the sion is made possible only by ignoring those func- works. A central problem with the city-county rela- tional dimensions of metropolitan organization that tionship is its bilateral nature. As more special dis- connect, mediate, and partially integrate separately tricts are created, however, the relationship may be- organized communities within a multijurisdictional gin to take on a multilateral character, and thus metropolis. perhaps facilitate further cooperation. Only limited evidence is found in St. Louis The immediate problems affecting St. Louis County of uncaptured economies of scale in service County are transitional. County government may no production. There is no compelling evidence that longer be able to maintain its role as a municipal the elimination of small municipalities or small service provider for 40 percent of the county's popu- school districts would significantly increase the effi- lation. The basic rules of municipal incorporation ciency of service production in those communities. and annexation, however, remain sound, although There is evidence, on the other hand, of inequi- some tinkering at the margins may be appropriate. A ties in service provision. A small number of small gradual period of incorporation and annexation municipalities and school districts in the county are would allow time for county government to adjust its disadvantaged, affecting disproportionately many of role. One possible adaptation is for county govern- the county's blacks and elderly citizens. Yet the ment to shift much more toward contract production magnitude of this problem is not overwhelming in its of municipal services. This step could increase the proportions, either with respect to municipal services level of competition among local service producers or elementary and secondary education. A con- and thus have a beneficial systemic effect through- cerned countywide constituency, organized through out the county. Pockets of unincorporated territory county government or through a special district or can be dealt with by contracting with adjacent mu- commission with authority to aid distressed commu- nicipalities. Any problems that remain after a period nities selectively, would clearly be able to overcome of incorporation and annexation can be addressed the more serious inequities that exist. Reorganization at that time. of municipal or school district boundaries is neither Local self-governance provides a solid founda- necessary nor sufficient to address these equity con- tion for metropolitan organization in St. Louis cerns. County. Extending the basic organizational dynam- These conclusions with respect to St. Louis ics of incorporation-annexation-cooperation to the County as a metropolitan county apply with some- presently unincorporated portion of the county will what less force to St. Louis City and County as a enable county government to become oriented more metropolitan area. Although a few special purpose toward its role as an overlying, general purpose juris- jurisdictions span city-county boundaries, the ab- diction and less toward its role as a de facto munici- sence of a general, overlying jurisdiction may be an pality in the unincorporated area. County govern- impediment to overall economic and political devel- ment clearly has important functions to perform-as opment. For fragmentation to work, it must be ac- the provider of an arterial street system, for exam- companied by a rich structure of overlapping juris- ple, and of important backup services to local po- dictions. This structure is much more elaborate lice. Perhaps most significantly, however, the county within St. Louis County than between St. Louis City government can play a much greater role in eco- and County. The separation of St. Louis City from nomic development, not only in the rapidly growing the county is clearly a greater barrier to economic unincorporated area, but especially in the county's and social progress in the larger metropolitan region small number of seriously distressed communities. than is jurisdictional fragmentation within the The entire county can benefit from such an effort. county. A range of options for reconstituting the re- The experience of the St. Louis area in metro- lation between city and county could be devised, in- politan organization has much to teach the rest of cluding not only reentry of the city into the county, metropolitan America. Jurisdictional fragmentation, but also the possibility of some broader representa- when combined with overlapping jurisdictions, can tional arrangement, such as a metropolitan council, provide an institutional framework for a dynamic of metropolitan organization that continually offers "See ACIR, Slate and Local Roles in the Federal System new opportunities for coordination and productivity (Washington, DC: ACIR, Report A-88, April 1982), pp. 446-448. The Commission has rescinded a 1981 recom- improvement. Jurisdictional fragmentation, there- mendation urging that small nonproducing local govern- fore, need not lead to functional fragmentation. Ef- ments be dissolved. See The Organization of Local Public fective metropolitan governance can emerge from Economies, p. 55. The Commission has taken the position that units of local government too small to produce services local self-determination and citizen choice, provided in-house may, nevertheless, usefully serve their citizens as that the basic rules of governance lie within a local service providers, while contracting for service production sphere of influence as well. Overall, such a process with private vendors or other government agencies. is necessarily complex, and difficult to capture in a "Anthony Downs, Urban Problems and Prospects, 2nd ed. (Chicago: Rand McNally College Publishing Company, snapshot case study. What is more, the process is 1976), Chapter 12. never finished. Only by continuing to study the func- '*See ACIR. The Organization of Local Public Economies. tional arrangements that emerge from different juris- "In a more consolidated system one would expect, all else dictional arrangements in various metropolitan ar- being equal, less exercise of public entrepreneurship. While eas, can we expect to improve the art of bureau chiefs in smaller jurisdictions may be in a position to be entrepreneurial, subordinate commanders in larger bu- metropolitan and local self-governance. reaucracies rarely are. The smaller number of elected public officials in more consolidated systems, too, reduces the potential for public entrepreneurship. "%chool districts are somewhat less dependent on self- l City of Town and Country v. St. Louis County, 657 determination, insofar as concurrent majorities are not nec- S.W.2d 598 (Mo. banc 1983). essarily required to merge school districts. Still, voter ap- In a council-manager form of government, the city manager proval districtwide is a requirement. How far this rule is the chief executive officer, responsible to and serving at departs from local self-determination in practice depends on the pleasure of the council, and the mayor is elected by the number of districts included in a proposed merger. Ex- council from its members, not separately elected by the peo- tensive historical research would be needed to learn whether ple. Some municipalities in St. Louis County have used the configuration of school districts in the county was af- home rule powers to create a position of city manager in ad- fected by this compromise of local self-determination. dition to a separately elected mayor. Still other municipali- leThe classic discussion of freerider and holdout problems is ties employ a city administrator, who reports to the mayor. found in Mancur Olson, The Logic of Collective Action The functional division of labor between mayor and man- (Cambridge: Harvard University Press, 1965). ager or administrator varies from one city to another. Thereader may be interested in comparing this discussion to Randy H. Hamilton refers to a related, but more general, the treatment of boundary review agencies in ACIR, The phenomenon in the extensive use of "citizen legislators" by Organization of Local Public Economies. American state and local governments. See his "Local Self- '*See John Mudd, "Beyond Community Control: A Neighbor- Government through Citizen Legislators: The Bedrock of hood Strategy for City Government," Publius 6 (Fall 1976) : Liberty" in Advisory Commission on Intergovernmental Re- 113-135. For further elaboration of this point, see Ronald lations (ACIR), Is Constitutional Reform Necessary to Re- J. Oakenon, "How Governance Structures Can Enhance invigorate Federalism? A Roundtable Discussion (Wash- Accountability and Responsiveness, " in James Perry (ed. ) , ington, DC: ACIR, Report M-154, November 1987), pp. Handbook of Public Administration, American Society of 17-25. Public Administration, forthcoming. See ACIR, The Organization of Local Public Economies '*See Robert Warren, Government in Metropolitan Regions: (Washington, DC: ACIR, Report A-109, December, A Reappraisal of Fractionated Political Organization 1987). (Davis: University of California, Institute of Governmental An additional advantage can be cited. Because many, in- Affairs, 1966). deed most elected officials of county municipalities have Theincorporation of the newly created City of Chesterfield other full-time jobs, most municipal meetings are scheduled in April 1988, was facilitated by the willingness of county to accommodate their work schedules, with evening and government to enter into a set of contracts for service pro- weekend meetings the predominant pattern. This pattern duction. including police services, in the new city's first works to the advantage of local citizens as well. Most of year of existence. them, like their officials, are employed full time. Not only a'For a discussion of the limits of interlocal competition, see are officials more accessible to citizens in the sense dis- ACIR, The Organization of Local Public Economies, pp. cussed in the previous section, but they tend to be more ac- 38-39. See also Ronald J. Oakerson, Roger B. Parks, and cessible in the conduct of their official duties. Henry Aaron Bell, "How Fragmentation Works-St. Louis a This distinction is the conceptual foundation for ACIR's re- Style," prepared for presentation to the Midwest Political port The Organization of Local Public Economies. Science Association annual meeting in Chicago, 1987. This count does not include Pacific, only a small portipn of =Dana L. Spitzer, Fostering Development in Metropolitan which lies in the county. St. Louis, a report submitted to John W. Hanley, Chair- man, The City-County Task Force of Civic Progress, De- One potential provision unit, the Village of Champ, has cember 1982. chosen not to make explicit provision for police services, re- =Report of the St. Louis County Annexation Study Commis- lying instead on the county police to enforce state law within sion, prepared by St. Louis County Department of Plan- its boundaries. ning, July 1985, p. 26. This count excludes specialized police forces such as that 24Too Many Governments? A Report on Governmental Struc- serving Lambert-St. Louis Airport, County Park Police, and ture in St. Louis City and County with Recommendations the Missouri Highway Patrol. for Change. Confluence St. Louis, February 18, 1987. 1°For further discussion, see ACIR, The Organization of Lo- =Market Opinion Research, Public Opinion on Government cal Public Economies. Reorganization Alternatives for the City of St. Louis and St. Louis County (Detroit: Market Opinion Research, April 32An additional group has recently entered this debate. The 1982). Local Government Service Association, comprised of local 28These generally high satisfaction ratings were also a feature elected officials from existing municipalities in the county, of earlier surveys in St. Louis City and County, for example especially those which would be likely to be eliminated if the that of The Metropolitan St. Louis Survey conducted in county or Confluence plans were adopted, is raising funds support of reorganization attempts in the late 1950s and from municipalities to combat reorganization proposals (St. early 1960s. See John C. Bollens, ed., Exploring the Met- Louis Post-Dispatch, November 3, 1987, p. 3A). ropolitan Community (Berkeley and Los Angeles: Univer- *Actually, the relevant constituencies are both citywide and sity of California Press, 1964). countywide because affirmative votes would be required in 27Spitzer, pp. 20-2 1. both jurisdictions to adopt a freeholder's plan. 2eKenneth F. Warren, "KTVI's Survey on Metropolitan Con- J4At least one advocate of the county plan, the county director solidation: A Summary Report," 1985. of planning, indicates that a countywide vote is essential to 2aThis poll, conducted in April 1987 by Attitude Research its success. He, like municipal officials, believes that re- Corporation, was reported in the St. Louis Post Dispatch, quiring concurrent majorities in existing municipalities prior November 27, 1987. to a change in their boundaries, would lead to voter rejection of many proposed mergers (personal communication, 30St. Louis County Department of Planning, A Comprehensive October 1987). Proposal for Local Government Reorganization in St. 23, Louis County, November 1987. "ACIR, State and Local Roles in the Federal System, p. 31St. Louis County Municipal League, Policy: Governmental 448. Organization, June 4, 1987. "Ibid., p. 449.

Recent Publications of the Advisory Commission on Intergovernmental Relations

Local Revenue Diversification: Local Income Taxes, SR-10,8/88,52 pp. Significant Features of Fiscal Federalism, 1988 Edition, Vol I, M-155,12/87, 128 pp. Vol 11, M-155 11, 9/88, 152 pp. Both Volumes Purchased Together Interjurisdictional Competition in the Federal System: A Roundtable Discussion, M-157, 8/88, 32 pp. State-Local Highway Consultation and Cooperation: The Perspective of State Legislators, SR-9, 5/88, 54 pp. The 1986 Federal Tax Reform Act: Its Effect on Both Federal and State Personal Income Tax Liabilities, SR-8, 1/88, 38 pp. Governments at Risk: Liability Insurance and Tort Reform, SR-7,12/87,42 pp. The Organization of Local Public Economies, A-109,12/87,64 pp. Measuring State Fiscal Capacity, 1987 Editon, M-156,12/87,152 pp. Is Constitutional Reform Necessary to Reinvigorate Federalism? A Roundtable Discussion, M-154,11/87,39 pp. Local Revenue Diversification: User Charges, SR-6,10/87, 64 pp. The Transformation in American Politics: Implications for Federalism, B-9R, 10187, 88 pp. Devolving Selected Federal-Aid Highway Programs and Revenue Bases: A Critical Appraisal, A-108,9/87, 56 pp. Estimates of Revenue Potential from State Taxation of Out-of-State Mail Order Sales, SR-5,9187, 10 pp. A Catalog of Federal Grant-in-Aid Programs to State and Local Governments: Grants Funded FY 1987, M-153,8/87,36 pp. Fiscal Discipline in the Federal System: National Reform and the Experience of the States, A-107,8/87,58 pp . Federalism and the Constitution: A Symposium on Garcia, M-152, 7/87, 88 pp. Local Perspectives on State-Local Highway Consultation and Cooperation, SR-4, 7/87, 48 pp. The reports of the Advisory Commission on Intergovernmental Relations are released in five series: the "A" series denotes reports containing Commission recommendations; the "M" series contains Commis- sion information reports ; the "S" series identifies reports based on public opinion surveys; the "B" series reports are abbreviated summaries of full reports; and the "SR series are staff information reports. Re- ports may be obtained from ACIR, 1111-20th Street, NW, Washington, DC 20575.