Policy Focus Report • Lincoln Institute of Land Policy

Evaluating Smart Growth State and Local Policy Outcomes

G r e g o r y K . I n g r a m a n d Y u - H u n g H o n g Evaluating Smart Growth: State and Local Policy Outcomes Gregory K. Ingram and Yu-Hung Hong

Contents

2 Executive Summary 4 Low-density Development and Smart Growth Policies 9 Growth Patterns and Trends 15 Natural Resources and Environmental Quality 19 Transportation 24 Affordable Housing 29 Fiscal Dimensions 33 Survey of Opinion Leaders 37 Conclusions and Recommendations 44 References

Policy Focus Report Series The policy focus report series is published by the Lincoln Institute of Land Policy to address timely public policy issues relating to land use, land markets, and property taxation. Each report is designed to bridge the gap between theory and practice by combining research findings, case studies, and contributions from scholars in a variety of academic disciplines, and from professional practitioners, local officials, and citizens in diverse communities.

About the Authors Gregory K. Ingram is president and CEO of the Lincoln Institute of Land Policy and cochair of the Department of International Studies. Contact: [email protected]

Yu-Hung Hong is a fellow at the Lincoln Institute of Land Policy and a visiting assistant professor at Massachusetts Institute of Technology. Contact: [email protected]

Copyright © 2009 by Lincoln Institute of Land Policy. All rights reserved.

113 Brattle Street Cambridge, MA 02138-3400, USA

Phone: 617-661-3016 x127 or 800-526-3873 Fax: 617-661-7235 or 800-526-3944 Email: [email protected] Web: www.lincolninst.edu

ISBN 978-1-55844-193-4 Policy Focus Report/Code PF020

ingram & hong ● Evaluating smart growth 1 ......

About this Report Contributors

The Lincoln Institute initiated a research project Policy Evaluation in late 2006 to evaluate the effectiveness of smart • Growth Patterns and Trends: Gerrit Knaap growth policies from 1990 to as far past 2000 as and Rebecca Lewis, University of Maryland data allowed. The analysis focused on four states • Natural Resources and Environmental Quality: with well-established statewide smart growth Terry Moore and Beth Goodman, ECONorth- programs (Florida, Maryland, New Jersey, and west, Oregon Oregon) and four states (Colorado, Indiana, Texas, • Transportation: Tim Chapin and Keith Ihlanfeldt, and Virginia) that offered a range of other land Florida State University management approaches. • Affordable Housing: Stuart Meck, Rutgers This report summarizes the findings and recom- Center for Government Services, and Timothy mendations of the complete evaluation, which MacKinnon, Monmouth University, New Jersey is published in the 2009 Lincoln Institute book, • Fiscal Dimensions: Robert W. Burchell and Smart Growth Policies: An Evaluation of Programs William R. Dolphin, Rutgers Center for Urban and Outcomes, edited by Gregory K. Ingram, Policy Research Armando Carbonell, Yu-Hung Hong, and Anthony Flint. This book is the source of the data and • Survey of Opinion Leaders and Regulatory statistics cited here unless otherwise noted. Analysis: Allan Wallis and Tom Clark, University of Colorado Denver The goal of the evaluation was to examine the effectiveness of various policies in achieving five State Case Studies commonly identified smart growth objectives: • Florida: Tim Chapin and Keith Ihlanfeldt, • promote compact development; Florida State University • protect natural resources and • Maryland: Gerrit Knaap and Rebecca Lewis, environmental quality; University of Maryland • provide and promote a variety of • New Jersey: Stuart Meck, Rutgers Center transportation options; for Government Services • supply affordable housing; and • Oregon: Terry Moore and Beth Goodman, • create net positive fiscal impacts. ECONorthwest Using 52 indicators based on U.S. Census Bureau • Colorado: Allan Wallis, University of Colorado data and other state and local datasets, several Denver research teams compared differences in perfor- • Indiana: Eric D. Kelly, Ball State University mance among the selected states and between the groups of smart growth and other states. • Texas: Robert G. Paterson, Rachael Rawlins, Another team surveyed opinion leaders on their Frederick Steiner, and Ming Zhang, University perceptions about the efficacy of smart growth of Texas at Austin programs, and other researchers prepared case • Virginia: Casey Dawkins, Virginia Tech studies on the political, environmental, and regu- latory conditions in the eight selected states.

ingram & hong ● Evaluating smart growth 1 ......

Executive Summary

Portland, Oregon his evaluation of the effectiveness • supply affordable housing; and of smart growth policies in the • create net positive fiscal impacts. United States focused on four states with well-established statewide smart No state did well on all performance mea- growthT programs (Florida, Maryland, New sures, although individual states succeeded Jersey, and Oregon) and four other states in one or more of their priority policy areas. (Colorado, Indiana, Texas, and Virginia) Maryland was successful in protecting that demonstrate a range of other land man- natural resources through its land preser- agement approaches (Ingram et al. 2009). vation programs and state funding for the The evaluation was objectives-based and purchase of farmland conservation ease- examined the extent to which five specific ments. New Jersey policies that responded smart growth objectives were achieved, to state supreme court decisions led to an based on measureable and comparable affordable housing approach that slowed performance indicators primarily during house price escalation and encouraged the decade from 1990 to 2000: rental and multifamily housing production. • promote compact development; Oregon’s commitment to establishing • protect natural resources and environ- urban growth boundaries was able to reduce mental quality; development on farmland in the Willamette • provide and promote a variety of Valley. The state also performed well in transportation options; reducing traffic congestion growth by

2 policy focus r e p o r t ● L i n c o l n I nstitute of Land Po l i c y ingram & hong ● Evaluating smart growth 3 ......

encouraging commuters to use transit and • Smart growth policies should make use by systematically planning for bicyclists and of economic incentives, such as pricing pedestrians. and tax policies, that have shown promise At the same time, some smart growth in other countries. states failed to achieve objectives in policy • Smart growth programs need to consider areas that were not given high priority the income distribution consequences during the study period, such as providing of their policies. affordable housing in Oregon and Maryland, and managing the spatial structure of urban Sustainability and Monitoring growth in Florida. of Programs The message is clear: achieving smart • Credible commitment from different growth is possible, but states must remain levels of government is crucial for the focused on their key policy goals. No single successful implementation of smart approach is right for all states. For example, growth programs. Colorado has no statewide smart growth • Improvements in measurement and program, but it outperformed some states collection of data, particularly related to with such policies by supporting local gov- environmental quality and public finance, ernment actions to pursue effective land use are needed to better monitor program planning within a regional context. The more performance. successful states use a variety of regulatory • More evidence is needed about the nature controls, market incentives, and institutional of interactions among smart growth poli- policies to achieve their objectives. cies—particularly those related to land use, transportation, and housing affordability. R ecomme n datio n s • Clearer definition of performance indi- Program Structure and Transparency cators and measurement of their attain- • The design of smart growth programs ment would facilitate the evaluation of and supporting regulations and incentives smart growth programs and contribute to should be guided by a vision of sustain- their technical and political sustainability. able and desirable development outcomes. • Any top-down or bottom-up smart Although this evaluation of smart growth growth policies must be coordinated at programs concentrates primarily on state- the regional level to be able to achieve wide performance during the 1990s, the their desired objectives. findings and recommendations will be use- • Policy makers must articulate the means ful for formulating growth management of achieving smart growth objectives and policies in today’s context of high energy specify implementation mechanisms, costs, historic housing market volatility, and rather than just declare objectives. increasing pressures to reduce greenhouse gas emissions. Many smart growth objectives Functional Linkages for Policy Design are precisely the outcomes posited to address • The design of growth management policies these current challenges facing state and should take account of interactions among local policy makers. policies and coordination across relevant agencies.

2 policy focus r e p o r t ● L i n c o l n I nstitute of Land Po l i c y ingram & hong ● Evaluating smart growth 3 ......

C hapter 1 Low-Density Development and Smart Growth Policies

Baltimore, Maryland

ew public policy issues are as con- planning for compact urban growth and tested as urban sprawl. Across the transit-oriented development. These growth United States, people are debating management approaches have attracted the issue of low-density develop- much public attention and research, but they mentF at the urban fringe as state and local have received little systematic evaluation. governments try to reconcile growing Several states have applied smart growth demands for new housing and commercial policies for decades, and others are just development with needs to protect open beginning to use them to address emerging space. The debate over sprawl often pits issues in the twenty-first century. the public good against private self-interest. While most people agree that protecting A n H istorical V iew of natural resources and open space is impor- G rowth Patter n s tant, many also value their property rights Low-density development at the urban and resist policies that may reduce the fringe has been prevalent in the United value of their land holdings. States since the 1940s. The average amount In response to this challenge, several U.S. of developed land per capita increased from jurisdictions have implemented smart growth 0.32 acres in 1982 to 0.38 acres in 2002. principles since the early 1970s through More important, the amount of newly devel-

4 policy focus r e p o r t ● L i n c o l n I nstitute of Land Po l i c y ingram & hong ● Evaluating smart growth 5 ......

oped land per added resident over this period illustrates that the average density in the averaged about 0.6 acres—nearly twice the Northeast was the highest among the four level of average land consumption. Growth regions in 1982, and more than twice that in population will increase overall land in the Midwest and South. Between 1982 consumption, and income growth can explain and 2003, the incremental density was lower much of the growth in area per person. than the average density in all regions, indi- Figure 1 shows that between 1982 and cating that all development was oriented 2002 the number of acres of developed toward lower densities. The West experi- land increased by 46 percent, while the U.S. enced rapid population growth, but man- population and personal income grew by 24 aged to keep its incremental density higher and 77 percent, respectively. This is consis- than the other regions (see Fulton et al. tent with the finding that land consumption 2001 for similar findings). will increase by about 30 percent if personal income doubles. If the land area per person T he E vol u tio n of S mart remained constant, population growth G rowth P olicies would be responsible for growth in land area In the face of this decreasing density and of 24 percent, leaving 22 percent of the 46 the spread of development at the fringe of percent total area due to other factors. If many urban areas, some states and localities all of this additional 22 percent growth was began to put policies in place to shape settle- caused by the 77 percent increase in income, ment patterns. By the early 1990s, these each 10 percent increase in income would efforts came to be knows as “smart growth” have caused a 3 percent increase in land area programs. What is now termed smart growth per person, a relationship similar to that has evolved from a continuous process of found in estimates of the demand for lot state land use policy development that has size (Glaeser, Kahn, and Rappaport 2008). coalesced around a set of objectives: pro- However, regional trends in population mote compact development; protect natural density are certainly not uniform. Figure 2 resources and environmental quality; create

figure 1 Percent Growth in Personal Income Influences Growth in Developed Land and Population, 1982–2002

200

180 Personal Income

160 Developed Land Population 140 Notes: Personal income is in

rcen t Pe 2005 dollars. Population and developed land estimates do 120 not include Alaska. Sources: U.S. Census Bureau 100 (1990c; 2000c); U.S. Census Bureau (1990d; 2000d); U.S. Census Bureau (2007); and 80 U.S. Department of Agriculture 1982 1987 1992 1997 2002 (1982; 1987a; 1992; 1997; 2003).

4 policy focus r e p o r t ● L i n c o l n I nstitute of Land Po l i c y ingram & hong ● Evaluating smart growth 5 ......

figure 2 The second wave, from the 1980s into the Average and Incremental Densities Vary Across U.S. Regions, early 1990s, marked a shift from regulating 1982–2003 and controlling growth to planning that was aimed at promoting economic growth and 6.16 protecting natural resources. The deployment of infrastructure also became more impor- tant as a land use planning tool. The second- 3.52 3.46 2.67 wave smart growth states included Florida,

1.10 Georgia, Maine, New Jersey, Rhode Island, 1.06 Midwest Northeast Vermont, and Washington (DeGrove 1992). West The third wave, beginning in the mid- 1990s, was distinguished by the addition of positive incentives to influence growth and 2.72 by more growth-accommodating policies. 1.39 Some states shifted from land use regulation, South urban growth boundaries, and requirements Average Density 1982 for local comprehensive plans to urban re- (Persons per acre) vitalization, zoning reform, and better coor- Incremental Density 1982–2003 (Added population per acre of dination of state agencies and their growth newly developed land) policies. More emphasis was also placed at

Sources: U.S. Census Bureau (2007); U.S. Department of Agriculture (2003). the local, metropolitan, and regional levels, and less on the hegemony of statewide transportation options and walkable neigh- programs. The third wave brought several borhoods; supply affordable housing; additional states into the movement, includ- generate net positive fiscal impacts; encour- ing Maryland, Minnesota, Pennsylvania, age community collaboration and facilitate Tennessee, and Utah (DeGrove 2005). transparent and effective development A fourth wave of smart growth policies is decision making processes (Smart Growth still emerging. The need to respond to climate Network 2009). change, environmental challenges, and soar- The smart growth movement can be ing energy costs, along with a new emphasis divided into four waves. The first three waves on investments in public infrastructure, have were defined by John DeGrove (1984; 1992; bolstered the demands being placed on smart 2005). We propose a fourth wave to charac- growth initiatives. Because automotive travel terize the emerging features of the smart produces a large share of greenhouse gas growth movement in the twenty-first century emisions, support is increasing for land use (Ingram at el. 2009). The first wave com- policies that foster more compact develop- menced in the 1970s with the enactment of ment patterns, transit use, and walking. growth management programs to advance New regional approaches are likely to environmental protection in California, encourage development proposals that ad- Colorado, Florida, Hawaii, North Carolina, here to a smart growth framework. Market Oregon, and Vermont. These programs forces are also encouraging more compact, were based on the regulation and control mixed-use development as households of land development either throughout the attempt to limit their travel costs and state or within specially designated zones achieve other energy savings. (DeGrove 1984).

6 policy focus r e p o r t ● L i n c o l n I nstitute of Land Po l i c y ingram & hong ● Evaluating smart growth 7 ......

Florida A n O pport u n e T ime for while others did little or nothing. These E val u atio n eight states constitute a purposive and not Smart growth programs in some states are a random sample as part of the research now in their fourth decade, but new environ- methodology (box 1). mental objectives have raised the stakes on The analysis revealed that the treatment their success. With many years of experience varied greatly across the four smart growth behind us, and the likelihood that reliance states, producing a range of outcomes that on growth management policies will grow in overlap with some of those in the other the future, this is an opportune moment to selected states. Outcomes and policies were evaluate how effective smart growth policies thus found to be more continuous across and programs have been at achieving their the eight states rather than dichotomous goals and objectives. between the two groups of states. This evaluation is based on a comparison This evaluation addresses two key ques- between four states that had statewide smart tions. First, does the presence of state-level growth policies in place by 2000 (Florida, smart growth programs result in objectively Maryland, New Jersey, and Oregon) and measurable improvement in performance? four other states that did not (Colorado, Second, to the extent that smart growth Indiana, Texas, and Virginia). Some of these programs are successful, what underlies this latter states did facilitate local and regional success? Conversely, if they fail, what are smart growth initiatives by enabling local the causes of their shortcomings? governments to promulgate local options,

6 policy focus r e p o r t ● L i n c o l n I nstitute of Land Po l i c y ingram & hong ● Evaluating smart growth 7 ......

Box 1 Research Methodology

he evaluation measured the achievement of five smart growth objectives in each of the Teight states. The analytic methods used varied according to the data available, ranging from descriptive statistics to fixed-effect regression models (Ingram et al. 2009, 10–20). The focus was on changes in performance indicators over time, given that current levels of many measures (e.g., population density) reflect the cumulative effects of past policies, technolo- gies, and relative prices. The effects of recent policies are likely to be observed only in cur- rent changes in performance indicators. It is also likely that some smart growth objectives reinforce each other, and others are antagonistic.

To make comparisons across states over multiyear periods, the evaluation developed a set of performance indicators that were defined consistently over time and available for all states. These indicators relied heavily on data from the U.S. Census Bureau and other nationally collected datasets uniformly available at the state and county levels. The analysis generally starts in 1990 and continues as far past 2000 as data allow, but focuses primarily on the decade of the 1990s. In addition to the performance indicators, the evaluation also analyzed how opinion leaders perceive the effectiveness of smart growth programs, and how imple- mentation has changed over time.

8 policy focus r e p o r t ● L i n c o l n I nstitute of Land Po l i c y ingram & hong ● Evaluating smart growth 9 ......

C hapter 2 Growth Patterns and Trends

Fort Worth, Texas major objective of smart growth per square mile in New Jersey to 35 in policies is to alter the spatial dis- Oregon). Their population and employment tribution of population and employ- growth rates over time vary much less, how- ment, principally by increasing the ever. As figure 3 shows, on average, popula- Adensity and intensity of development, pro- tion increased more slowly in the smart moting compactness, and slowing the sprawl- growth states (15.9 percent) than in the other ing of development to rural and undeveloped selected states (19.4 percent) from 1990 to areas. Four measures of growth patterns are 2000, as did employment (22.5 percent used to assess relative changes in spatial struc- versus 24.8 percent) from 1994 to 2004. ture in the eight states: land use patterns, spatial concentration, urbanization, and L a n d u S E Patter n s centralization (Ingram et al. 2009, 22–45). Land uses vary considerably across the A baseline review of the size and growth states. For example, about half of state area of the eight case study states in 2000 shows is rangeland in Texas, cropland in Indiana, that they range widely in population (from forestland in Virginia, and federal land in more than 20 million in Texas to 3.4 million Oregon. The area of developed land, which in Oregon); size (from 265,000 square miles is most relevant to smart growth policy, in Texas to 7,500 square miles in New Jersey); increased in each of the eight states in every and population density (from 1,115 persons five-year interval from 1982 through 1997.

8 policy focus r e p o r t ● L i n c o l n I nstitute of Land Po l i c y ingram & hong ● Evaluating smart growth 9 ......

figure 3 Employment Growth Generally Exceeded Population Growth The largest proportional increase in devel- oped land occurred in Florida, followed by 40 Virginia; the smallest increase was in Indi-

Florida ana. The average proportional increase was 35 26 percent for the four smart growth states and 21 percent for the four other states. Virginia Because the states differ so widely in land 30 use and growth, these proportional increases Other Selected State Average Texas Colorado can be made comparable by relating the 1994–2004 25 increase in each state’s developed land area

yment, Maryland Smart Growth State Average to the increase in population over similar 20 periods. Figure 4 shows the ratio (marginal

New Jersey land consumption) as the increase in square Oregon 15 miles of developed land per 1,000 new resi- Indiana dents. While the average for smart growth states is lower than for the other selected 10 states, the best performers—Oregon and rcent Change in Emplo Pe Colorado—are from the two different 5 groups (box 2).

0 S patial C o n ce n tratio n 0 5 10 15 20 25 30 35 40 The distribution of population over space Percent Change in Population, 1990–2000 can be measured by the Gini coefficient,

Note: Employment counts are derived from the U.S. Census Bureau’s Zip Code Business Patterns which is an index of inequality based on the in 1994 and 2004, with 1994 the earliest year for which data were available. Counts do not in- Lorenz curve measuring how evenly a vari- clude government workers, farm workers, or part-time or self-employed persons. Sources: U.S. Census Bureau (1990b; 1996; 2000b; 2006). able is spread. When activities are uniformly

figure 4 Developed Land Generally Increased Less in Smart Growth States than in Other Selected States

1.00

0.90 Smart Growth States 0.80 Other Selected States 0.70 0.60 0.50 0.40 0.30 0.20 Marginal Land Consumption 0.10 0.00 Florida Maryland New Jersey Oregon Colorado Indiana Texas Virginia

Notes: Averages are 0.61 for the smart growth states and 0.71 for the other selected states. Growth in developed land is measured from 1987 to 1997 in square miles. Population growth is measured from 1990 to 2000. Marginal land consumption is square miles per 1000 additional residents. Source: U.S. Department of Agriculture (2000).

10 policy focus r e p o r t ● L i n c o l n I nstitute of Land Po l i c y ingram & hong ● Evaluating smart growth 11 ......

figure 5 Population and Employment in Most States Became Less Concentrated

1.0

0.9

0.8

0.7 Gini Coef cient

0.6

0.5 Florida Maryland New Jersey Oregon Colorado Indiana Texas Virginia Smart Growth States Other Selected States

1990 Population 2000 Population 1994 Employment 2004 Employment

Sources: U.S. Census Bureau (1990b; 1996; 2000b; 2006); GeoLytics (2002).

distributed, the Gini coefficient is zero; when its concentration declined more than that activity is concentrated in one place, it is one. of population over the decade. The average Decreases in the spatial Gini coefficient over reduction in Gini coefficients for the smart time therefore indicate that the development growth states was greater than for the other pattern is becoming more dispersed. Spatial states, for both population (-.007 versus Gini coefficients vary from 0.25 to 0.90 across -.002) and employment (-.021 versus -.011). all U.S. states, and are usually high in states This outcome is generally counter to smart with only one large city and low in states growth objectives. While these differences with no large cities and dispersed populations. have similar patterns for most states, they Here the concern is more about the change are not statistically significant across states. in concentration over time than its level, Gini coefficients for population and which reflects the legacy of the past more employment also were calculated for each than the effects of current policies. Increased large metropolitan area (with population concentration (higher Gini coefficients) would over one million) in the eight case study generally be consistent with smart growth states. The results are similar to the state- policies. wide outcomes, but with larger Gini co- Figure 5 shows statewide Gini coefficients efficients. Again, the average reduction in calculated from census tract data for popula- Gini coefficients in the smart growth states tion and from zip code data for employment was greater than that in the other selected over ten-year intervals. Oregon is the only states for both population (-.019 versus state where population concentration in- -.017) and employment (-.044 versus -.028). creased and employment concentration did Echoing the state results, Portland was the not decrease. While employment was typi- only metropolitan area where population cally more concentrated than population, concentration did not decrease.

10 policy focus r e p o r t ● L i n c o l n I nstitute of Land Po l i c y ingram & hong ● Evaluating smart growth 11 ......

Box 2 Urban Growth Management in Oregon and Colorado

rom 1982 to 1997, Oregon and Colorado distin- growth over 20 years. Some communities also identify urban F guished themselves by having smaller increases in reserved areas intended to accommodate growth over a longer marginal land consumption than any of the other case time horizon. study states. Oregon experienced a decline in the amount The state also provides incentives in the form of grants and of developed land per capita over the 1990s, and in Colo- technical assistance to jurisdictions undertaking planning func- rado 88 percent of the population lived in only 5 percent tions. In the 1980s, the state and federal governments provid- of the census tracts during the same period. ed $24 million in planning grants to local governments, or Oregon has one of the first and best-known statewide nearly 63 percent of the planning budget during that period smart growth programs in the country (Ingram et al. (Rohse 1987). Between 1997 and 2007, these planning 2009, 188–198). The 1960s development boom in the grants declined to $12 million. Oregon also defers taxes on state consumed productive farmland and raised fears farmland and forestland. A recent study estimates the total about pollution, deteriorating quality of life, and loss of amount of taxes deferred between 1974 and 2004 at more the state’s economic base. In response, Oregon passed than $4.8 billion (Richmond 2007). the 1973 Senate Bills 100 and 101 that emphasized Colorado’s approach to managing urban growth in the early the need to protect its agricultural and forestry lands 1970s was to enable local governments to engage in planning by establishing an urban growth boundary (UGB). and to implement land use controls, but not to have a mandat- The UGB identifies and separates land that can be urban- ed statewide growth control program (Ingram et al. 2009, 200– ized from land that must remain rural, and encourages 208). The Colorado Constitution and Local Government Land compact development. Oregon’s urbanization goal re- Use Control and Enabling Act granted counties and municipali- quires all cities to define, adopt, and plan development ties the authority to plan and regulate land use. In principle, within growth boundaries that provide enough land to the Land Use Commission and the Department of Local Affairs accommodate projected residential and employment could override local planning permission for land development

Portland, Oregon

12 policy focus r e p o r t ● L i n c o l n I nstitute of Land Po l i c y ingram & hong ● Evaluating smart growth 13 ......

u R B A n izatio n Smart growth programs seek to encourage infill development in urbanized areas and reduce the spread of development to adjoin- ing rural areas. To assess performance on these objectives, population growth was classified by three locations: areas denoted as urban in 1990; those newly urban between 1990 and 2000; and those rural in both 1990 and 2000. The smart growth states had a larger share of new residents settle in urban and newly urbanized areas, and a smaller share in rural areas (figure 6). Oregon had the highest share of popula- tion growth in already urbanized areas at 49 Denver, Colorado percent; New Jersey was second at 45 percent; in preserved areas. In practice, however, the state and Colorado was third at 38 percent. Indi- role has been to assist, not direct, local govern- ana had the lowest share at 6 percent. Using ments to identify and adopt guidelines for matters this same classification for large metropolitan of state interest. areas, the results were similar to those at the During the 1990s, the Great Outdoors Colorado state level. Portland had the highest share Trust Fund was established to receive half of the of additional residents settle in already proceeds of the Colorado Lottery. A portion of the urbanized areas in the 1990s (59 percent), trust’s funds can be awarded for open space plan- while Miami–Ft. Lauderdale had the ning grants. Consistent with general citizen sup- second highest (54 percent). port for preserving open space, Colorado ranked first in the nation between 1999 and 2004 in the C e n tralizatio n dollar value of voter-approved open space bonds To measure changes in the centralization and tax measures (Trust for Public Land 2005). of population and employment over time, each large metropolitan area in the eight Colorado also provides several examples of volun- tary regional collaborations designed to manage growth. The establishment of the Denver Region figure 6 Smart Growth States Kept More Population Council of Governments Metro Vision 2020 is Growth in Urbanized Areas most noteworthy. It includes the Mile High Com- pact and has created a voluntary growth boundary 50 Smart Growth States covering the six-county area along the Front Range 40 Other Selected States of the Rocky Mountains where more than 60 30 percent of the state’s population resides. Other elements of Vision 2020 call for reinforcing spines 20 opulation Growth

of development along transit corridors. Passage P 10

of the $4.7 million FasTracks bond initiative in rcent of 1990–2000 Pe 0 2004 also provided significant support for the Urban in 1990 Urbanized Rural in 1990 and 2000 implementation of Vision 2020. after 1990 Sources: U.S. Census Bureau (1990b; 2000b); GeoLytics (2002).

12 policy focus r e p o r t ● L i n c o l n I nstitute of Land Po l i c y ingram & hong ● Evaluating smart growth 13 ......

figure 7 Metropolitan Area Population and Employment Decentralized Less in Smart Growth States

9

8 Smart Growth States 7 Other Selected States 6 5 4 oint Decrease in Share 3 2 1

rcentage P Pe 0 <5 miles <10 miles <20 miles <5 miles <10 miles <20 miles Population 1990–2000 Employment 1994–2004

Sources: U.S. Census Bureau (1996; 2006).

case study states was divided into concentric four states, marginal land consumption per rings with radii of 5, 10, 20, and 30 miles. new resident was lower, the share of new The shares of population and employment population locating in urban areas was located within each ring were calculated at higher, and population and employment the beginning and end of a ten-year period. decentralization was lower than in the four As might be expected, employment was more other selected states. Smart growth states centralized than population. The shares also added a smaller share of new popula- of population and employment within the tion in rural areas. At the same time, 5-, 10-, and 20-mile rings decreased or re- however, the concentration of population mained the same in all metropolitan areas. and employment declined more in the smart Employment decentralized more than popu- growth states than in the other states. lation. Figure 7 shows that there was less When ranked in terms of overall perfor- decentralization in the metropolitan areas mance, the top three states were Oregon, of the smart growth states, a result consis- Colorado, and New Jersey, with Florida tent with the analysis of urbanization above. ranked eighth. Oregon performed well across most measures including land use, S u mmar y urbanization, and concentration. Colorado’s Overall, the changes in growth patterns in strong showing indicates that smart growth the smart growth states show some consis- outcomes can be attained without a man- tency with smart growth objectives. In these datory statewide smart growth policy.

14 policy focus r e p o r t ● L i n c o l n I nstitute of Land Po l i c y ingram & hong ● Evaluating smart growth 15 ......

chapter 3 Natural Resources and Environmental Quality

Virginia hile smart growth policies are L a n d C o n servatio n intended to improve the envi- While all of the case study states support ronment and protect natural land trusts and related conservation ease- resources, the strength of the ments, state policies differ in many details. linkagesW between specific programs and objec- Maryland and Oregon have programs to tives varies. Smart growth policies often protect open space and environmentally relate directly to land use, including land sensitive land and to preserve agricultural conservation, but only indirectly to air and land. Maryland provides state funding to water quality through impacts on transpor- purchase conservation easements on farm- tation and development patterns. land. New Jersey preserves agricultural lands Data are reasonably accessible for land by purchasing development rights, and pro- use and land conservation measures that tects environmentally sensitive lands through are consistent over time and across states, regional planning. Florida has purchased but it was impossible to obtain comparable over 2.5 million acres of environmentally data for air and water quality. As a result, sensitive land, but has yet to fund its pro- all natural resource and environmental gram for protecting agricultural lands. performance measures in this evaluation Among the other selected states, Colo- pertain to the use and conservation of rado offers state tax credits for private con- land (Ingram et al. 2009, 46–57). servation easements, purchases conservation

14 policy focus r e p o r t ● L i n c o l n I nstitute of Land Po l i c y ingram & hong ● Evaluating smart growth 15 ......

land, and uses lottery proceeds to fund park As figure 8 indicates, the smart growth and conservation programs. Indiana has a states experienced smaller losses per new modest land trust program funded from sales resident in both land categories. Maryland of affinity license plates. Texas has three lost the least amount of resource land per programs aimed at the preservation of forest- new resident (0.38 acres), while Indiana land, one of which involves state purchase (0.90 acres) and Oregon (0.88 acres) lost the of conservation easements. Virginia has most. Virginia lost the least amount of farm- no statewide program, but facilitates local land per new resident (0.04 acres), while efforts to adopt conservation policies. Colorado (2.36 acres), Indiana (1.63 acres), and Oregon (0.89 acres) lost the most. R eso u rce L a n d Oregon’s loss is surprising given the state’s a n d Farmla n d goal of protecting farmland. Further anal- The analysis relied on two comprehensive ysis revealed, however, that the loss was datasets available at five-year intervals. The primarily in the sparsely populated eastern first, the National Resource Inventory, pro- part of the state. The densely settled Willa- vides information on several land use cate- mette Valley region around Portland actu- gories, five of which were aggregated into a ally increased the amount of farmland single measure termed resource land: crop- per added resident by 0.05 acres. land, pastureland, rangeland, forestland, and conservation reserve program land. The L a n d T r u sts a n d second dataset, available from the National C o n servatio n P rograms Agricultural Statistics Service, collects infor- All eight case study states support the place- mation on farm acreage through a census ment of private land in land trusts and of of farms. Changes in the amounts of resource farmland in conservation programs. During land and farmland were related to the the two overlapping 15-year periods shown change in population in each of the eight in figure 9, smart growth states performed state cases. less well than the other selected states, but the within-group performance varied widely. figure 8 In 2005, New Jersey (3.6 percent) and Mary- Smart Growth States Lost Less Land per New Resident land (2.5 percent) had the highest percent- than Other Selected States ages of their areas in conservation easements 0.80 held by land trusts, while Oregon (0.1 per-

0.70 cent) and Florida (0.2 percent) had among the lowest. 0.60 In terms of farmland in conservation 0.50 programs, Colorado (5.6 percent) performed 0.40 best in 2002, followed by Maryland and 0.30 Oregon (both at 2.8 percent). The share of 0.20 area in land trusts is thus a poor predictor 0.10 of the share of farmland in conservation pro- Acres Lost per New Resident 0.00 grams, except in Maryland where both are Resource Land 1982-–1997 Farmland 1987–2002 reasonably high.

Smart Growth States Other Selected States

Sources: U.S. Department of Agriculture (1987b; 2000; 2002).

16 policy focus r e p o r t ● L i n c o l n I nstitute of Land Po l i c y ingram & hong ● Evaluating smart growth 17 ......

figure 9 Smart Growth States Protected Smaller Shares of Their Area in Land Trusts and Farmland Conservation Programs

3.00

2.50 Smart Growth States

2.00 Other Selected States

1.50

rcent Area Pe 1.00

0.50

0.00 1990 2005 1987 2002 State Area in Land Trusts Farmland Area in Conservation

Notes: Land trust data excludes land owned by The Nature Conservancy. Farmland data includes land enrolled in Conservation Reserve and Wetlands Reserve programs. Sources: Land Trust Alliance (n.d.; 2005); U.S. Department of Agriculture (1987b; 2002).

S tate Parkla n d figure 10 A final indicator related to natural resources Smart Growth States Lost Their Lead in State Parkland is the amount of land devoted to state parks. After the 1990s Acres of parkland per 1,000 persons, the 25.00 level of service measure used by the National Smart Growth States Recreation and Park Association, has changed 20.00 Other Selected States

over time in the two groups of states. As sons figure 10 shows, the smart growth states were 15.00 slightly ahead of the other selected states on this indicator in 1990, but the difference 10.00 between the two groups was negligible by 5.00 2006. In that year, Colorado had the most r Acres per 1000 Pe parkland per 1,000 persons (42.5 acres), 0.00 followed by New Jersey (38.5 acres). Virginia 1990 2006 had the lowest service level (8.1 acres) and Indiana the next lowest (10.3 acres). Source: National Association of State Park Directors (n.d.).

S u mmary level, Maryland had the highest average The evidence on natural resource and ranking across all measures (box 3), with environmental quality measures is mixed, New Jersey and Colorado tied for second. with neither group of states clearly outper- Indiana had the lowest average ranking. forming the other in terms of protecting Colorado again performed well despite its undeveloped areas. At the individual state lack of a statewide smart growth program.

16 policy focus r e p o r t ● L i n c o l n I nstitute of Land Po l i c y ingram & hong ● Evaluating smart growth 17 ......

Box 3 Land Conservation Programs in Maryland

aryland is the gatekeeper to the largest and The emphasis on land use grew in the 1990s, beginning M most productive estuary in the United States, the with the Economic Growth, Resource Protection, and Plan- Chesapeake Bay. Because most of Maryland lies within ning Act of 1992 (the Growth Act) and the Smart Growth the watershed, the health of the bay has been a major Areas Act of 1997, which took an inside/outside approach driver of the state’s land use and environmental policies in an effort to direct growth to Priority Funding Areas for many years. (PFAs) while preserving undeveloped areas through the Rural Legacy Program. This evaluation revealed that Maryland had the largest percentage increase in acres of farmland enrolled in land The centerpiece of the Smart Growth Areas Act attempted preservation programs among the eight selected states to influence development decisions by restricting “growth- (Ingram et al. 2009, 166–175).­­­ About 2.5 percent of related” state spending to specific areas. County govern- Maryland’s land was privately conserved in 2005, the ments were required to designate certain areas as PFAs, second highest level among the selected states. In addi- which included all incorporated municipalities; heavily tion, Maryland preserved 343,000 acres—about 5 per- developed areas inside the circumferential highways cent of its total land area—through state and county around Baltimore and the Maryland suburbs of Washing- transfer of development rights and purchase of devel- ton, DC; and other areas meeting specific state criteria. opment rights programs (Lynch et al. 2007). The counties were required to map their PFAs and submit Beginning in the 1960s, the Maryland General Assembly their plans to the Maryland Department of Planning (MDP) and various governors proposed and enacted a series of for review and comment. The goal was to use the power land use laws designed to protect the environment. These of the state budget as an incentive for smarter growth. laws were intended to help the state acquire parkland, State programs were geared either to support develop- protect forests and wetlands, reduce soil erosion, preserve ment within the PFAs or to protect undeveloped land farmland, and regulate storm water runoff. Much of the outside them. focus turned to the Chesapeake Bay following passage of the Chesapeake Bay Agreement in 1983.

Chesapeake Bay, Maryland

18 policy focus r e p o r t ● L i n c o l n I nstitute of Land Po l i c y ingram & hong ● Evaluating smart growth 19 ......

chapter 4 Transportation

Dallas, Texas

mart growth proponents view trans- sistently across states are reasonably available, portation as a major determinant including census data that record commute of land use patterns and an impor- mode and travel time. Data from the Texas tant manifestation of the success Transport Institute, which estimates annual ofS smart growth policies. They argue that delay per peak-period traveler and a peak- expanding transport options, altering trans- period travel time index, have been used to port pricing, and fostering pedestrian-friendly examine levels and changes in congestion. settings yield less single-occupant car travel, Census data are available every 10 years less congestion growth, and more trips by for all states and municipalities, and the transit, biking, and walking. These patterns Texas Transport Institute data are available are associated with more compact, mixed- annually for 85 U.S. metropolitan areas. Con- use, and dense urban forms. This evaluation gestion data were used for cities with popula- therefore looked at performance indicators tions of one to three million, and seven of related to mode choice and traffic conges- the eight case study states (all but New Jersey) tion to assess how they are associated with had at least one such metropolitan area with smart growth programs, which reflect different congestion data. Vehicle miles traveled policy approaches (Ingram et al. 2009, 58–75). (VMT) was one explanatory variable used Transportation data that are defined con- to analyze the change in congestion.

18 policy focus r e p o r t ● L i n c o l n I nstitute of Land Po l i c y ingram & hong ● Evaluating smart growth 19 ......

C omm u te M odes increase the share of bike/walk commutes Data on mode choice for work trips were or at least retard its decline. But as figure tabulated for cities and counties with an 12 indicates, while the bike/walk share was average density of at least 50 persons per generally higher in the smart growth states, square mile. Counties were sorted into three its share declined over time and was essen- density categories—very high (more than tially unrelated to population density. The 500 persons per square mile), high (101 to exception to this pattern is Oregon, where 500), and medium (50 to 100)—to control the bike/walk share increased from 1990 for the level of urbanization. to 2000 by more than 10 percent—likely Figure 11 shows that the transit share of reflecting the state requirement that local work trips varied greatly with population governments produce bike and pedestrian density in both the smart growth and other plans as part of their transportation plans selected states. The transit share of work (box 4). trips across the country during the 1990s was 6 percent in very high density counties, C o n gestio n 1 percent in high density counties, and 0.5 The relationship between the form of percent in medium density counties. development and traffic congestion is much Work trip transit shares in all four smart debated. Some analysts believe that dense, growth states exceeded U.S. averages, as compact development promotes transit use did those in Colorado. In addition, average and shorter automobile trips, while others work trip transit shares increased from 1990 contend that decentralization reduces the to 2000 in the smart growth states in all distances from home to work and spreads density categories, but generally declined in car travel more widely over existing trans- the other selected states except Colorado. port capacity. Since smart growth programs Given that smart growth programs typically seek to reduce congestion, assessing typically provide bike lanes, bike racks, the change in travel delays over time is sidewalks, and priced parking, they should essential.

figure 11 Work Trip Transit Shares Started Higher and Rose in Smart Growth States

9

8 Smart Growth States 7 Other Selected States 6 5

ercen t 4 P 3 2 1 0 1990 2000 1990 2000 1990 2000 Very High Density Counties High Density Counties Medium Density Counties

Sources: U.S. Census Bureau (1990e; 2000e).

20 policy focus r e p o r t ● L i n c o l n I nstitute of Land Po l i c y ingram & hong ● Evaluating smart growth 21 ......

figure 12 The Bike/Walk Share Generally Started Higher in Smart Growth States, but Declined During the 1990s

6 Smart Growth States 5 Other Selected States

4

3 ercent P 2

1

0 1990 2000 1990 2000 1990 2000 Very High Density Counties High Density Counties Medium Density Counties

Sources: U.S. Census Bureau (1990e; 2000e)

figure 13 Annual Increases in Traffic Delays in Smart Growth States Declined After Smart Growth Programs Were Introduced

2.50 y,

2.00 s of Dela

1.50

1.00 1982–2003

0.50

Annual Increase in Hour 0.00 Florida Florida Maryland Maryland Oregon Colorado Indiana Texas Virginia Before After Before After Smart Growth States Other Selected States

Source: Ingram et al. (2009, 65).

Figure 13 shows average annual increases growth states (1.73 hours) exceeds that in in peak-period hours of delay, including the other selected states (1.31 hours). Yet data for Florida and Maryland before and the initiation of smart growth programs in after their respective programs were initi- Florida and Maryland reduced the annual ated. Oregon’s start date preceded the avail- increase in traffic delays, thus providing able data, and no data were available for New some evidence of program success. Jersey. The results are mixed. The average The results from a sample of six smart annual increase in congestion in the smart growth states indicate that smart growth

20 policy focus r e p o r t ● L i n c o l n I nstitute of Land Po l i c y ingram & hong ● Evaluating smart growth 21 ......

Box 4 Transportation Planning in Oregon

regon’s integration of transportation planning with O land use management is among the most advanced in the country (Ingram et al. 2009, 188–198).­­ The state’s transportation plans must be based on an inventory of local needs and consider all modes (cars, freight, bicycle, pedestrian, public transit). In 1993 the state created the Oregon Transportation and Growth Management Program (TGM), a partnership between the Oregon Department of Land Conservation and Development and the Oregon Department of Transportation. One of the primary goals of TGM is to make walking, biking, and using transit safe and convenient. The program also aims to enhance the state’s road system as a whole and to improve the ability of commercial enterprises to move goods and services along the highways (Oregon Transportation and Growth Management Program 2009).

TGM’s funding from the Federal Highway Administration enables the state to leverage its own investment with fed- eral dollars. TGM recognizes that scattered development without good connections between local destinations in- creases the need for driving and that well-planned devel- opment with good street and walkway connections im- proves transportation options (such as walking and transit-oriented development; and bicycle and pedestrian biking) and can reduce automobile usage. networks. From 2007 to 2009, TGM granted $3.8 million in financial and technical assistance to 60 local transpor- Thus TGM promotes planning concepts including mixed- tation projects throughout Oregon. use, compact development; good connectivity between local destinations; revitalizing downtown and main streets TGM also sponsors community workshops, lecture series, where good transportation options are already available; and other events to improve public understanding of land use and transportation planning concepts. Through its Quick Response and Transportation System Plan Assessments pro- grams, TGM offers direct design assistance and assessments of transportation needs to communities. These functions aid local governments in their search for grants to carry out necessary projects. Its Code Assistance program helps local governments revise zoning and development codes.

22 policy focus r e p o r t ● L i n c o l n I nstitute of Land Po l i c y ingram & hong ● Evaluating smart growth 23 ......

programs have a statistically significant and S u mmar y behaviorally meaningful effect on conges- Analysis of the transportation indicators, tion. The travel delay regression indicates especially work-trip transit ridership and that smart growth programs reduced the changes in congestion, provides reasonably annual increase in delay by 2.2 hours for strong evidence that smart growth programs every year that the program was in place. are associated with desirable outcomes. For example, if the travel delay had been While the evidence on bike/walk commutes increasing by 4.0 hours per year, smart was less compelling than that on transit, growth programs would reduce that number smart growth states had somewhat higher to 1.8 hours per year. shares of work trips by these modes. An attempt was also made to relate the When performance across the three congestion reduction from smart growth major indicators was aggregated for each programs to three underlying causal factors state, Oregon ranked at the top. That state —increased population density, increased did very well in transit and bike/walk com- transit ridership, and changes in VMT. Taken mutes, and was the top smart growth state together, these variables were found to explain in terms of congestion. Indiana and Texas no more than 20 percent of the effect of were at the bottom of the overall rankings. state smart growth programs on congestion. It is noteworthy that the four states that performed best in the rankings on trans- portation (Oregon, Virginia, Colorado, and New Jersey) also performed best on growth patterns and trends.

22 policy focus r e p o r t ● L i n c o l n I nstitute of Land Po l i c y ingram & hong ● Evaluating smart growth 23 ......

chapter 5 Affordable Housing

Boulder, Colorado

hile improving the affordabil- added such a requirement in 2003 (Ingram ity of housing is a common et al. 2009, 76­­­–87). goal of smart growth programs, ­­ the emphasis placed on this H o u si n g Val u es objectiveW varies across the case study states. The first measure used to assess housing New Jersey has the strongest state-level affordability was the change in median program, which stems from the Mt. Laurel I housing values from 1990 to 2000. Figure and II state supreme court decisions requiring 14 shows that median values rose in all eight municipalities to provide realistic opportuni- states, with the largest percentage increase ties for low- and moderate-income housing (118 percent) in Oregon. Although New Jer- on a regional fair-share basis. sey posted the smallest percentage increase, Florida requires that local plans include a it had the highest median housing value of housing element. Oregon has a requirement the eight case study states in both 1990 and for provision of “needed housing,” while 2000. Colorado had the second highest Maryland has no specific state-level housing median value in 2000, followed by Oregon mandate. None of the other selected states and Maryland. Housing in the smart growth had a state-level affordable housing require- states was clearly more expensive than in ment during the 1990s, although Virginia the other selected states.

24 policy focus r e p o r t ● L i n c o l n I nstitute of Land Po l i c y ingram & hong ● Evaluating smart growth 25 ......

Even so, the average increase in median M u ltifamily a n d prices was lower in the smart growth states R e n tal H o u si n g (31 percent) than in the other selected states States that successfully promote affordable (58 percent). As a result, the difference in housing are likely to produce more multi- median house values between the two family and rental units. As figure 15 shows, groups shrank from $34,000 in 1990 to all eight case study states added varied $25,000 in 2000. The source of the price shares of rental units during the 1990s. increases in the smart growth states could Multifamily units made up a larger average have been on the demand side (greater amen- share of new housing in the smart growth ities), the supply side (higher regulatory states (21 percent) than in the other selected costs), or both. states (13 percent). Smart growth states were

figure 14 Percentage Increases in Median House Value Varied Widely Among the Case Study States

$180,000 4% 94% 160,000 1990 2000 24% 118% 140,000 42%

alue 31% 120,000 22% 73% 100,000 32% 80,000 60,000 Median House V 40,000 20,000 0 Florida Maryland Jersey Oregon Colorado Indiana Texas Virginia U.S.

Smart growth state average increase = 31 percent. Other selected state average increase = 58 percent. Average

Note: Includes all owner-occupied units. Sources: U.S. Census Bureau (1990a, table H061A; 2000a, table H85).

figure 15 Smart Growth States Added a Larger Share of Multifamily Units During the 1990s

35 30 Rental Multifamily 25 20 15 10

rcent of Units Adde d Pe 5 0 Florida Maryland Jersey Oregon Colorado Indiana Texas Virginia Smart Growth States Other Selected States

Sources: U.S. Census Bureau (1990a; 2000a).

24 policy focus r e p o r t ● L i n c o l n I nstitute of Land Po l i c y ingram & hong ● Evaluating smart growth 25 ......

Box 5 The Council on Affordable Housing in New Jersey

he Council on Affordable Housing (COAH) is a key had to provide realistic opportunities to meet their fair share T planning agency that administers the New Jersey Fair of low- and moderate-income housing in their regions. Housing Act (Ingram et al. 2009, 177–187).­ The enact- Under the Fair Housing Act, COAH’s responsibilities in- ment of this legislation was a response to a series of clude defining housing regions; estimating moderate and state supreme court decisions, known as the Mt. Laurel low-income housing needs; setting criteria and guidelines cases, which dealt with affordable housing and exclusion- for municipalities to determine and address their fair ary zoning. In the first case, Southern Burlington County share numbers; and reviewing and approving hous- NAACP v. Township of Mount Laurel, 67 N.J. 151, the court ing elements/fair-share plans and regional contribution ruled in 1975 that developing municipalities have a con- agreements for municipalities. Once its housing element stitutional obligation to provide a realistic opportunity for and fair share plans are approved, the municipality has the construction of low- and moderate-income housing. a degree of protection from Mt. Laurel–type lawsuits.

In the second case, Mt. Laurel II (92 N.J. 158) in 1983, During the 1990s, the council used an allocation formula the court held that all municipalities should share the to establish goals for all municipalities in the state. As obligation to provide the opportunity for the development of 2008, COAH had begun round three and moved to a of affordable housing, and provided specific judicial guide- “growth share” formula that bases the affordable housing lines for municipalities to follow, so as to fulfill their con- determination for a municipality on the actual growth of stitutional obligation. Municipalities that enacted zoning market-rate units and nonresidential development.

Cranbury, New Jersey

26 policy focus r e p o r t ● L i n c o l n I nstitute of Land Po l i c y ingram & hong ● Evaluating smart growth 27 ......

also less likely to add housing in rural figure 16 areas. New Jersey had the highest shares Housing Cost Burdens Were Generally Higher in the Smart Growth of both rental and multifamily units of States, Especially for Renters the eight case study states (box 5), while Maryland had the lowest rental share 30 and Colorado had the lowest multi- 25

family share. 20

15 H o u si n g C ost B u rde n Affordability is determined by both 10 Housing Cost as rcent of Income Pe housing prices and household incomes. 5 The housing cost burden is the percent 0 of income spent on housing, commonly 1989 1999 1989 1999 measured by the ratio of median house Owners Renters prices and rents to median household Smart Growth States Other Selected States income. Figure 16 indicates that renter and owner housing cost burdens were Note: Cost-burdened owners and renters are defined as those paying 30 percent or more slightly higher in the smart growth states of income for housing. Source: U.S. Census Bureau (1990a; 2000a). than in the other selected states. The share of income spent on housing changed little from 1989 to 1999, except that figure 17 The Share of Cost-burdened Owners Rose in the Smart Growth owners in smart growth states saw their States in the 1990s cost burdens edge up from 22 percent to 23 percent. 40 In both groups of states, the housing 35 Smart Growth States cost burden for renters was consistently 30 Other Selected States 25 higher than for owners. The renter bur- 20 den fell slightly in all states except Oregon 15 (up 5.5 percent) and Colorado (up 1.1 Households 10 percent). The owner cost burden rose 5 rcent of Cost-burdened Pe 0 the most in Indiana (15.6 percent) and 1989 1999 1989 1999 Oregon (13.7 percent), and decreased Owners Renters Note: Cost-burdened households spend 30 percent or more of income for housing. the most in Texas (-3.8 percent) and Note: Cost-burdened owners and renters are defined as those paying 30 percent or more Virginia (-2.3 percent). of income for housing. A generally accepted standard of Source: U.S. Census Bureau (1990a; 2000a). affordability is that housing costs should be less than 30 percent of household the share of cost-burdened renters fell or was income. Accordingly, a specific indicator unchanged. of affordability (or its lack) is the share The shares of both cost-burdened owners of households whose housing cost bur- and renters were higher in the smart growth den exceeds 30 percent of income. As states than in the other selected states. Oregon shown in figure 17, the share of cost- posted the largest increase in the share of burdened owners rose between 1989 cost-burdened owners (5.8 percent), followed and 1999 in both groups of states, while by Maryland (3.8 percent). Texas was the only

26 policy focus r e p o r t ● L i n c o l n I nstitute of Land Po l i c y ingram & hong ● Evaluating smart growth 27 ......

state to show a decline. The share of cost- S u mmar y burdened renters increased only in Oregon These results indicate that smart growth (2.5 percent) and Maryland (0.1 percent), programs that lack an affordable housing and fell the most in Texas (-3.9 percent). element have been associated with increases Note that these changes in housing cost in housing cost burdens, especially for owners. burdens do not account for any offsetting While smart growth states experienced a cost reductions, such as for transportation, smaller increase in median housing prices that may be associated with smart growth and added a greater share of multifamily programs. units in the 1990s, they also had higher shares Statistical regressions were used to anal- of cost-burdened owners and renters than yze the determinants of the change in the the other selected states. shares of cost-burdened owners and renters Regression results indicate that the across the case study states. Regressions that smart growth states had greater increases in hypothesized a uniform effect from smart the share of cost-burdened owners than of growth programs found a statistically signi- renters. New Jersey, with its court-mandated ficant relationship. Smart growth programs affordable housing requirement, was first were associated with increased shares of overall in a composite ranking across all cost-burdened households. housing affordability indicators. Oregon Additional regressions that allowed each ranked last, having experienced the largest state to have an independent effect found increases in housing values, housing cost that the shares of cost-burdened renters and burdens, and shares of cost-burdened owners increased the most in Oregon and households. the least in Texas. But New Jersey and Florida—smart growth states that require affordable housing elements in local plans— performed better than Oregon and Mary- land for owners, and better than Oregon, Maryland, Virginia, and Colorado for renters.

28 policy focus r e p o r t ● L i n c o l n I nstitute of Land Po l i c y ingram & hong ● Evaluating smart growth 29 ......

C hapter 6 Fiscal Dimensions

Indianapolis, Indiana

mart growth programs seek to population density of each state. The anal- concentrate economic activity in ysis looked at ten variables related to econo- areas that are already developed mic development for which comparable data and to control growth in undevel- were available for the decades from 1980 Soped and rural areas. Fiscal revenues and to 1990 and from 1990 to 2000: population expenditures have a role to play in these growth and density, households, employment, efforts. Some central questions are the extent personal income, retail sales, tax base, hous- to which revenues from developed areas are ing values, multifamily units, and journey- sufficient to pay for expenditures, and how to-work travel times. the balance of revenues and expenditures The shares of incremental growth in compares between smart growth and each variable that occurred in each state’s other states. urban/suburban and rural/undeveloped To examine these issues, counties in the counties were calculated for both decades. eight case study states were separated into The proportion in the 1980s was then sub- two groups according to their population tracted from that in the 1990s to provide densities—rural/undeveloped, and urban/ a simple summary statistic measuring the suburban or otherwise developed (Ingram et change in the distribution of economic acti- al. 2009, 88–115). The densities that defined vity. For example, if rural counties received the two categories varied with the average 8 percent of a state’s population growth in

28 policy focus r e p o r t ● L i n c o l n I nstitute of Land Po l i c y ingram & hong ● Evaluating smart growth 29 ......

figure 18 Smart Growth States Had Less Growth in Rural/Undeveloped Areas in the 1990s than in the 1980s

Multifamily Units House Value

Tax Base

Retail Sales

Income Smart Growth States Employment Other Selected States Households

Population

Average

-4 -2 0 2 4 6 8 10 12 14 16 Percentage Point Change in the Rural Share of Growth in the 1990s Relative to the 1980s.

Source: U.S. Census Bureau (1980; 1990c; 2000c); Woods and Poole Economics, Inc. (2005).

the 1980s and 13 percent in the 1990s, the made it necessary to impute values for many net increment for rural counties would be missing variables before expenditures and 5 percent. revenues could be analyzed. First, a signifi- Figure 18 shows how much the shares of cant amount of information for New Jersey statewide growth in rural/undeveloped coun- and Texas was missing, because local juris- ties changed from the 1980s to the 1990s on dictions did not report it. Second, some eight of the ten variables. The only activity expenditures and revenues are not counted, where that share decreased was multifamily especially if they are nonrecurring or in the housing in the smart growth states. In all form of intrajurisdictional transfers (table 1). other cases, the rural/undeveloped share of Figure 19 shows the ratios of the respec- growth was equal to or larger in the 1990s tive growth of expenditures and revenues than in the 1980s. It is striking that the rural (in constant dollars) in the 1990s and in the growth rate for all activities in the smart 1980s. A ratio over 1.0 indicates that growth growth states was less than or equal to that accelerated; that is, its absolute magnitude in the other selected states. This indicates in the 1990s exceeded that in the 1980s. This that smart growth states were more success- figure reveals three trends: (1) both expendi- ful in fostering density in urban/suburban tures and revenues grew faster in rural areas areas and in moderating the growth of devel- than in urban/suburban areas; (2) revenues opment in rural/undeveloped areas. grew faster than expenditures in all areas; The analysis of fiscal impacts was based and (3) smart growth states had lower growth on public finance data drawn from the U.S. in both expenditures and revenues than the Census of Governments. However, the 2002 other selected states. The faster growth in census had two serious data problems that rural areas reflects the increase in develop-

30 policy focus r e p o r t ● L i n c o l n I nstitute of Land Po l i c y ingram & hong ● Evaluating smart growth 31 ......

Table 1 Expenditures and Revenues in the Evaluation

Expenditures Revenues Included Excluded Included Excluded Local public infrastructure Transfers between the same Property, sales, income franchise, Impact fees for capital and services investments or different levels of government lodging, fuel, and other taxes improvements

Salaries/wages and other Public services provided Fees collected from inspections, expenditures by community associations building permits, and ordinance filings

Debt service to support Educational expenditures in Traffic and parking fines large-scale capital project noneducational budgets of selected municipalities and counties

Purchases of computers, Municipal subsidies to charter schools Service charges for contracted solid office furniture, and regular waste removal, animal control, and vehicles special assessments for improvements New components of expenditures Intergovernmental transfers such as start-up computer costs

New spending due to the addition of new government divisions or annexation

Source: Ingram et al. (2009, 104–105).

figure 19 Aggregate Expenditures and Revenues Increased Less in Smart Growth States than in Other Selected States

1.6 Smart Growth States 1.4 Other Selected States 1.2

1.0

0.8

0.6

0.4

0.2

Ratio of 1990s Growth to 1980s 0 Urban/Suburban Rural/Undeveloped Urban/Suburban Rural/Undeveloped Expenditures Revenues

Source: U.S. Census Bureau (1982; 1992; 2002).

ment as documented in figure 18, while the change in revenues to the change in the other patterns may reflect prudent fiscal expenditures. When this ratio is greater than management in the smart growth states. 1.0, revenues are growing faster than expen- Fiscal impact, a common metric used to ditures. The analysis indicates that the smart evaluate fiscal performance, is the ratio of growth states had a more favorable fiscal

30 policy focus r e p o r t ● L i n c o l n I nstitute of Land Po l i c y ingram & hong ● Evaluating smart growth 31 ......

balance (1.2) in the 1990s than the other mental expenditures. It is noteworthy that selected states (1.0) in urban/suburban areas. the more favorable fiscal balances in the What explains this outcome? Analysis of smart growth states result from larger tax the change in tax bases and tax rates reveals increases. This suggests that these states are that the smart growth states increased taxes generally more supportive of the public somewhat more than the other selected sector than the other selected states, in terms states to strengthen their fiscal positions. of both the regulatory structure underlying smart growth programs and the provision of S U MM ar y financial resources to the public sector. New Overall, smart growth states did better than Jersey performed best overall on measures the other selected states in controlling the of the distribution of growth and fiscal growth of economic development activities impact, followed by Florida and Maryland. in rural areas, and in achieving a favorable The lowest ranking states were Indiana balance of incremental revenues over incre- and Virginia.

32 policy focus r e p o r t ● L i n c o l n I nstitute of Land Po l i c y ingram & hong ● Evaluating smart growth 33 ......

C hapter 7 Survey of Opinion Leaders

Baltimore, Maryland

his evaluation also included a sur- that the costs of smart growth policies and vey of 117 state and local opinion the time required to complete the review leaders—about 15 in each of the process had “become a lot higher” than eight case study states. The ques- respondents in the other selected states. Ttionnaire covered the period from 2000 to Opinion leaders generally had similar views 2007 and addressed five major topics related about the role of government in smart growth to smart growth programs: effectiveness in policies, except that those from the other achieving goals; effectiveness of sanctions selected states were more likely to believe and incentives; public participation; costs of that state governments should defer to regulatory compliance; and the government’s local governments on such issues. role in guiding land use decisions. The survey Responses on the effectiveness of achiev- was careful to differentiate between state- and ing goals, of public participation, and of local-level efforts and activities because sev- sanctions are summarized in figure 20. Opin- eral of the other selected states enable local ion leaders felt that smart growth states had governments to apply smart growth policies been more effective than the other selected (Ingram et al. 2009, 116–133). states at the state level, but that the other Respondents in the smart growth states selected states had been nearly as effective were two to three times more likely to believe at the local level. These views reaffirm

32 policy focus r e p o r t ● L i n c o l n I nstitute of Land Po l i c y ingram & hong ● Evaluating smart growth 33 ......

figure 20 Survey Respondents Viewed State-level Programs in the Other Selected States as Least Effective

4.0

3.5

3.0

2.5

2.0

1.5

1.0 erage Score (5=highest ) Av 0.5

0.0 Achieving Goals Public Participation Effective Sanctions

Smart Growth States, State Level Other States, State Level Smart Growth States, Local Level Other States, Local Level

Source: Ingram et al. (2009, 118–125).

figure 21 Perceptions of State-level Effectiveness in Achieving Smart Growth Goals Vary More than Those of Local-level Effectiveness Across All States

4.0

State Level 3.5 Local Level

3.0

2.5

2.0

1.5

1.0 erage Score (5=Most Effective) Av 0.5

0.0 Florida Maryland New Jersey Oregon Colorado Indiana Texas Virginia Smart Growth States Other Selected States

Source: Ingram et al. (2009, 119).

34 policy focus r e p o r t ● L i n c o l n I nstitute of Land Po l i c y ingram & hong ● Evaluating smart growth 35 ......

earlier findings in this report that states with- figure 22 out statewide programs, such as Colorado, Smart Growth States Have Stronger State and Local have succeeded in achieving smart growth Residential Regulations objectives through locally implemented policies. State Smart Growth Requirements This point is reinforced in figure 21, which shows by state the opinion leaders’ views of Strong 20 NJ the effectiveness of state and local govern- OR ments in achieving smart growth objectives.

The lack of a statewide smart growth prog- FL MD ram apparently contributes to increased Average Score=12

local activism in the other selected states that may exceed the perceived effectiveness 10 CO IN TX of programs in smart growth states. SCORE VA The strength of state regulatory regimes was assessed as a composite of five attributes: Wharton state requirements for local planning; state Residential Land specification of the size of communities that Use Regulatory Index must plan; and state requirements for inter- Weak 0 nal consistency, vertical consistency, and hori- MEAN -1.0 SD +1.0 SD zontal consistency. The Wharton Residential Low High Land Use Regulatory Index was used to measure the strength of local housing devel- Smart Growth States Other Selected States

opment regulation (Gyourko, Saiz, and Note: SD=Standard deviation. Note: SD=Standard deviation. Summers 2006). Source: Gyourko, Saiz, and Summers (2006). Figure 22 shows that both state and local Source: Ingram et al. (2009, 146). regulations are strong in the four smart growth states. Colorado, which does well on many smart growth performance indicators, also has relatively strong local regulations, sug- gesting that if they are reasonably consistent within a state they can produce smart growth outcomes similar to those in states with strong state-level regulation (box 6). In addition, whether a state is subject to Dillon’s Rule (a legal doctrine holding that localities have only those powers specifically delegated by state law) has little relation to the presence of strong state or local regula- tions. For example, Maryland is a Dillon’s Rule state and Oregon is not, yet both have strong state and local regulations (Richard- son and Gough 2003).

34 policy focus r e p o r t ● L i n c o l n I nstitute of Land Po l i c y ingram & hong ● Evaluating smart growth 35 ......

Box 6 Local Government Planning Roles in Colorado

olorado currently has no statewide mandated growth with local comprehensive plans. In effect, such plans C management program. Instead, the state’s approach are advisory and presumed to be modified when zoning is has been largely to create a toolbox of planning powers amended. Similarly, the state does not require that local that local governments can adopt. The few mandatory capital improvements—including those by school dis- requirements often reflect federal requirements devolved tricts, utilities, or other units of local government—be by the state to local or regional jurisdictions. State agen- consistent with local plans. Local governments primarily cies offer fairly modest technical support for planning use their own powers to protest a land use permitted (Ingram et al. 2009, 200–208). by an adjacent jurisdiction that would have a negative impact on their own jurisdictions. Although there is no mandatory state planning require- ment, there are many locally initiated regulations. Under To encourage regional collaboration and planning, the Colorado Revised Statutes, counties and municipalities state has created several programs designed to work as with a certain population level or growth rate must pre- voluntary partnerships and/or incentive programs, such pare and adopt a master plan. The Department of Local as the Colorado Wetlands Partnership, Colorado Voluntary Affairs annually calculates which local governments meet Clean-up and Redevelopment Act, and Greater Outdoors certain growth thresholds, and receives plans for advice Colorado. In addition, the Colorado Division of Housing and comment. Nevertheless, the state has no authority and the Colorado Housing Finance Authority make re- to approve plans or to enforce recommendations. sources available to local governments that want to pro- vide low-income housing. Tax incentives were established In addition, since Colorado has no state or mandatory in 2000 to encourage developers to build low-income regional plans, there is no consistency requirement. In rental housing. fact, the state does not mandate internal consistency

Rocky Mountains, Colorado

36 policy focus r e p o r t ● L i n c o l n I nstitute of Land Po l i c y ingram & hong ● Evaluating smart growth 37 ......

C hapter 8 Conclusions & Recommendations

Richmond, Virginia

he evidence presented here does resources and environmental protection. not sustain the premise that state- In contrast, Oregon and Maryland ranked wide programs are either necessary low on affordable housing, an objective that or sufficient to attain all smart growth their programs did not emphasize, and Tobjectives, although most statewide programs Florida ranked low on spatial structure. clearly make progress on one or more of them. The programs adopted by both the smart While the sample smart growth states as a growth and other selected states differ greatly group only marginally outperformed the in their details, even beyond their emphasis other selected states overall, one of the smart on specific objectives, but some common growth states performed best on each objec- patterns and linkages exist among some tive. At the same time, however, another smart objectives. For example, the four states with growth state often performed well below the highest rankings on spatial structure also average on each objective. rank highest on transportation, supporting There is a marked tendency for smart the idea that transportation and develop- growth states to perform well in an area that ment patterns are closely related. is a high priority for that state (table 2). Thus, The correlations in table 3 also support Oregon ranked highest on spatial structure that conclusion. The second strongest cor- and transportation measures, New Jersey on relation is between environmental protection affordable housing, and Maryland on natural and fiscal dimensions, which suggests that

36 policy focus r e p o r t ● L i n c o l n I nstitute of Land Po l i c y ingram & hong ● Evaluating smart growth 37 ......

Table 2 State Performance Rankings Vary Among Smart Growth Objectives Spatial Environmental Housing Fiscal Structure Protection Transportation Affordability Dimensions Smart Growth States Florida 8 6 5 5 3 Maryland 6 1 6 7 4 New Jersey 3 2 4 1 1 Oregon 1 6 1 8 8 Other Selected States Colorado 2 2 3 6 2 Indiana 7 8 7 4 6 Texas 5 4 7 2 5 Virginia 4 4 2 3 5

Source: Ingram et al. (2009, 146)

Table 3 Correlations Across Smart Growth Objectives Show Mixed Effects Spatial Environmental Housing Fiscal Structure Protection Transportation Affordability Dimensions Spatial Structure 1.00 Environmental Protection 0.33 1.00 Transportation 0.76 0.09 1.00 Housing Affordability -0.17 0.06 -0.32 1.00 Fiscal Dimensions -0.08 0.62 -0.13 0.41 1.00

Source: Ingram et al. (2009, 147)

land preservation and conservation occur growth objectives provided supportive and more often in states with strong fiscal bal- enabling conditions for local governments ances and modest development in rural to pursue those objectives. This is the case in areas. Colorado, where several metropolitan areas On the other hand, the performance on the Front Range have implemented on fiscal dimensions is independent of the similar smart growth programs—essentially performance on transportation and spatial simulating a statewide program. Texas and structure. The correlation between housing Virginia also facilitate or do little to constrain affordability and transportation is somewhat local smart growth initiatives. Indiana does negative, indicating that the factors associated very little to support smart growth policies at with high transit use may also raise housing either the state or local levels and performs prices. These correlations suggest potential poorly across all five objectives. synergies and antagonisms among various Because smart growth programs involve smart growth objectives, and highlight the strengthened regulatory controls on develop- need for careful program design. ment, higher housing prices are often seen As noted earlier, states without statewide as an inevitable consequence. This evalua- programs that did well in achieving smart tion, however, does not support the view

38 policy focus r e p o r t ● L i n c o l n I nstitute of Land Po l i c y ingram & hong ● Evaluating smart growth 39 ......

that smart growth is always associated with These findings support several recom- increases in housing prices or reductions in mendations that can be grouped under three housing affordability. The evidence for New headings: program structure and transpar- Jersey indicates that housing affordability ency; functional linkages and program design; can be achieved in a state that performs well and program sustainability and monitoring. on the main smart growth objectives when it As state and local governments spend ever mandates the promotion of affordable hous- greater amounts of time and money in ing. Smart growth states with weak or no pursuit of smart growth objectives to meet affordable housing mandates at the state new challenges such as climate change, it is level perform poorly in this area. essential to improve the efficacy of policies, While no state performs well across all the transparency of objectives and their objectives, the rankings do support the con- performance indicators, and the evidence clusion that smart growth programs can and that objectives are being met. do achieve smart growth goals. Thus, this evaluation counters what might be termed R ecomme n datio n s o n “the smart growth impossibility theorem,” P rogram S tr u ct u re a n d that is, the conjecture that historic growth T ra n spare n c y patterns and behaviors are so entrenched A vision of sustainable and desirable that they cannot be changed. The mixed development outcomes needs to in- results and imperfect performance of the spire and motivate smart growth smart growth states seem to reflect the pri- programs and inform the package of orities of specific programs and how the policies, rules, incentives, and regu- states focused their efforts. lations that support such programs.

Boulder, Colorado

38 policy focus r e p o r t ● L i n c o l n I nstitute of Land Po l i c y ingram & hong ● Evaluating smart growth 39 ......

Florida and Oregon best embody the Rocky Mountains have taken a similar ap- evidence for this recommendation. Florida proach to urban development, with Denver has a long history of promoting and requir- making an aggressive effort to slow sprawl. ing planning at all governmental levels. But This consistent policy approach at the local when its outcomes were evaluated in terms level essentially mimics an effective state- of smart growth objectives, the state per- wide program and provides coherence at formed poorly compared with other states the regional level. in this study. Florida has not articulated a Smart growth programs implemented by coherent strategy for managing growth or a local governments with no regional coordi- vision of what constitutes desirable develop- nation are unlikely to yield good outcomes ment outcomes. Instead, it has followed a because of negative spillover effects from process-oriented approach that—at times communities pursuing their parochial in- because of unfortunate policy interactions terests. In addition, state-level programs that —has essentially encouraged low-density are poorly coordinated or do not take account development on the fringe of metropolitan of policy interactions across agencies also areas. In contrast, Oregon’s smart growth will perform poorly. program, whose mission is clearly described in its initiating legislation, has focused strongly Smart growth policies should artic- on preventing development on agricultural ulate the means of achieving objec- lands adjacent to urban areas and has had tives and specify implementation measurable success in achieving that mechanisms, rather than just objective in the Willamette Valley. declare objectives. It is easy for policy makers to endorse legis- Smart growth policies can be im- lation that extols the virtues of smart growth, plemented on either a top-down or such as improved environmental quality, a bottom-up basis, but an approach more affordable housing, and reduced traffic that coordinates policies at the regional congestion. But without confronting the chal- level is a minimal requirement to lenges of how to achieve those objectives, achieve smart growth objectives. such legislation is little more than statements This evaluation found that states with state- about preferred states of the world. wide smart growth programs tended to do Real progress on the smart growth somewhat better across a range of perfor- agenda has come in states that have grap- mance indicators than most states that had pled with implementation and come up no such programs. Oregon and New Jersey with new means of achieving smart growth have succeeded in more than one aspect objectives. For example, Maryland’s initia- of smart growth. tives have gained notice because of their Exceptions to this pattern were often avowed use of incentives to promote smart found in Colorado and sometimes in Texas growth, which are seen as an advance over and Virginia, which do not have statewide prior regulatory or command-and-control policies but do allow local governments to approaches. pursue extensive growth management prog- The results of this evaluation suggest rams. Colorado actually performed better that achieving smart growth objectives is than several states with statewide smart extremely challenging. Further progress is growth programs. A handful of its metro- needed in developing and refining policies politan areas on the eastern edge of the that both build on past experience and try

40 policy focus r e p o r t ● L i n c o l n I nstitute of Land Po l i c y ingram & hong ● Evaluating smart growth 41 ......

Houston, Texas

new approaches to policy implementation infrastructure capacity existed largely on that are politically feasible, technically the fringes of urban areas, and that the sound, and economically efficient. concurrency policy helped to force develop- ment to those areas and thus exacerbated R ecomme n datio n s o n sprawl. While infill development was a F u n ctio n al L i n kages a n d policy objective in Florida, needed infra- P rogram D esig n structure investments were not forthcoming. The design of growth management As another example, none of the eight policies should take account of inter- case study states requires government actions among policies and coordinate approval of conservation easements, raising well across relevant agencies. the possibility that such easements can be In some cases, policies are put in place that inconsistent with existing land use plans. focus narrowly on specific objectives. But those policies are part of a larger framework Smart growth policies should make and may have potential synergies or antago- use of economic incentives, such as nisms with other policies. pricing and tax policies, that have One outstanding example is Florida’s shown promise in other countries. concurrency policy, which required that devel- While Maryland has instituted policies that opment can occur only where there are ade- use incentives, such as facilitating growth quate infrastructure and other public facilities. in areas already served by infrastructure, In retrospect, it is clear that Florida’s unused relatively few programs have used pricing

40 policy focus r e p o r t ● L i n c o l n I nstitute of Land Po l i c y ingram & hong ● Evaluating smart growth 41 ......

mechanisms to attain smart growth objec- tion of affordable housing is an achievable tives. In transportation, for example, no and important objective. U.S. city has implemented congestion fees Other aspects of smart growth programs, along the lines of those in London and especially those associated with transporta- Singapore, even though the evidence indi- tion, are also likely to produce benefits for cates that these programs have successfully low-income residents in terms of improved reduced peak-period congestion and pro- access to employment or lower travel costs. moted transit use. Policies and programs that promote walking, While many municipalities levy impact cycling, and transit accessibility and use are fees on new development, these charges are likely to expand the opportunities of low- based primarily on estimates of the capital income households, many of whom may investment attributable to the development, have limited transportation options. Improv- not on estimates of negative externalities ing transit accessibility more generally is also (such as increased congestion, emissions, likely to benefit low-income workers who and greenfield development). Several muni- are current transit users. cipalities do impose additional charges or levies, including the requirement that devel- R ecomme n datio n s o n opers provide a percentage of affordable P rogram S u stai n abilit y dwellings within a larger project or contrib- a n d M o n itori n g ute to an affordable housing fund. However, Because smart growth programs outside of proposals such as cap-and-trade must be implemented over a long schemes that allow trading of emissions, period to achieve results, success prices and fees have modest roles in smart requires credible governmental growth plans. This is an area that deserves commitment to these policies. more focus. Some of the smart growth states included in this evaluation benefited from consistent Smart growth programs need to application of policies over time, while consider carefully the income distri- others did not. Oregon has sustained its bution consequences of their policies. focus on the adequacy of state-mandated With the exception of affordable housing, plans for development. In contrast, Mary- smart growth programs generally pay little land’s Governor Glendening implemented attention to their income distributional effects. an innovative set of smart growth policies While most smart growth programs do in- that his immediate successor largely ignored. clude affordable housing elements, these New Jersey has made a highly credible policies are working well in only a few states commitment to one aspect of smart growth —most notably New Jersey. Affordable hous- programs following the state supreme court’s ing policies are critical because many smart Mt. Laurel decisions requiring implementation growth program elements have been asso- of policies to promote affordable housing. ciated with higher housing prices. Many observers attribute the state’s success Whether price pressures come from the in providing affordable rental housing to the demand side (because of the greater attrac- subsequent oversight of policies and imple- tiveness of smart growth areas) or the supply mentation that executive decision making or side (because new constraints on low-density inattention cannot undo. The need for cred- housing are not offset by incentives for multi- ible governmental commitment strengthens family housing) is still unclear, but produc- the preference for state-level smart growth

42 policy focus r e p o r t ● L i n c o l n I nstitute of Land Po l i c y ingram & hong ● Evaluating smart growth 43 ......

policies, since it may be difficult to make about the interactions among land use, trans- such commitments at the local or even portation, and emissions, as well as between regional levels. land use policies and housing affordability. Policies are being put in place whose success Measurement and collection of data, depends on specific expectations about such particularly related to environmental interactions. For example, although there is quality and public finance, are inad- still substantial uncertainty about the mag- equate and need to be improved. nitude of the relationship between residen- Several objectives of smart growth prog- tial densities and vehicle miles traveled, the rams, particularly those related to the envi- results expected from many smart growth ronment, lack sufficient data to allow formu- policies depend importantly on assumptions lation of performance indicators that can be about the nature of this relationship. tracked over time or compared across juris- dictions. State-level data are particularly Clearer definition of performance weak for assessing air and water quality and indicators and measurement of their for creating measures related to flora and attainment would facilitate the evalu- fauna. Comparable data are currently avail- ation of smart growth programs and able only for performance measures related contribute to their technical and to land uses. political sustainability. In the area of local public finance, exist- This evaluation makes clear that few smart ing datasets should be able to support the growth programs define specific performance formulation of performance indicators, but indicators to measure the success of their they are missing extensive amounts of data policies. In addition, the states have not and do not readily sustain comparisons over collected the data necessary to monitor their time or across states. performance or to carry out an evaluation Finally, data on population and many of overall program effectiveness. Oregon’s related measures are collected for geograph- benchmark reporting system and Maryland’s ic areas whose boundaries change over time. National Center for Smart Growth Research Reconfiguring zonal systems to make them and Education are among the rare efforts consistent over time is very costly for indi- to create monitoring capacity. vidual studies and would be done more Smart growth programs benefit from efficiently at a centralized level. well-defined and measurable objectives, as well as from procedures to monitor relevant Smart growth programs would ben- performance measures that demonstrate efit from more research on important how the programs can achieve those objec- interactions among policies used to tives. Effective monitoring can help main- achieve different objectives. tain voter support for smart growth policies Much is known about the determinants of and guide adjustments when the evidence urban form and household behavior in urban indicates that policies are not working. settings. However, stronger evidence is needed

42 policy focus r e p o r t ● L i n c o l n I nstitute of Land Po l i c y ingram & hong ● Evaluating smart growth 43 ......

references

DeGrove, John M. 1984. Land, growth and politics. Oregon Transportation and Growth Manage- ______. 2000a. 2000 census summary file 3. www. Washington, DC: Planners Press. ment Program. 2009. 2007-2009 Biennial Report. census.gov/Press-Release/www/2002/sumfile3.html http://www.oregon.gov/LCD/TGM/docs/ ———. 1992. The new frontier for land policy: Planning 0709biennialreportfinal.pdf ———. 2000b. Census 2000: Summary tape file .1 and growth management in the states. Cambridge, MA: http://factfinder.census.gov Lincoln Institute of Land Policy. Richardson, Jesse J., and Meghan Z. Gough. 2003. Is home rule the answer? Clarifying the influence ______. 2000c. Census of population and housing. ———. 2005. Planning policy and politics: Smart of Dillon’s Rule on growth management. Washington, Washington, DC: Government Printing Office. growth and the states. Cambridge, MA: Lincoln DC: Brookings Institution. Institute of Land Policy. ______. 2000d. Census of population income survey. Richmond, Henry. 2007. Oregon’s public Washington, DC: Government Printing Office. Fulton, William, Rolf Pendall, Mai Nguyen, investment in conservation, prosperity, and fairness. and Alicia Harrison. 2001. Who sprawls most? www.oregonpublicinvestment.com/ ———. 2000e. Means of transportation to work for How growth patterns differ across the U.S. Washington, workers 16 years and older (Question P30 in the 2000 DC: The Brookings Institution, Survey Series. Rohse, Mitch. 1987. Land use planning in Oregon: Decennial Census). A no-nonsense handbook in plain English. Corvallis: GeoLytics. 2002. CensusCD 1990 long form in 2000 Oregon State University Press. ———. 2002. U.S. census of governments. Washing- boundaries. East Brunswick, NJ: GeoLytics, Inc. ton, DC: Government Printing Office. Smart Growth Network. 2009. Smart Growth Glaeser, Edward L., Matthew E. Kahn, and Online. http://www.smartgrowth.org/about/principles/ ———. 2006. Census zip code business patterns: Jordan Rappaport. 2008. Why do the poor live default.asp 2004. http://factfinder.census.gov in cities? The role of public transportation. Journal of Urban Economics 63:1–24. Trust for Public Land. 2005. Value of voter-approved ———. 2007. Revised intercensal population estimates. open space and tax measures, 1999 to 2004. TPL www.bea.gov/regional/spi/default.cfm?satable=summary Gyourko, Joseph, Albert Saiz, and Anita Summers. LandVote Database and Mapping System. 2006. A new measure of the local regulatory environment www.tpl.org U.S. Department of Agriculture. 1982. Summary for housing markets: The Wharton residential land use report of the natural resources inventory. http://www.nrcs. regulatory index. Working Paper 558. Philadelphia: U.S. Census Bureau. 1980. Census of population and usda.gov/technical/NRI/ University of Pennsylvania, Zell/Lurie Real housing. Washington, DC: Government Printing Estate Center. Office. ———. 1987a. Summary report of the natural resources inventory. http://www.nrcs.usda.gov/technical/NRI/ Ingram, Gregory K., Armando Carbonell, Yu- ———. 1982. U.S. census of governments. Washing- Hung Hong, and Anthony Flint. 2009. Smart growth ton, DC: Government Printing Office. ———. 1987b. U.S. census of agriculture. policies: An evaluation of programs and outcomes. Cam- www.agcensus.usda.gov/ bridge, MA: Lincoln Institute of Land Policy. ———. 1990a. 1990 census summary tape file 3. http://factfinder.census.gov/servlet/DatasetMain ———. 1992. Summary report of the natural resources Land Trust Alliance. 2005. 2005 national land PageServlet?_ds_name=DEC_1990_STF3_&_ inventory. http://www.nrcs.usda.gov/technical/NRI/ trust census report. www.landtrustalliance.org/about-us/ program=DEC&_lang=en. land-trust-census ———. 1997. Summary report of the natural resources ———. 1990b. Census 1990: Summary tape file 1. inventory. http://www.nrcs.usda.gov/technical/NRI/ ———. n.d. 1990 Census Data Cumulative http://factfinder.census.gov Totals. Unpublished. ———. 2000. Summary report: 1997 national resourc- ———. 1990c. Census of population and housing. es inventory. Washington, DC, and Ames, Iowa: Lynch, Lori, Karen Palm, Sabrina Lovell, and Washington, DC: Government Printing Office. Natural Resources Conservation Service and Jay Harvard. 2007. Using agricultural and forest land Iowa State University, Statistical Laboratory. values to estimate the budgetary resources needed to triple ———. 1990d. Census of population income survey. Maryland’s preserved acres. College Park: University Washington, DC: Government Printing Office. ———. 2002. U.S. census of agriculture. of Maryland, Harry R. Hughes Center for www.agcensus.usda.gov/ Agro-Ecology, Inc. ———. 1990e. Means of transportation to work for workers 16 years and older (Question P045 in the ———. 2003. National resources inventory 2003 National Association of State Park Directors. n.d. 1990 Decennial Census). Annual NRI. www.nrcs.usda.gov/technical/NRI/2003/ Summary of state park acreage. Unpublished. Landuse-mrb.pdf www.naspd.org ———. 1992. U.S. census of governments. Washington, DC: Government Printing Office. Woods and Poole Economics, Inc. 2005. Complete economic and demographic data source (CEDDS). Wash- ———. 1996. Census zip code business patterns: ington, DC: Woods and Poole Economics, Inc. 1994. http://factfinder.census.gov

44 policy focus r e p o r t ● L i n c o l n I nstitute of Land Po l i c y Ordering Information To download a free copy of this report or to order copies of the printed report, visit www.lincolninst.edu and search by author or title. For additional information on discounted prices for bookstores, multiple-copy orders, and shipping and handling costs, send your Lincoln Institute of Land Policy inquiry to [email protected]. www.lincolninst.edu

Lincoln Institute is a private operating foundation whose mission is Production Credits P RO J E C T M A NAG E R & E ditor to improve the quality of public debate and decisions in the areas Ann LeRoyer of land policy and land-related taxation in the United States and D E S I G N & P roduction : DG Communications/NonprofitDesign.com around the world. The Institute’s goals are to integrate theory and P rinting : practice to better shape land policy and to provide a nonpartisan Recycled Paper Printing, Boston

forum for discussion of the multidisciplinary forces that influence Printed on recycled paper using soy-based inks. public policy. The Institute seeks to inform decision making through education, research, demonstration projects, and the dissemination of information through publications, our Web site, and other media. Photographs

Lincoln Institute programs bring together scholars, practitioners, F ront C over public officials, policy advisers, and involved citizens in a collegial Top left: JupiterImages.com Top right: Travel Portland learning environment. Bottom left: Photos.com Bottom right: JupiterImages.com

Other IndexOpen.com: 2 JupiterImages.com: 4, 7, 9, 37 Northern Light Productions: 12 Ann LeRoyer: 13, 36 iStockphoto.com: 15, 18, 19, 22, 29 Anne Battis: 24, 39 Alan Mallach: 26 Gerrit Knaap: 33 Ming Zhang: 41

113 Brattle Street Cambridge, MA 02138-3400 USA

Phone: 617-661-3016 or 800-526-3873 Fax: 617-661-7235 or 800-526-3944 Web: www.lincolninst.edu Email: [email protected]

44 policy focus r e p o r t ● L i n c o l n I nstitute of Land Po l i c y Evaluating Smart Growth State and Local Policy Outcomes

The findings of this report support three sets of recommendations that may be useful to state and local policy makers as they formulate smart growth objectives in the current context of high energy costs, historic housing market volatility, and increasing pressures to reduce greenhouse gas emissions.

Program Structure and Transparency Sustainability and Monitoring of Programs • The design of smart growth programs and • Credible commitment from different levels of supporting regulations and incentives should be government is crucial for the successful imple- guided by a vision of sustainable and desirable mentation of smart growth programs. development outcomes. • Improvements in measurement and collection of • Any top-down or bottom-up smart growth policies data, particularly related to environmental quality must be coordinated at the regional level to be and public finance, are needed to better monitor able to achieve their desired objectives. program performance. • Policy makers must articulate the means of • More evidence is needed about the nature achieving smart growth objectives and specify of interactions among smart growth policies— implementation mechanisms, rather than just particularly those related to land use, trans- declare objectives. portation, and housing affordability. • Clearer definition of performance indicators Functional Linkages for Policy Design and measurement of their attainment would • The design of growth management policies should facilitate the evaluation of smart growth take account of interactions among policies and programs and contribute to their technical coordination across relevant agencies. and political sustainability. • Smart growth policies should make use of eco- nomic incentives, such as pricing and tax policies, that have shown promise in other countries. • Smart growth programs need to consider the income distribution consequences of their policies.

ISBN 978-1-55844-193-4

ISBN 978-1-55844-193-4 Policy Focus Report/Code PF020