INFORMATION TO USERS

This manuscript has been reproduced from the microfilm master. UMI films the text directly from the original or copy submitted. Thus, some thesis and dissertation copies are in typewriter face, while others may be from any type of computer printer.

The quality of this reproduction is dependent upon the quality of the copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bfeedthrough. substandard margins, and improper alignment can adversely affect reproduction.

In the unfikety event that the author did not send UMI a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion.

Oversize materials (e.g., maps, drawings, charts) are reproduced by sectioning the original, beginning at the upper left-hand comer and continuing from left to right in equal sections with small overlaps.

Photographs included in the original manuscript have been reproduced xerographically in this copy. Higher quality 6" x 9” black and white photographic prints are available for any photographs or illustrations appearing in this copy for an additional charge. Contact UMI directly to order.

ProQuest Information and Learning 300 North Zeeb Road. Ann Arbor, Ml 48106-1346 USA 800-521-0600

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. NOTE TO USERS

This reproduction is the best copy available.

UMf

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. SYMBOLS AND SUBSTANCE: HOW BLACK CONSTITUENTS ARE COLLECTIVELY REPRESENTED IN THE UNITED STATES CONGRESS THROUGH ROLL-CALL VOTING AND BILL SPONSORSHIP

DISSERTATION

Presented in Partial Fulfillment of the Requirements for

The Degree of Doctor of Philosophy in the Graduate

School of The Ohio State University

By

Valeria Sinclair-Chapman. M_A.

The Ohio State University 2002

Dissertation Committee:: Approved by Professor Janet M. Box-Steffensmeier. Adviser

Professor William E. Nelson C/ Adviser Political Science Graduate Program Professor Herbert Weisberg

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. UMI Number 3049115

Copyright2002 by Sindair-Chapman, Valeria Michelle

Ail rights reserved.

UMI*

UMI Microform 3049115 Copyright 2002 by ProQuest Information and Learning Company. Alt rights reserved. This microform edition is protected against unauthorized copying under Title 17. United States Code.

ProQuest Information and Learning Company 300 North Zeeb Road P.O. Box 1346 Ann Arbor. Ml 48106-1346

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Copyright by- Valeria Sinclair Chapman 2002

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ABSTRACT

Representing the policy preferences of any group is a complex and dynamic

enterprise. Representatives make decisions about what group interests to publicly

support and which legislative tools to use to communicate that support to group members

and colleagues. The representation of group preferences can be substantive, i.e.

narrowly-tailored to achieve tangible, relatively immediate, policy outcomes. Or. the

representation of group preferences can be symbolic, i.e. broadly-tailored to give group

members voice in the legislature. This study expands conventional approaches to black

representation by developing a theory of symbolic representation that accounts for

legislative activities undertaken with the objective of giving psychological reassurance to

group members that representatives are working in their interests and responsive to their

needs without the condition that measurable policy outcomes be an immediate goal.

Using data I collected on House membersr roll-call voting behavior and bill

introductions, as well as data from a recent national black opinion survey. I create several

measures of substantive and symbolic representation to analyze how House members

represent black interests at the policy-making and agenda-setting stages of the legislative

process. 1 find that descriptive representatives (those sharing similar social

characteristics, e.g.. race) and partisan representatives (those sharing the same party

affiliation) represent black policy preferences differently depending on the stage of the

n

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. legislative process being examined and whether the representation is symbolic or

substantive.

Descriptive representatives1 use of bill sponsorship to advocate black concerns is

essential to black representation. Importantly, rney are considerably more likely than

their colleagues to advocate black concerns symbolical!}. Descriptive representatives

behave less distinctively than their colleagues at the level of roll-call voting and

occasionally cast votes that are contrary to the substantive representation of black

concerns. New survey data shows that although blacks are generally supportive of black

interest policy proposals, their preferences for symbolic or substantive policy proposals

are marginally influenced by the type of representation—descriptive or partisan—which

they receive.

iii

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Dedicated to:

Chris and David Chapman

James and Serena Sinclair

Jada Louise Lamb (1969-1997)

IV

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ACKNOWLEDGMENTS

I wish to thank my adviser. Janet Box-Steffensmeier for intellectual support and

encouragement. Early on. she believed in the possibilities tor this project when I myself

had doubts. Her enthusiasm about m y work and love for quality scholarship made this

thesis possible.

I thank William E. Nelson for reliably asking tough questions and helping me son

through the answers. Prolessor Nelson's Knowledge is boundless and his intellectual

demands have sharpened the focus of this project. His unwavering confidence and

support have been invaluable since the beginning of my graduate career.

I thank Herbert Weisberg and Samuel Patterson for lending their incredible

knowledge and expertise to this project. These men generously rearranged their

schedules to help facilitate the completion of this dissertation and their willingness to do

so is greatly appreciated.

I also wish to thank Katherine Tate for her support and encouragement throughout

this project. Even after leaving my graduating institution. Katherine has continued to

offer intellectual and personal insight. She has been instrumental in teaching me what it

means to be a professional in this discipline.

I also wish to acknowledge the support and assistance of my colleagues at the

University of Rochester. Dick Fenno. Gerald Gamm. Fred Harris. James Johnson.

v

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Richard Niemi. and Christian Grosce provided valuable advice and comments at various

stages o f the writing process.

A special thanks is owed to Lauren Parr and Bronwen Van Hooft for helping me

get organized. For moral support. I thank Rhonda and Rick Parr. Angela Hemphill.

Thehna Jackson. Sabrina Miller. Donald Wilson. Gloria Hampton. Melanye Price.

Yolanda Keller-Bell. and more recently. Gay Byron. I am indebted to Gloria and

Melanye for their willingness to shuttle paperwork back and forth to the graduate school

while 1 was hundreds of miles away in .

I am deeply indebted to Chris Chapman for his patience and support during the

highs and lows of my completing this project. He was my partner long before this project

began and for more than a decade he has helped me seek my own center and define who I

really am. I thank my son. David Chapman, for his innocent demands to play in my

office, type on the computer, and sit on my lap for a cuddle while I was working in the

wee hours of the morning. Throughout this occasionally trying process, he has ever

reminded me of life's greatest joy—to love and be loved in return. I thank my parents.

Jim and Serena Sinclair, for their support—monetary and emotional—throughout

graduate school and for their unending confidence that I could do this. I thank my

brother. Tim. for his love, support, and listening ear. His confidence in me has

encouraged me to always aim higher. I admire him more than he could possible know. I

also thank Jada Lamb for her lasting friendship despite distances separating us through

time and space.

VI

jtr " m i. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. I thank Paul Gariington and Heartsight for spiritual guidance and support as 1

neared the end of this project. And most importantly. I thank God for the divine plan that

has allowed this experience to unfold exactly as it should.

This research was supported by an award from the Madison Scott Memorial

Foundation at The Ohio State University and by a Ford Foundation Dissertation

Fellowship. I might never have started this research agenda, and certainly would not

have been so creative in my approach, without the opportunities for learning and research

provided by the Women's Research and Educational Institute (WREI) congressional

fellowship program

vii

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. VITA

March 22. 1969...... Bom— Landover. Maryland.

199 1...... B.A.. Political Science. University of at Asheville.

199 2...... M_A„ Political Science. The Ohio State University.

1992-1993...... Women's Research and Education Institute (WREI> Congressional Fellow.

1994-199 7...... Teachinr and Research Assistant* The Ohio State University.

1995-199 6...... Research Assistant. Polimetrics Research Center

1999-200 0...... Ford Foundation Dissertation Fellow.

2000-Presen t ...... Instructor. University of Rochester.

FIELDS OF STUDY

Major Field: Political Science

viii

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. TABLE OF CONTENTS

Page Abstract...... ii

Dedication ...... iv

Acknowledgments ...... v

Vita ...... viii

List o f Tables ...... xii

Chapters:

I. Introduction ...... I

1.1 Types of Black Representation ...... 5 1.1.1 Descriptive Representation ...... 6 1.1.2 Symbolic Representation ...... 8 1.13 Substantive Representation ...... 12 13 Defining Black Interests ...... 15 13 Toward a Theory of Black Representation ...... 16 1.4 Research Methods and Data ...... 20 1.5 Chapter Outline...... 21 1.6 Conclusion ...... 22

Representing Racial Group Interests ...... 24

2.1 Racial Group Representation and the Electoral Connection ...... 25 2.2 The Electoral Connection and Ideological Congruence ...... 29 2 3 The Electoral Connection and Partisan Congruence ...... 33 2.4 The Electoral connection and Descriptive Representation ...... 35 23 Conclusion ...... 37

3. Black Representation Through Roll Call Votes ...... 40

3.1 Analyzing Black Substantive Representation ...... 42

IX

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 3.1.1 Policy Congruence Analysis of Black. Interests ...... 45 3.1 J? The Black Opinion Inaex ...... 47 3.2 Data and Methods ...... 50 33 Results and Discussion ...... 52 3.3.1 Comparing LCCR and BOI Scores ...... 55 33.2 Difference o f Means for LCCR and BOI Scores ...... 57 3 3 3 Support for Subjective Black Interests by Race and Pam - ...... 58 3.4 Conclusion ...... 59

4. Black Representation Through Bill Sponsorship ...... 67

4.1 Bill Sponsorship and Representation...... 70 4.2 Theoretical Consideration: Symbolic Representation ...... 75 43 Data and Methods ...... - 80 4.4 Results and Discussion ...... 83 4.5 Conclusion ...... 88

5 Black Support for Substantive and Symbolic Bills ...... 97

5.1 Analyzing Black Policy and Representation Preferences ...... 98 5.1.1 Black Assessments of the Value of Legislative Activities ... 100 5.1.2 Legislator Expectations and Representation Strategies 104 53 Data and Methods ...... 108 5.3 Results and Discussion ...... 110 5.4 Black Preferences for Symbolic and Substantive Policy Proposals.. 115 5.5 Conclusion ...... 117

6. Conclusion ...... *...... 125

6.1 Black Representation through Bill Sponsorship ...... 129 6.2 Black Representation through Roll Call Votes ...... 131 63 Black Preferences for Symbolic and Substantive Policies ...... 133 6.4 Directions for Future Research ...... 133 6.5 Conclusion: Symbols and Substance in Black Representation 135

Appendix A. Description of NBES Survey Data ...... 137 Appendix B. Summary Descriptives of U.S. House Representatives in Sample— 139 Appendix C. Survey Questions Used to Create the Black Opinion Index (BOI) 142

x

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Appendix D. Roll-Call Votes Used to Create the Black Opinion Index (BOI) 144 Appendix E. Issues in Leadership Conference on Civil Rights (LCCR) Scores—. 145 Appendix F. Measurement of Variables ...... 146 Bibliography ...... - 147

XI

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. f

LIST OF TABLES

Table Page

3.1 Ordinary Least Squares Regression Model of Representatives' Support for Objective Black Interests (LCCR Scores >. 104 Congress ...... 64

3.2 Ordered Probit Regression Model of Representanvesr Support for Subjective Black interests (BOI Scores). 104“ congress ...... 65

3.3 Mean LCCR and BOI Scores by Race and Party of Representative. 104th Congress ...... 65

3.4 Percent Support for Black Preferences on BOI Issues by Representatives’ Race and Party. 104th Congress ...... 66

4.1 Sponsors of Black Interest Bills (One or More) by Member Race and Party, 104th Congress ...... 93

4.2 Percentage of Black Interest Bills Sponsored by Member Race and Party, 104th Congress ...... 94

43 Types of Bills Introduced as a Percentage of All Black Interest Bills Sponsored by Member Race and Party, 104th ...... 94

4.4 Ordinary Least Squares Regression Model of Direct and Indirect Black Interest Bills, 104th Congress ...... 95

4 3 Ordinary Least Squares Regression Model of Symbolic and Substantive Black Interest Bills, 104th Congress ...... 96

5.1 Black Support for Hypothetical Black Interest Bills ...... 120

53 Black Support for Black Interest Bills by Type of Representation 121

53 Black Support for Black Interest Bills by Ideological. Partisan, and Race Identification ...... 122

xii

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 5.4 The Effect of Descriptive and Partisan Representation on Support for Substantive Bills ...... 123

5.5 The Effect of Descriptive and Partisan Representation on Support for Symbolic Bilk ...... 124

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER 1

INTRODUCTION

The relevance of race to political representation has long been a topic of study in

black politics. Many black politics scholars assumed, either implicitly or explicitly, that

black elected officials would best represent the political interests of black constituents

(Walters 1992. Parker 1990. Barker 1988. Marable 1985. Carmichael and Hamilton

1967). However, exponential increases in the number o f blacks in elected office at all

levels of government, along with recent scholarship on race-based districting and

representation have led inexorably to the question of whether ensuring more black faces

in elective office really is a requisite of adequate black political representation at the

congressional level (Lublin 1997. Swam 1994. Guinier 1992. Themstrom 1987. Davidson

1984. Grofman 1982).

What role, if any. does the race of the representative play in shaping the tendency

to advocate on behalf of black constituents? This question is important because race is a

major source of social cleavage in American politics. The fact that the courts continue to

grapple with constitutional questions regarding how the rules governing fair

representation in the Voting Rights Act should be interpreted and how they should be

implemented by Congress and the states points out how extremely relevant race is to

questions o f fair and equal representation for minorities.

I

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Is there a tradeoff between descriptive representation and the substantive

representation of black interests? The political debate rases on with regard to the lone

and short-term goals of redistricting to achieve greater minority representation- The tact

remains that the number of blacks in elected office has increased considerably since the

implementation of the 1965 Voting Rights Act and continues to increase moderately in

response to continued efforts, e.g.. Kousser 1992, Cavanagh 1992, Williams 1990. Still,

research has revealed a cleavage produced within the Democratic Party by the diverging

interests of white Democrats seeking to maintain incumbency and minority Democratic

candidates aggressively seeking supportive constituencies < Gameron. et. al.. 1996. Lublin

1997. Grofman. Griffin, and Giazer 1992). And. researchers continue to find mixed

results regarding whether tne creation of majority-black and majoritv-minority districts

helps or harms black representation goals (Cameron et. al. 1996. Epstein and CFHallorar 8 1996. Lublin 1997, Hill 1996. Swain 1993).

Scholars of gender politics (Barrett 1995. Vega and Firestone 1995. Thomas

1994), and minority politics (Hero and Tolbert 1995. Tate 1994, Swain 1993. Hero 1992.

Bobo and Gilliam 1990. Browning, Marshall, and Tabb 1984. Welch and Hibbing 1984)

have sought to directly examine the issue of descriptive representation in the

representation of interests. At the state politics level, research has concluded that blacks

and women do sometimes present different and more descriptively reflective policy

agendas than non-group members (Bratton and Haynie 1999, Bameilo and Bratton 1998,

see also Thomas 1994. Nelson 1991, Thomas 1991. Thomas and Welch 1991. Miller

1990). For instance. Bratton and Haynie (1999) conclude that blacks and women share

distinctive policy interests in bill sponsorship, even alter controlling for constituency

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. influence. There is also evidence that racial differences among blacks and whites exist in • ** roll-call voting and bill sponsorship (Bamello and Bratton 1998). At the congressional

level, results are mixed regarding whether the race of the representative makes a

difference in how well black interests are represented. Carol Swain (1993) was the first

to conclude, using empirical evidence, that the race of the representative was not

necessarily relevant for the representation of black interests. Other research has revealed

differences between black and white legislators voting behavior on specifically racial

issues (Whitby and Krause 1998, Whitby 1997. Hall and Heflin 1994).

Substantive representation research among scholars of black representation has

resulted in a conceptualization that rests almost exclusively on evidence of congruence

between representatives1 voting behavior and liberal ideological measures or scholars

objective assertions of black interests based, for instance, on national statistics that showr

higher raies of poverty among blacks. If we are to advance our understanding of black

representation, it behooves scholars to broaden the definition of black interests and to

explore the remaining components of representation. Representatives can conceivably

perform well in one aspect of representation and fail dismally in another. As it stands, we

have no idea how these various components tit together in the minds of constituents. It is

possible that an interaction between two or more of these components contributes to

making constituents “feel” represented, while lopsided performance on any one results in

just the opposite.

This dissertation expands notions of black representation in several ways. First. I

consider substantive representation in terms of both objective and subjective interests.

Rather than asking only, “who represents what black Americans need?” based on

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. objective indicators.” I also consider, “who represents what black Americans want?”

based on aggregated black opinion poll results. These two concepts are discussed more

fully later in this chapter and empirically tested in chapter 3.

Second. I develop a theory ot symbolic representation that accounts for the

symbolic actions that representatives engage in to give voice to group concerns. Bill

sponsorship behavior among white and black representatives provides an arena for

theory-building and empirical testing in chapter 4. I investigate what kinds of black

interest bills are sponsored by legislators and then code bills on three measures: I) Do

they broadly reflect objective black interests te.g.. poverty remedies, education, and civil

rights)? 2) Do these bills directly or indirectly reflect black interests i i.e.. whether blacks

are primary or secondary beneficiaries)? 3) Are the bills substantive or symbolic in focus

(i.e.. whether tangible policy outcomes would or would not result if the bill were

enacted)? This approach adds a balanced alternati ve to explaining black representation in

purely substantive terms. Explored specifically in the context of black representation* the

theory developed here has implications for our understanding of what representation is in

the larger American political context.

Finally. 1 use survey data to consider how African American citizens respond to

substantive and symbolic overtures by African American representatives. Are some

blacks more responsive to symbolic overtures than others? What role does group

identification or assessments o f “common fate” play in determining how blacks respond?

The answers to these questions will shed light on how substantive and symbolic

representation efforts are received by various elements of the black community.

4

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The use of bill sponsorship data as well as roll-call votes, in addition to a

connection to recent black opinion survey data give this study a unique focus on the issue

of black representation. Together, this triangular approach will contribute to a more

thorough understanding of the substantive and symbolic components of representation

and should make an important contribuuon to the literature on legislative studies,

congressional representation, and black representation, in particular.

1.1 Types of Black Representation

Political scientists have relied on the seminal work of theorist. Hannah Pitkin (1967).

to guide them in systematically identifying and cataloging types ot representation when

they occur. She identified three modes of representation - descriptive (shared social

traits like gender, religion, and race), symbolic (symbol-making to engender trust), and

substantive (acting in the interests of the represented), two of which, namely descriptive

and substantive, have dominated studies of black representation.

Ten years later. Eulau and Karps (1977) built on Pitkin’s work and that of several

other theorists and empirical scholars to better specify the components of representation

for empirical study in political science. They identified four primary components of

representation, or more accurately, responsiveness: policy (responsiveness to policy

concerns of constituents often through voting), service (responsiveness to individual or

group needs in the constituency through securing particularized benefits), allocation

(responsiveness to district needs through pork-barrel politics), and symbolic

(responsiveness to the affective needs of constituents through the use of public gestures

which engender trust). The approach that is generally used by scholars studying black

5

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. representation draws largely from these two sources and results in the loose typology

found in the next sections.

Before moving on to a more detailed discussion, a brief model o f the three central

types of representation used in the analysis of black representation is presented. If

representation were envisioned as a vertical continuum, descriptive representation would

take residence at the low end as it the easiest criteria that a representative may fulfill,

requiring no effort on the pan of the representative to actually represent except that she

simply be herself and let her obvious or self-promoted social demographic characteristics,

e.g.. race, gender, or religion, speak for her. Substantive representation would reside at

the high end as it has the most impact on the tangible needs and wants of the constituents

and arguably requires the greatest degree of public commitment to represent constituent

interests. Symbolic representation would fall someplace in between the two extremes as

it requires the representative to undertake some degree of action on behalf of the

represented be it through speech-making or bill sponsorship in Congress, or through the

use of body language or word choice. Representatives use activities on the Hill and other

subtle cues at home in the district to demonstrate as symbolic gestures that say to

constituents, “I am one of you” (Fenno 1978.115).

1.1.2 Descriptive Representation

Descriptive representation can be conceived of as “representation by the

numbers.” Proponents of black descriptive representation ground their arguments for

increased numbers of blacks in elected office on two complementary propositions. First,

is the argument that a truly representative assembly must include blacks, a group once

6

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. fonnaUv excluded from participation in legislatures, in roughly the same proportion as

their numbers in society to be considered legitimate by the previously excluded group.

Second is the argument that black interests are best represented by black elected officials

who. in addition to possessing simple congruence of skin-color, are likely to have

intimate knowledge of or shared experiences with other blacks regarding for instance,

discrimination (Walters 1992, Parker 1990. Barker 19$8, Marable 1985. Carmichael and

Hamilton 1967V

Noting that the percent of blacks m congress falls significantly short of their

percent in the population, descriptive representation advocates question whether the

legislature is truly "representative" and call lor proactive steps to increase the number of

blacks in elected office. Many empirical scholars have questioned whether there is a

direct relationship between descriptive representation and increased substantive

representation of black interests. While shared racial identity is not a necessary or

sufficient predictor of black representation, the jury is still out regarding the precise

nature of the relationship between descriptive and substantive representation (Lublin

1997. Swain 1994. Guinier 1992, Thernstrom 1987. Davidson 1984. Grofman 1982).

Beginning in the early 1990s, discussion regarding black representation evolved

from an emphasis on descriptive representation to an emphasis on substantive

representation, that is. evaluating the extent to which black policy preferences are

supported by their representatives, whether black or white. Still, recent trends have found

both theorists and empirical scholars returning to previous questions regarding the value

of descriptive representation (Canon 1999. Mansbridge I999.WilIiams 1998, Whitby

1997). As Kenny Whitby aptly states in The Color o f Representation. “the interaction of

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. descriptive and substantive representation may be crucial because of the possible

connection between black constituentsr desires and the racial composition of the

legislature" (1997.6). * *

1.1.2 Symbolic Representation

To bener understand the nature of black representations it is essential to examine

its symbolic dimensions. Symbolic acts aimed at giving voice to group interests, agenda-

setting. and offering alternative views of political possibilities are integral to enhanced

political deliberation that addresses the concerns o f disadvantaged groups. According to

Mansbridge (1999) disadvantaged groups benefit from descriptive representation in four

contexts- two of which add to the quality of deliberation in legislatures through 1)

facilitating communication in contexts of mistrust and. 2) facilitating innovative thinking

in contexts of uncrystallized. not fully articulated interests.1 As I describe in chapter 4.

symbolic representation is particularly important as a source o f innovative thinking that

defines political alternatives to the status quo. Symbolic representation also serves to

highlight the achievements and experiences of disadvantaged group members that

otherwise might not be recognized by members of advantaged groups. Elected officials

who represent blacks symbolically are at once communicating empathy with black

constituents and advocating black concerns in the legislature. Symbols are. in fact, the

basic currency of political exchange between representatives and their constituents as

1 The other two contacts relate to the benefits of representation outside o f the legislature. Descriptive representation provides opportunities for members of previously excluded groups to demonstrate their ability to rule and act as role models for other group members. 8

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. weii as between representatives and their colleagues in the legislature (Elder and Cobb

1983. Edelman 1960).

Many scholars have interpreted symbolic representauon as being synonymous

with the concept of descriptive representation- asserting that elected officials who

represent certain demographic characteristics of their constituents are at once fulfilling

both descriptive and symbolic representation roles (Whitby 1987. Smith 1990, Gilliam

1996). Scholars of black politics, lamenting the fact that electing black officials does not

automatically lead to tangible benefits for black constituents, adopted and transformed

this term to refer to “the failure of black tand by extension, any) elected officials to

advance the policy interests of their group (Swain 1993. 1081. Swain (1993) envisions

black representation as a continuum moving from descriptive representation at one end to

substantive representation at the other. Unlike the vertical continuum I presented earlier

in this chapter, Swain's approach does not recognize the importance and distinctiveness

of symbolic representation. As she puts it. symbolic representation is, “descriptive

representation that is not accompanied by substantive representation” (1993,5).

Pitkin (1972), however, suggests a clear distinction between descriptive and

symbolic representation. While she defines descriptive representation as a physical

concept, the “idea of correspondence or likeness... [and] importance of resembling one's

constituents,” she defines symbolic representation in psychological terms based on the

belief of constituents in their representative and the activities that the representative

undertakes to foster belief, trust, and confidence (1972.107).*

2 Swain concedes that the common usage of the term, "symbolic representation” does not aeree with Pitkin's original conceptualization, but also points out the difficultv of using 9

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. In Murray Edelman's 1964 classic, regulatory policy implementation was studied

in several contexts to examine “divergence between the political and legal promises on

the one hand and resource allocation and group reactions on the other...” (1964.23). He

developed the term, symbolic reassurance. to refer to the condition where groups most in

need of promised benefits do not receive them and yet seem quiescent and reassured by

thepromise of tangible goods even in the absence o f measurable change. In those cases,

the regulatory promises served as symbolic reassurance to groups whose interests the

policy supposedly sought to satisfy.

Drawing on Pitkin (1967) and Edelman (I964i. among others. Eulau and Karps

describe the concept ot symbolic responsiveness as that “which involves public gestures

of a son that create a sense of trust and suppon in the relationship between the

representative and the represented” (1977. 241). Appointing minorities to high-level

positions, visiting economically-depressed areas with the press in tow, public hearings,

and fiery floor speeches are all examples of symbolic actions utilized by representatives

from the highest federal offices down to local city councils. Beyond symbolic gestures,

Eulau and Karps suggest that much of the' legislation sponsored at all levels of

government is symbolic, “[having] not the slightest chance of ever being passed and,

more often than n o t,... not intended to be passed" (1977,247).

Elder and Cobb maintain that “symbols are the currency of the [political]

communication process” (1983, 9). To the extent that representatives are able to

the concept in empirical analysis. She captured the essence of the empirical predicament with this concept when she wrote thatr"[u]nlike descriptive representation, which can be discerned by the presence of shared demographic characteristics, or substantive representation, which can be identified through activities, symbolic representation is more ambiguous ...”(1992:108). 10

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. effectively communicate their policy positions and their degree of policy responsiveness

through bill sponsorship, explaining votes cast, and credit-claiming for individual and

I party legislative successes, symbols piay a useful role in making complex legislative

action digestible for constituents who may be iess interested in. or aware of. details than

they are o f the general timbre of their legislator 's activities.

Although symbolic representation is admittedly a concept that is difficult to

operationalize, the examination of symbolic aspects of behavior traditionally thought of

as purely substantive provides one option tor improving measurability. Both voting and

bill sponsorship in legislatures can take on svmbohc qualities. For instance,, a House

member's vote tor or against a Constitutional ban against burning the United States flag

can be conceptualized as a symbolic cue to voters about patriotism on the one hand, or

support for individual freedom on the other On occasion. House members may vote for

legislation or sponsor bills with the objective o f giving voice to constituent preferences or

demonstrating to voters that their concerns really do matter in Washington.

The theory of symbolic representation developed here highlights legislative

activity undertaken with the objective o f giving psychological reassurance to constituents

that representatives are working in their interests and are responsive to their needs

without the condition that measurable policy outcomes be an immediate goal? Here, I

* It seems evident at this stage that the implications of symbolic representation need not be limited solely to questions of black representation. Farm interests could be represented symbolically by representatives who were fanners in their pre-political lives as well as by those who were not. By the same token, black interests can be represented symbolically by representatives who are not black. Classic representation theory, previous research, and empirical data support the effort to isolate and quantify symbolic dimensions of representation. Although I concentrate on the symbolic dimensions of bill sponsorship, future research should consider the svmbolic dimensions of a range of 11 !

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. focus on the symbolic qualities of bill sponsorship and construct a theory of symbolic

representation in that arena. Bills are symbolic when their goals are purely symbolic as

in the naming of a post office alter an historical black figure, as well as when the bill

sponsorship effort itself is symbolic. These purely symbolic and hybrid symbolic bills

will be discussed further in chapter 4.

1.13 Substantive Representation

Much of the current research on the representation of biack interests focuses on

questions of substantive representation Researchers have used policy congruence

analysis of roll-call votes to examine the degree to which representatives respond

favorably to the policy preferences of black constituents. Building on a rich tradition

established by Miller and Stokes' locus on constituency attitudes (1963) and Verba and

Nie's policy focus (1972;. scholars have examined the congruence between

representatives of both races and their black constituencies on a range of ideological

measures and policy preferences.

David Canon (1996) characterizes strict policy congruence theory as the most

demanding of representation theories. Indeed. Robert Bernstein (1989) asserts that the

idea of constituency control is actually a myth that pervades political science theorizing.

He raises the bar to a generally unattainable level when he argues that researchers must

distinguish between “coincidence and control” when evaluating the extent of constituent

control over legislators' actions (for a discussion see Narcisse. 1997). Coincidental

political behavior including campaigning,voting, floor speaking, and co-sponsoring bills. For example, see Hill and Hurley's (2001) analysis of symbolic floor speeches in the Senate. 12

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. congruence occurs when representatives’ votes are the result of some value they share

with constituents. Constituent control occurs only when the representative votes with the

plurality of district opinion and against his own ideological or party preferences. Most

scholars aspire to demonstrate constituent influence rather thancontrol, and thus assert

that there are instances when constituent opinions, preferences, and beliefs do influence

their representatives’ decisions. For instance. Kingdon (1989) finds that constituency

opinions on highly salient issues are quite important to representatives’ decision-making.

Most black representation studies are based on aggregate roll-call analysis, which

regresses ideological group indices like ADA (Amencans for Democratic Action). LCCR

(Leadership Council for Civil Rights), or COPt iLommittee on Political Education)

scores on district characteristics. A primary reason lor the use of these summary

measures is that interest group ratings are easily accessible in one place, while sufficient

survey data on black opinion usually is not. Whitby nicely catalogs the strengths of using

roll-call analysis in the study of black substantive representation: 1) roll-calls can give

researchers a general idea of the level of importance parties and individual legislators

place on black substantive concerns 2) comparisons of legislators’ roll-call votes can

allow researchers to draw valuable inferences about policymaking behavior in relation to

black interest without the heavy costs of interviewing 435 members of the House of

Representatives, 3) roll-call analysis can help establish changing trends in black

responsiveness over time, and tell researchers whose interests are being represented in the

statutes and whose are not (Whitby 1997.9-10).

Despite the overwhelming use of policy congruence analysis by scholars seeking

to understand and evaluate black representation, the method is not without its problems

13

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. (discussed more thoroughly in chapter 3). First. roBscaQ votes are censored samples

where, due to the committee process of weeding out extremely, liberal and conservative

positions, only the middle range o f potential policy proposals are ever voted on in Bouse

floor votes.

The second problem, related to the first, is that black legislators themselves are a

marginalized group within the legislature due to their extremely liberal attitudes, voting

records, and legislative agendas (Guinier 1995). Barker and Jones (1994) assert that even

among the most active black legislators u will be a rare occasion when a black incumbent

can show a record ot tangible accomplishment in meeting the immediate problems of his

or her district. Supporting this assertion is recent empirical evidence at the state-level

that black legislators are considerably less Iikeiy to see their legislation pass (Bratton and

Haynie 1999)/

Finally, the use of liberal interest group scores, which may be only tangentially

related to black interests, presents another quandary for researchers. Analysis based on

ratings of liberalness cannot answer the question of who represents black interests when

those interests are not consistent with the liberal agenda of the Democratic party, nor can

they purport to assess responsiveness to specifically racial concerns when so few of the

votes included in the score cards are actually related to race.

* At the national level. Anderson. Box-Steffensmeier. and Sinclair (20(G) find that blacks were not less likely to get bills passed. In fact, their data show that blacks were slightly more likely than whites to see positive action on the house floor regarding their bills, at a statistically significant level dining the 10?ttt Congress (see also. Canon 1999). 14

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1.2 Defining Black. Interests

Studies o f substantive black interests have primarily analyzed objective black

interests. Objective interests are defined by observable phenomena which affect a

designated group, in this case blacks, in a meaningful and often costly way. Indicators of

objective black interests include labor, poverty, education, and healthcare statistics that

reveal racial disparities among blacks and whites and disproportionate disadvantage

among blacks. For instance, statistics that show black unemployment levels at double and

triple that o f whites, particularly in urban centers, indicates that legislative proposals

regarding employment remedies for the poor and the revitalization o f urban economic

centers are in the objective interests o f blacks (Swain 1994,7). Although objective black

interests are often linked to evidence of racial disparity, they need not be. If black

unemployment is in the double digits, comparisons to white unemployment levels are not

necessary to deem joblessness as a legitimate cause for concern and action in the black

community.

Subjective interests are defined by aggregating the opinions o f group members

regarding a range of racial and non-racial issues. Subjective interests are simply a

measure o f what people or groups say they want According to the 1996 NBES, the

majority o f blacks, 64 percent, support more federal spending to combat crime; therefore,

legislative proposals for building more jails are in blacks’ subjective interests.

It can be argued that objective interests concernneeds in the black community,

while subjective interests concern warns (Canon 1999). Needs often dictate wants,

however, such is not always the case (for a counter-argument see Canon 1999,26-31).

15

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. For instance, high poverty rates in the black community result m a need to combat the

poor living conditions, joblessness, and low wages that plague the black poor. It comes

as no surprise, then, that this need is translated into opinion survey results that reveal that

a majority o f blacks prefer more government intervention in the aforementioned areas.

On the other hand, evidence abounds regarding racial disparities in sentencing and

executions in capital punishment cases. Nevertheless, a plurality o f blacks supports the

death penalty. In the latter case, it would seem that black needs (objective interests) are

contradictory to black wants (subjective interests).

Focusing exclusively on the needs of the black community does not give an

adequate account o f what black people say they want Understanding who advocates for

the wants of a particular constituency is a noteworthy enterprise in its own right. As

Williams (1998, 171) notes, “[t]o be more certain of the relationship between^,

descriptive and substantive representation,~.we require further empirical comparisons o f

the political attitudes o f marginalized groups within the electorate and those o f political

elites.* In chapter 3 ,1 present a comparison of the voting behavior of white and black

political elites from districts with varying black populations as a step in that direction.

Are African American representatives in tune with black political opinion on issues, and

if so, on which issues is agreement found most often? Conversely, if there is a

disjunction, under what circumstances is it evidenced?

13 Toward a Theory of Black Representation

It is interesting to note that explicitly stated theories o f black representation are

few and far between (for exceptions see Whitby 1997 and Hill 1995). The theoretical

16

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. framework I use to address the question o f who represents black interests is captured in

the gender politics debate over thepolitics o f ideas, versus the politics o f presence

(Phillips 1995). The politics of ideas refers to good government politics where party

platforms and campaign promises are, in a sense, contractual agreements that are

faithfully executed once parties and individuals are elected to office. This theory

“suggests a broadly secular understanding o f politics as a matter o fjudgment and debate,

and expects political loyalties to develop around policies rather than around people”

(Phillips 1995,1).

The politics of ideas dictates that representation be viewed as a relationship

between constituents and party platforms. In other words, the message is more important

than the messenger and representatives’ social characteristics are not relevant. Here,

representation is essentially a race-neutral enterprise. Constituents’ interests are best

served when issues are well-defined and parties have taken distinct policy positions.

Because clearly defined issues, policies, and party positions are fundamental to the

politics of ideas argument, this theory is ill-equipped to account for representation

involving new issue advocacy and radical policy change. For the purposes of this

dissertation, the theoretical perspective embodied in Phillip’s politics o f ideas will be

referred to as either the politics o f parties or partisan representation, interchangeably.

This perspective takes into account the importance o f party in the representation o f black

interests at certain stages o f the legislative process.

The politics o f presence holds that, if democratic governments are deemed to be

valid, some degree o f descriptive representation must be evident. This theory calls into

question the validity of representative governments that do not reflect significant

17

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. characteristics o f the citizens they purport to represent. For example, the politics of

presence theory questions the validity o f an all-male legislature representing the interests

o f women. The issue is not whether men can represent women, but instead, whether a

representative body that appears to discourage the participation o f women can be said to

legitimately represent their interests.

Legitimate representation cannot occur under circumstances where entire groups

are excluded or only negligibly included. The very presence o f representatives from

previously excluded groups matters to constituents who share their social characteristics.

Their presence in the legislature raises the potential that new issues will be raised, new

policies advocated, and new alternatives considered (Mansbridge 1999, Williams 1998).

The point is not so much that every black person be represented by a black legislator, but

rather that black representatives be present in the legislature in the event that routine

legislative decisions become relevant to racial concerns, for instance, when racial ♦ interests and party interests diverge.5 The battle over welfare reform provides a good

example o f such an occasion. Despite President Clinton’s push for reform and support

from many white Democrats, nearly every Black Democrat, with one exception,

expressed opposition. Fundamentally, descriptive representation matters most in arenas

o f policy innovation where newly included representatives can transform the political

landscape by raising new policy issues or expressing distinct policy preferences.

5 The politics o f presence argument can be applied to other previously excluded groups, including other racial and ethnic groups, as well as women. The presence o f members from outside the status quo at the deliberation table enhances the likelihood, though by no means assures, that distinct perspectives from the status quo w ill be voiced. 18

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Shared group experiences among blacks may differentiate them behavior from

non-group members. Empirical evidence demonstrates,that blacks approach politics with

a strong sense o f racial group identity. A large majority of blacks believe that what

happens to black people as a group affects their own lives (see Gurir^ Hatchett, and

Jackson 1989; Dawson 1994, Tate 1994). Further evidence shows that electoral voting

continues to be racially polarized (Giles and Hertz 1994, Parker 1990, Williams 1990).

And, in low-information elections, candidate demographics affect voting behavior by

providing voters with stereotypical cues that help them decide how to vote (McDermott

1996). For instance, both black and white voters stereotype black candidates as being

more liberal than whites and more competent on and concerned about minority rights

(McDermott 1996, 12). Finally, in an analysis of the relationship between trust and

representation, constituent stereotyping of representatives’ behavior is a cornerstone

assumption (Bianco 1997). William Bianco writes that, “[cjonstituents form beliefs

about their representatives using stereotypes: they reason from a representatives observed

actions and attributes to form judgments about her unobserved policy goals” (Bianco

1997, 3). Stereotyping behavior is necessary because unless all potential policy issues

are explicitly debated during campaigns, representatives will invariably have to use their

own judgment to make decisions at some point. Citizens use race, among other

stereotypes, as a cue about how likely the representative is to share a common interest.

The scholarly debate about whether the party or the race of the representative is

most important in the representation o f black interests reflects the give and take between

the theories o f the politics of presence (descriptive representation) and the politics of

ideas (partisan representation). This is not to suggest the normative conclusion that one

19

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. theory leads to better representation than the other. The road to fairer representation for

blacks is likely found in the interaction between the two approaches and not in the false

opposition between one and the other (Phillips 1995,25).

1.4 Research Methods and Data

The data in this project come from several sources: voting and bill sponsorship

data from the Thomas online legislative databases and black opinion survey data from the

1996 National Black Election Study (NBES).6 The unit of analysis is House members

from districts with 15 percent or greater black populations. More than three-quarters o f

House districts have black populations smaller than 15 percent The sampling method

used ensures the inclusion of every black Democratic House representative and white

legislators with black populations ranging from15 to 55 percent in the 104th Congress.7

This data allows for an examination o f how a member’s race, party, and size o f the black

population in the district affect black representation.*

I collect and code data on bill sponsorship and roll-call votes for each House

member in the sample to examine the relationship between legislative activity and the

collective representation o f black Americans.9 Each legislator’s agenda is coded for the

6 The Thomas web address is http://thomas.Ioc.gotv. The 1996 National Black Election Study is archived at the ICPSR. Katherine Tate was the principal investigator. 7 Reps. J.C. Watt (R-OK) and Rep. Gary Franks (R-CT), the lone black Republicans in the 104th Congress, are excluded beeause their districts were less than 15 percent black. * See Appendix A and Appendix B for more detailed descriptions o fthe data. 9 Representation can be understood both as a dyadic relationship, he., the representation relationship between a representative and the district constituency and as a collective relationship, i.e., the representation relationship among members o f a legislature and members o fa group (Weisberg 1978, Powell 2001,13). For example, a Democratic constituent in a district represented by a Republican legislator is not represented in a 2 0

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. proportion of black interest bills sponsored, the proportion o f black interest bills with

blacks as direct or indirect beneficiaries, and the proportion o f black interest bills that are

symbolic or substantive. Each member’s roll-call votes on she issues are coded for

whether they match the plurality o f black opinion, as expressed in the NBES. In addition,

district descriptions and representative characteristics come from a variety o f sources,

includingThe Almanac o fAmerican Politics, Congressional Districts in the 1990s, and

Politics in America 1994 and 1996. Model specifications, variable operationalization,

and expected relationships are presented in the related empirical chapters.

1.5 Chapter Outline

My dissertation is comprised of six chapters. Chapter 2 explores factors that

motivate members to represent groups in their constituency, with specific focus on racial

group representation. Chapter 3 considers who represents black interests at the voting

stage of the legislative process. It begins with a brief discussion of policy congruence

theory and goes on to address the strengths and limitations o f policy congruence analysis.

The chapter concludes with an analysis of congressional roll-call data for empirical

evidence on who represents black interests. Chapter 4 examines who represents black

interests at the bill sponsorship stage o f the legislative process. It introduces my theory

o f symbolic representation in Congress. The final portion o f chapter four presents the

results o f an empirical analysis ofblack interest bill sponsorship data.

' dyadic relationship, but m ay have his policy preferences represented by other Democratic representatives in the legislature in a collective representationrelationship- 21

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chapter 5 presents an analysis o f constituent support for symbolic and substantive

bills based on survey responses from the 1996 National Black Election Study.

Specifically, it addresses how differences hr group identification, socioeconomic status,

and the type of representative affect blacks’ responsiveness to representatives’

substantive and symbolic policy proposals.

Finally, chapter 6 concludes the dissertation with a summary of the empirical

findings, a discussion of my theoretical contributions, and an assessment of the

implications o f this study for future research. I find that partisan representation must be

balanced with descriptive representation and race advocacy for black interests to be fully

represented. At the policy-making stage of the legislative process in the House of

Representatives, the party o f representatives is the most important factor in determining

whether objective and subjective black interests will be represented. However, at the

agenda-setting stage o f the process, the race o f the member is the most important factor in

determining whether black interests w ill be represented through the introduction o f black

interest bills, especially bills aimed at directly benefiting blacks and bills that give voice

to black concerns symbolically.

1.6 Conclusion

The theoretical and empirical analysis in this project contributes to the literature

on the nature of black representation, and congressional representation in general.

Although partisan representation is an important factor in the representation o f black

interests, descriptive representation is also important depending on the type o f legislative

activity being examined. This examination o f both roll-call votes and bill sponsorship

22

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. w ill add to our understanding o f how interests ace represented daring different stages o f

the policy process. The analysis o f black support for symbolic and substantive policy

proposals w ill contribute further to our understanding o f constituency-representative

linkages in representation. This dissertation demonstrates that a stronger link exists

between descriptive and substantive representation than previously found m roll-call vote

analysis.

I

23

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER 2

REPRESENTING RACIAL GROUP INTERESTS

This project tests the hypothesis that black Democratic legislators will be more

likely to support and advocate black interests than them white Democratic and Republican

colleagues. Thus, all things being equal, black representatives will be more likely to

sponsor and vote for black interest bills than other representatives. In the language o f the

theories o f presence and parties, the argument is that the politics o f presence, based on

shared experiences with groups that were previously excluded from the legislature,

enhances the representational relationship between elected officials and constituents by

giving representatives a self-interested motivation to move beyond responsiveness into

advocacy o f shared group interests.

Given what we know about the electoral connection between legislators and

constituents, under what circumstances would an elected official find it advantageous to

pursue diverse and sometimes divisive racial group interests? And, to the extent that all

elected officia ls are driven by sfnrrilar goals, why would w e expect a representative’s race

to influence his or her behavior?

2 4

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 2J Racial Group Representation and the Electoral Connection

Members o f Congress are purposeful actors who are responsive to their constituents

because this helps them achieve their own personal goals. Motivational theories of

congressional behavior establish three primary goals that drive representatives* actions.

Reelection is generally accepted as paramount (Mayhew 1974); however, the potentially

competing goals o f making good public policy and gaining political power are also

important (Fenno 1973). Thus, legislators may be motivated to represent black interests

when there is a clear connection between black constituents and representatives* electoral

fortunes. Additionally, legislators may also support black interests when black policy

preferences coincide with their own ideas about good public policy or when doing so

enhances their own political ambitions. In the remainder o f this chapter, we will examine

the extent to which these theoretical connections have been observed between legislators

and their black constituents. I will also consider the implications that theories o f group

consciousness and linked fate raise for racial group representation (Gurin, Hatchett, and

Jackson 1989, Dawson 1994, Tate 1994).

Because reelection is the proximate objective, representatives routinely consider the

potential electoral costs or benefits in the district prior to making decisions that are

traceable such as casting recorded votes, sponsoring and cosponsoring legislation, and

publicly pushing for or against policy proposals, hi the absence o f complete information

about constituents’ preferences on the thousands o f policies representatives consider each

session, they attempt to anticipate the reactions of various constituent groups.

Representatives behave in-ways that reduce the likelihood o f raising the ire o f attentive

publics, motivating otherwise inattentive and unfriendly publics, or giving challengers

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. information to use against them in the future (Arnold 1990, Kingdon 1981). The end

result then, is that although legislators are not unyieldingly yoked to the demands of their

constituents, the electoral connection operates to keep them at least minimally attentive to

the preferences o f politically relevant groups in the district. As Pitkin (1967) notes in her

treatise on the concept o f representation, representatives need not choose between being

delegates or trustees of their constituents, they must only behave in ways that

demonstrate a general consistency with their constituents’ expectations. This chapter

reviews the extant literature to assess the following questions: I) Are representatives

more responsive to blacks as their numbers in the district increase? 2) Does the electoral

connection m aintain its straightforward expectations for the representative-constituency

relationship when race becomes a factor in the political equation?

The motivational theories outlined above suggest that demographic factors like the

racial composition o f the district or the race o f the legislator should have little bearing on

the fundamental operation of the electoral link. Racial factors should rarely, if ever,

influence the behavior o f individual members. Frequent elections safeguard the will of

constituents, forcing representatives to either pay attention to constituent expectations or

risk being thrown out o f office. Therefore, the relative size (and not the racial makeup)

of various groups in the district population should be an important predictor of

responsiveness. As a given group increases its proportion o f the district’s voting-age

population, legislator responsiveness to the group should also increase. Before sketching

a theory o f racial group representation, Whitby (1997, 82) acknowledges that traditional

theories o f representation disregard the implications o f racial politics. He notes that,

26

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. [ijmplicit in the theoretical literature on representation is the notion that representatives should be judged in terms o f them legislative activities, not on die extent to which they mirror the social traits o f the people they formally represent From the perspective o f traditional theory, the race o f the member oughtn o affect t his or her policy-making behavior.

The expectation o f a linear positive electoral relationship based on relative size and

potential voting power can be confounded, however, when representatives’ perceptions o f

their constituencies are taken into account Representatives are particularly attentive to

constituents who comprise their personal and primary constituencies made up o f personal

friends and their strongest most loyal supporters (Fenno 1978, 18-27). They look to

members of these groups for cues on how to present themselves and their agendas to

voters and pay careful attention to what members o f these constituencies think o f them.

Representatives strategically weigh interactions with outside groups in relation to the

perceived electoral benefits and costs. What amount o f effort-from the representative is

required to win support from these other groups and will attentiveness to these other

groups threaten support from personal, primary, and key re-election supporters? When

representatives perceive that their interactions with or advocacy for certain groups w ill be

controversial or compromising, they weigh how important votes from the groups will be

in future elections. They further consider how little or how much time and effort they

need to expend to get the votes they want.

Fenno (1978) describes a representative who chose not to spend much time with

members o f his reelection constituency from a certain county despite the fact that he

routinely received a sizable portion o f their vote. The representative, Congressman A,

described the district as, “heterogeneous, disorganized, and full of factions...” before

going on to say, “... I can get 50 percent o f the vote without campaigning there at all; and

27

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1 couldn't get more than 75 percent if I campaigned there all the time” (Fenno 1978,68).

Congressman A knew that he conld expend very little or no political capital m the area

and stm get votes because, as long as he won the primary election, the constituents in his

party in Omega County had no other candidate to whom to turn.

Representatives make similar calculations o f time expenditures and political benefits

when racial politics in the district are polarized or contentious. When racial groups in

the district express diverging or competing preferences or when legislators believe racial

divisions exist in the electorate, they may “choose sides.” In such cases, the preferences

of a sizable racial minority may be disregarded in favor of a small, united and

oppositional majority. Two anecdotal accounts highlight this aspect o f representatives’

electoral strategies.

In the first account, a white representative from the rural South accepts support from

black voters, who as a group are not part o f his primary constituency, but allocates little

time to courting their support. He notes that, “‘[tjhe black people who know me know

that I will help them with their problems,’” but he does not put much effort into courting

or increasing his share of black votes (Fenno 1978,68). In the second account, a black

legislator, Congressman F, from a majority-black urban northern district with an

approximately one-third white population discounts white support when he thinks

strategically about his electoral fortunes. As Fenno (1978,115) describes it, “[h]e thinks

o f his reelection constituency as the ‘whole black community.’ He will take whatever

white support comes his way. And he knows some will. But he does not count on any

when he calculates.”

28

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. In these two examples, the race o fthe representatives, the placement o f various racial

groups in their perceptions o f important constituencies, and the nature o f racial politics in

these two districts affected their electoral calculus. Rather than actively courting and

advocating for sizable racial group minorities in the district, these representatives

calculated the likelihood o f getting the votes without committing the political resources.

These two accounts suggest that the electoral connection may not be as straightforward

when racial politics is considered, particularly when a single party dominates the district,

racial politics are contentious, and representatives are not dependent on votes from

outside group members to win the primary election.

Empirical research on the relationship between the racial composition o f the district

and legislators’ responsiveness is plentiful. Scholars have examined the impact of the

percentage of blacks in the district on three key measures of black interest

responsiveness: 1) the likelihood of electing a liberal over a conservative, 2) the

likelihood of electing a Democrat over a Republican, and 3) the likelihood of electing a

black representative over a white representative. In the following sections, I w ill review

extant research on the each of these measures of responsiveness before explicating the

potentially confounding impact o f race.

22 The Electoral Connection and Ideological Congruence

Legislators’ liberal voting behavior is an important measure o f responsiveness to

black interests for a number of reasons. First, blacks are extremely supportive of both

liberal policies and liberal candidates. Robert Singh, in an analysis o f the Congressional

Black Caucus, characterizes blacks as “an exceptional constituency” that “ .. tenaciously

29

j

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. adhere[s] to convictions about the role o f government that the vast majority o f whites

simply do not (or no longer) share” (1998,11-14). Empirical evidence demonstrates that

blacks consistently support liberal governmental responses to the problems of

unemployment, poor-quality schools, lagging health care access, and urban decay (Lublin

1997, Tate 1994, Dawson 1994). Scholarship on racial group policy preferences also

indicates a “racial divide” in public opinion: while black support o f liberal policies and

candidates has persisted overtime and in spite o f increasing economic polarization, white

support has lessened (Kinder and Winter 2001, Canon 1999, Singh 1998, Tate 1994).

Finally, the black policy agenda is an extension o f the Democratic Party liberal

agenda which emphasizes economic protections for the working class, civil rights

protections for women and racial/ethnic minorities, and safe-keeping for the nation’s

most vulnerable. Thus, the liberal agenda embodies elements o f the black agenda that

stress universal appeals for redistributive social programs to provide a safety net

America’s less fortunate citizens.10

That being said, we must note that the liberal agenda does not fully encompass the

radical racial agenda promoted by advocates of black interests. As reported by Singh

(1998), Sniderman and his colleagues (1993) contend that the early civil rights era racial

agenda o f equal opportunity achieved through the pursuit o f non-discriminatory policies

has given way to a “race-conscious” agenda that demands that race be taken into account

in the distribution o f material benefits in education, employment, government contracts,

and electoral arrangements. Arguing in support o fa race-conscious transformative black

10 See Cohen (1999) for discussion of the limitations o f the black policy agenda, specifically as it relates to cross-cutting issues. Also, see Smith (1996) for a critique of the liberal agenda’s promotion ofblack concerns. 30

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. agenda, Smith (1996,210) rejects deracialized strategies for achieving black policy goals

in favor o f a “race-specific strategy that goes hunting where the ducks are.” According to

this view, problems that manifest themselves in race-specific ways (like unemployment

for blacks in the double-digits) demand race-specific solutions- However, despite its

limitations, the liberal agenda is central to the realization ofblack policy priorities. Thus,

liberal voting and the initiation o f liberal policy proposals in Congress are important

indicators o f responsiveness to blacks’ policy needs.

Political scientists have studied the relationship between the size of the district’s

black population and liberal voting for several decades. V.O. Key’s (1949) richly

descriptive analysis o f the behavior o f white southerners prior to the passage o f the 1965

Voting Rights Act reveals one of the possible relationships between the racial

composition o f the district and legislator voting behavior. Before southern blacks

reclaimed legal access to the ballot, representatives from districts with the largest black

populations were more likely than others to vote conservatively. Whites who resided in

“black-belt” districts most often elected conservative candidates who vowed to maintain

white political dominance.

Following the legislative triumphs o f the civil rights movement, scholars began to

examine the prevalence o f this “white backlash theory” in the new political context of

black enfranchisement Using a variety o f barometers, e.g., votes on civil rights bills or

conservative coalition scores, scholars found that prior to the 1980s, southern legislators’

responsiveness to blacks did not exhibit the positive linear pattern that electoral

connection theories would predict (Black 1978, Bullock 1981, Combs, e t al. 1984). The

initially negative relationship documented by Key (1949) and later Black (1978) prior to

31

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. the passage o f the 1965 Voting Rights Act, became curvilinear in the decade immediately

following passage (Comb, et al. 1984, Bollock 1981). Although the positive linear

relationship predicted by the electoral connection between responsiveness and the

district’s black percentage was evidenced in the voting behavior o f non-southern

representatives, support of black civil rights and liberal voting by southern

representatives was limited to representatives from districts with relatively small black

populations or those from districts with black majority populations.

By the mid-1980s, the significance of the racial composition of the district was

declining as the voting behavior of southern Democrats became more like the behavior o f

their non-southem partisans (Whitby 1985, Whitby 1987, Whitby and Gilliam 1991).

Scholars also found a positive effect from the interaction of race and urbanization on

legislators’ liberal voting behavior (Combs, et. al. 1984, Whitby 1985, Whitby 1987,

Fleisher 1993). There is a general consensus, however, that black political participation

has contributed to the moderation o f white southerners on racial policies (Whitby 1997,

87-90). This liberalizing effect more intense among non-southem representatives than

among southerners. Lublin (1997) and Cameron, Epstein and O’Halloran (1996) find

that larger black populations are required in southern districts to attain the same liberal

impact as in non-southem districts.

The most recent work on the liberalizing effects o f black population size finds mixed

results. In a pooled time-series analysis o f Poole-Rosenthal Nominate scores from 1972-

1992, Lublin (1997) finds that the percent black in the district has a liberalizing effect on

legislators’ voting behavior, especially when blacks constitute 40 percent or more o f the

districts’ population. He concludes that although, “white backlash may still make

52

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. representatives less responsive to African Americans than might otherwise be the case,...

it does not prevent blacks from exerting influence over their representatives” (Lublin

1997,88). At no point during the timeline o f his study does Lublin find that increasing

the percentage of blacks in the district increases conservative voting behavior among

legislators. Likewise, Cameron and his colleagues (1996) in a cross-sectional analysis of

LCCR scores in the 103rd Congress find that increasing the percent of blacks in the

district does not result in white backlash; however, they find that the racial composition

of the district has little measurable effect on increasing liberal voting behavior among

legislators.

23 The Electoral Connection and Partisan Congruence

The percent o f blacks in the district is positively related to the probability o f electing

Democrats to office (Cameron, et. al. 1996, Lublin 1997). This is not particularly

surprising given that the majority o f blacks identify themselves as Democrats and that

blacks are the most loyal constituency for Democratic Party candidates, e.g., Frymer

1999, LubHn 1997, and Tate 1994.

Empirical studies find that Democratic representatives are more responsive to black

interests than Republicans. Using several measures of liberal and race-specific policy

support, Swain (1994) found that Democratic representatives in the 100th Congress were

considerably more supportive o f black interests than Republicans. Likewise, both pooled

thne-senes (Lublin 1997) and longitudinal cross-sectional analyses (Whitby 1997) of

partisan differences in liberal voting behavior find that Democrats outpace Republicans

in black responsiveness using Poole-Rosenthal scores and LCCR scores, respectively.

33

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Ia concert with my findings for the 104th Congress discussed in Chapter 3, Canon

(1999,177) finds that the mean LCCR scores for black and white Democrats were more

than 73 and 56 percentage points higher, respectively, than the mean score for white

Republicans. In a refutation of white backlash theories, Lublin (1997) finds that the

percent black in the district was not positively delated to electing Republican

representatives between 1972 and 1992. Although the exact percentage of blacks

required in a district to elect a Democratic representative varies by region, extant

scholarship indicates that a minimum black population of 40 percent is necessary

(Lublin 1997) with somewhat higher percentages, between 45 and 47 percent, in the

South (Cameron, e t al. 1996).

The electoral connection based on shared party identity is challenged, however, when

race becomes a factor. Some scholars have suggested that black electoral support o f and

leadership roles in the Democratic Party may doom the Party to minority status in

Congress for the foreseeable fixture. This “new orthodoxy” in the Democratic Party

blames black loyalty (rather than white racism) for the flight of conservative and working

class whites from the Democratic Party rolls." Proponents of this approach urge

Democratic Party leaders and candidates to distance themselves from “black” issues if

they want to regain political power (Frymer 1999, Smith 1996). The affect of this

ambivalence toward blacks at the national party level on our expectations for the district-

level electoral connection suggests that white Democrats may be less willing to

11 See Smith (1996,259), Chapter 10 for a thorough discussion o fthis “’new orthodoxy” that “blame[sj black political demands for the rightward turn in American politics and for the left’s collapse.” 34

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. vigorously pursue the issues that concern blacks in their districts lest they risk alienating

white voters.

The apparent breakdown o f the electoral connection is further heightened by the

unparalleled loyalty o f blacks to the Democratic Party. In recent years, black loyalty has

been only reluctantly rewarded by the Democratic Party as it has retreated toward the

center and abandoned its left-leaning liberal agenda in favor o f centrist politics. Because

blacks are essentially “captured’' by the Democratic Party, with no viable alternative

political party, Democratic representatives can benefit from their support while allocating

little time or resources to courting or maintaining it (Frymer 1999).

2.4 The Electoral Connection and Descriptive Representation

Electing black legislators is also viewed as important for black representation, but

disagreement exists among scholars over the associated costs to black substantive

representation. The reasons for increasing black descriptive representation are many.

Some suggest that the mere goal o f blacks’ having role models in Congress is a sufficient

reason for encouraging descriptive representation. On the grounds o f democratic theory

and fair representation, others find that giving blacks the opportunity to elect a

representative o f their choice means designing districts where black candidates at least

have a fighting chance.12

a This argument goes both ways. See Cain (1992) “Democratic Theory” in Controversies in Minority Voting Rights. See*aIsOi Guinier (1994) for an assessment o f why the race o f the candidate is not necessarily the best indication of the exercise “choice.” 35

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Black Democrats are the most consistently liberal representatives in the House. This

suggests that they are also most consistently m accord with the liberal agenda that resides

centrally in the black political agenda (Swain 1994, Lublin 1997, Whitby 1997, Canon

1999). In addition, Whitby (1997) finds that, compared to other legislators, black

representatives consistently vote to strengthen and support civil and voting rights

legislation.

As evidence of their sensitivity to racial concerns, Canon (1999) finds that black

representatives address the preferences of both white and black constituents in a

“balancing approach.” Although they do not focus exclusively on black concerns, black

representatives allocate substantially more resources to black issues through voting, bill

sponsorship and cosponsorship, floor speaking, and staff allocations than their white

counterparts. Hall (1995) further finds that African American and Latino representatives

more intensely advocate issues that concern minorities during floor debate than their

colleagues. This suggests that black representatives may be the most likely source of

transformative politics on behalf o fblack interests.

Cathy Cohen (1999, 254) defines transformative leadership as “leadership which

seeks fundamental change by appealing to a greater moral purpose, transforming both

group members and political targets.” Transformative leadership w ill not likely emerge

in the context of moderation and marginalization that characterizes the electoral

connection between black constituents and centrist white Democrats. Thus, while white

Democrats may “vote the right way,” they may not translate that responsiveness into

advocacy o f black concems.

36

"'■* - *!■ •’ ^ Reproduced withpermission of the copyright owner. Further reproduction prohibited without permission. Finally, although many arguments o f enhanced black representation because o f the

presence o f black legislators rest on theories o f shared group consciousness and linked

fate, the electoral connection may simply be stronger for black legislators.13 Black

representatives must rely more heavily on black votes than other representatives to win

reelection. The central importance o f black constituents in their primary and reelection

constituencies contributes to black legislators7 high degree of responsiveness on nearly

all measures.

2.5. Conclusion

hi this chapter, I have considered the impact of race on the expectations for

representative-constituency relations based on the electoral connection. While a standard

interpretation o f the electoral connection would dismiss the influence of the racial

composition o f the district or the race o f the legislator on the representative-constituency

relationship, I have demonstrated that racial factors do indeed reshape the relationship.

The liberal agenda, while instrumental for black interests, falls far short of coming to

terms with the pressing needs of black life. And the liberalizing effect of blacks’

proportion o f the district population on the electoral connection has become less effectual

over time. The connection between black populations and the election of Democratic

candidates is strong, but cross-pressures within the Democratic Party may prevent white

Democrats from rewarding black voter loyalty with race-specific policies because o f the

“ Kenny Whitby (1997,82-86) sketches a theory o f black representation that accounts for why the presence o f black representatives enhances black representation that draws on literature about group consciousness, the social psychology o frace, and House members’ political perceptions of their constituencies. There is very little empirical work on this subject (but see Hutchings, McCIerkmg, and Guy-Uriel 2001). 37

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. risk o f alienating white voters who may exit to the Republican Party. Black legislators,

because o f a mix o f electoral factors and shared group interests, are the most consistently

responsive to black interests, regardless o f the district’s racial composition.

This discussion has not resolved the conflict between the politics o f presence and

the politics o fparties. Instead it has presented us with a paradox similar to the one David

Lublin highlights in his discussion o f the trade-off between descriptive and substantive

representation. Lublin's “paradox o f representation” notes that the costs associated with

creating districts with large enough black populations to elect black candidates in the

South are I) the draining o f black populations from surrounding districts and, 2) the

subsequent creation o f more conservative and more likely, Republican, districts.14 Along

with the creation of more Republican districts comes the increased likelihood of a

Republican-dominated legislature and the subsequent enactment o f fewer liberal policies.

Thus, blacks must accept a trade-off between their goals to achieve greater descriptive

representation (more blacks in elected office) and their goals to attain greater substantive

representation (enacting liberal policies). This chapter demonstrates that blacks must

accept a trade-off between racial group advocacy and substantive representation from

white Democrats as well.

What I term a “paradox of advocacy” considers representation as the active

pursuit of policies that serve to transform black social, economic, and political

14 In non-southem districts, the trade-off between achieving black descriptive and substantive representation is not so stark, hi northern urban districts, the proximity o f traditionally liberal communities, e.g^ large Jewish communities in addition to blacks, means that reducing blackpopulations in adjoining districts to create a majority-black district w ill not necessarily result in the adjoining districts becoming more conservative. Thus, black descriptive representation can be increased without jeopardizing black substantive representation in non-southem districts. 38

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. conditions, rather th an only enacting liberal policies. Libera! policies are necessary, but

not sufficient, to attain the goals o f social and economic justice that transformative black

leaders seek. The electoral connection promotes only the most minimal representation o f

black interests. Thus the liberal voting behavior that we commonly use as a measure o f

“responsiveness” to black interests, only serves blacks policy goals if blacks’ moderate

their preferences in favor o f a less demanding, less radical agenda, leaving many o f the

problems that desperately need to be addressed unresolved. The perverse conclusion

appears to be that blacks can gain, maintain, and increase their substantive representation

(liberal Democrats in power in Congress) only by relinquishing their fundamental goals

for political change.

In the chapters that follow, I explore this “paradox of advocacy” through an

analysis o f black representation through voting and bill sponsorship, taking into account

the race and party o f the member, and the racial composition of the district. Chapter 3

considers the representation o f black objective and subjective interests through roll-call

voting. Chapter 4 specifically examines the implications o f symbolic representation for

the advocacy of black interests. Chapter 5 then considers which blacks are most

responsive to symbolic policies. Finally, chapter 6 summarizes the results and presents

some concluding remarks.

39

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER 3

BLACK REPRESENTATION THROUGH ROLL-CALL VOTES

In the previous chapter, I addressed why and how elected representatives might be

motivated to engage in the representation o f a racial group, i argued that although the

electoral connection is often expected to operate in a race-neutral fashion, divisive racial I i politics could compromise the expectation o f candidate responsiveness based on group

size and voting power. When certain groups are “captured” by a political party or

candidate, or when candidates are not dependent on members o f the group for votes, the

electoral connection is compromised and group interests may be ignored.

In this chapter, I consider the effect o f representat^ves, race and party, and the size

of the district*s black population on roll-call voting behavior in favor o f objective and j subjective black interests. Recall from chapter 1 that objective black interests are defined

i by observable social, economic, and political circumstances that demonstrate a ! disproportionately negative impact on the black community. Assessments of objective j black interests are based on sources of information that are external to the black

| community like census data indicating racial disparities in poverty rates, housing

accessibility, employment, and educational quality. Alternatively, subjective black

interests are measured by assessments blacks make of their own preferences in public 40 . f

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. opinion surveys. Thus, an important distinction between objective and subjective black

interests is that measures o f the former are based on sources that are external to die black

community, while measures o fthe latter are based on internal sources.

The work o f both Canon (1999) and Swain (1994) points to the importance o f

subjective black interests in black representation, but to date, no one has presented a

systematic empirical analysis of the concept A primary reason for this is the lack of

sufficient survey data to gauge black interests. A battery of questions asked in the 1996

National Black Election Study on black policy preferences provides a good starting point

for such efforts. Because of the potential disconnect between the objective needs of a

group and their desired preferences, it is important that researchers further examine the

representation of subjective interests in addition to continued research on the

representation of objective interests. Understanding where black interests diverge

between needs and wants and who represents each will allow researchers to consider

black representation from external (need-based) and internal (preference-based) vantage

points.

Existing research shows that support for objective black interests may vary

according to representatives’ partisan affiliation and race, the size o f the district’s black

population, the size o f the district’s urban population, and region. Other variables,

including how long a member has served in office, may also influence the likelihood to

support black objective interests. How do these same variables affect representatives’

support for subjective black interests? Is that fact that African American legislators

often vote more liberally than African American policy preferences an impediment to the

41

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. collective representation o f black interests? The goal o f this chapter is to address the

above questions and in the process generate dialogue about the nature of black

representation particularly with regard to the representation o f black interests when

blacks selfdefine them.

3J Analyzing Black Substantive Representation

Building on a rich tradition established by Miller and Stokes’ focus on

constituency attitudes (1963) and Verba and Nie’s policy focus (1972), scholars have

traditionally used policy congruence analysis to assess the fit between elected officials’

roll-call votes on policy issues and the needs o f the African American community. In the

absence o f alternative data sources, evidence o f a liberal ideology has routinely been used

to indicate the likelihood o f a representative to vote in favor o f objective black interests.

Empiricai studies of objective black representation are based on aggregate roll-call

analysis which regress ideological group indices like LCCR (Leadership Conference on

Civil Rights), or COPE (Committee on Political Education) scores on district

characteristics. The resulting model is then interpreted as a model o f black representation

when it is actually a model of representatives’ ideology (see Lublin 1997). In effect,

representatives’ ideology -when liberal- is interpreted as a measure o f black interests so

that a member with a high LCCR score is interpreted as being highly supportive of

objective black interests.

A primary reason for the use o f these summary measures is that interest group

scores are easily accessible in one place while sufficient survey data on black opinion

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. usually is not. Liberal members may reasonably be expected to more often represent

objective black interests because, at the heart o f black interests, there is a core liberal

agenda supporting government intervention in remedying racial disparities in housing,

education, economics, heath care access, criminal justice, etc. Nationally articulated

during the black Civil Rights Movement, this core liberal agenda has remained

remarkably consistent over time and across socioeconomic groups in the black

community.

Policy congruence analysis o f black representation focuses on two factors: the

district*s racial composition and the race of the representative. On the side of

constituency influence, scholars have investigated whether changes in the district’s black

population have affected representatives’ voting behavior.

Research following the implementation o f the 1965 Voting Rights Act examined

how changes in districts’ black populations impacted on white representatives’ voting

behavior (Whitby 1997, Whitby and Gilliam 1991, Whitby 1987, Bullock 1985, 1981,

Combs, et. al. 1984). Scholars have found that over time, increases in black voter

participation have resulted in more liberal voting patterns for white legislators. Some of

this moderation is due in part to the interaction between urbanism and black populations.

When urbanism increased, so did black responsiveness (Whitby 1985. Combs, et. aL

1984).15 The decline in regional ideological differences between southern and non­

“ Fleisher (1993) argues that the percent urban is actually accounting for the increased white liberalism that can be expected in cities. Whitby (1987) argues that the percent urban is actually a proxy measure for the degree o f organization and resources in black urban communities. 43

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. southern Democrats has also accounted for some o f the moderation in support o f black

interests (Whitby 1997, Lublin 1997, Bullock 1995).

The interplay between race and responsiveness became an increasingly important

research question as the numbers o f black representatives at all levels o f government

increased. The question shifted from, “how does district composition impact substantive

responsiveness?,' to include “how does the race o f the representative impact substantive

responsiveness?,' and the unit of analysis shifted from percent black in the district to

include the race o f the legislator.

The number o f black representatives in the House grew dramatically during the

last decade. Sixteen new black representatives, 13 from the South, were elected to the

House in 1992, following the first round o f redistricting after the 1990 Census. There

were a total of 25 black House Representatives prior to the convening of the 103rd

Congress (1993-1994). The 103rd Congress saw that number increase to 38 black

representatives. With this remarkable increase in the number o f black representatives,

their numbers increased to 8.7 percent o f the House, but still fell short of proportionally

representing the U.S. black population at 12.1 percent. Increased numbers indicated

increased black descriptive representation, but did not necessarily indicate a

corresponding increase in black substantive representation.

Carol Swain (1993), for instance, found that white representatives from

predominantly black districts were just as likely as black representatives to support

objective black interests. This finding has been highly contested, generating a

44

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. groundswell o f research on the topic o f whal black representation is and who represents

black interests, e.g^, Tate 2002, Canon 1999, Lublin 1997, and Whitby 1997.

3J.1 Policy Congruence Analysis of Black Interests

With regard to objective black interests in particular, there are two

methodological limitations that make policy congruence analyses based on interest group

indices less than perfect indicators of black representation. The first is the censored

sample problem (King 1989). Roll-call votes are the end result o f a long and complex

process weeding out process in House committees. As Canon (1996, 8) notes, “only the

middle-range of the possible continuum of racial concerns ever make it to a roll-call

vote... Thus, statistical analyses based on this censored sample yield biased parameter

estimates.* The fact that the potential sample o f policies is already truncated by the time

it reaches the roll-call vote stage is an extremely important consideration for black

representation scholars since black interests (and other minority interests), by their very

nature, are already marginalized.

The second problem relates to the use o f summary interest group scores. The use

of liberal interest group scores is based on an assumption that black representation is

analogous to liberal representation. That is, scholars have used evidence of

representatives' liberal voting records, e.g., high Americans for Democratic Action

(ADA) or Leadership Conference on Civil Rights (LCCR) scores, as evidence o f higher

responsiveness to black interests since blacks are extremely liberal in their policy

preferences and voting behavior. However, survey data points out the fallacy of

assuming liberal attitudes for blacks across a range o f issues. For instance, blacks are not

45

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. as liberal as whites with regard to homosexuality and abortion issues (Tate 1994). Black

public opinion is not uniformly liberal and is often at odds with the liberal agenda

regarding women's rights, gay rights, abortion, and school prayer (Narcisse 1997). Black

and white liberal representatives can hardly be said to represent black interests on issues

where they so squarely oppose one another.

Additionally, many of the issues included in these scorecards are of only

tangential interest to blacks and are sometimes in direct conflict with the majority blacks.

For instance, o f the 14 issues selected by the LCCR for the 103rd Congress, Canon

(1996,8-9) points out that only two related directly to black interests (D.C. Statehood and

an amendment to delete the Racial Justice Act), four related tangentially to blacks (two

motor voter bill votes, the Brady handgun control bill, and an education bill), and the

remaining six were not specifically relevant to the black community. Similarly, I find

that only one o f the 20 issues in the LCCR report for the 104th Congress related directly

to black interests (affirmative action for minority broadcasters), and four were related

indirectly (welfare reform, drug sentencing disparities, banning assault weapons, and

increasing the minimum wage).16

While recognizing some o f the drawbacks o f using roll-call votes to examine the

representation o f black interests, researchers have nevertheless rightly pointed out the

utility and general reliability of such measures (Swain 1993,13-16 and Whitby 1997,9-

11, but see also Canon 1999).17 Researchers have overcome some of the limitations of

16 See Appendix E for list o f issues in the LCCR report for the 104th Congress. l7In a rather forceful statement, Canon (1999, 31) concludes that the ideological indices like ADA and LCCR scores are flawed measures of black representation and are 46

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. traditional roll-call analysis by using votes on key issues objectively determined to be in

the interest o fblacks as dependent variables, e.g., civil rights votes where blacks are clear

beneficiaries. While this approach is certainly legitimate, it requires researchers to

determine the racial nature of roll-calls. An alternative to this method is to simply use

blacks’ opinions on various policy issues found in surveys and polls to guide the choice

of which roll-call votes to examine. Doing so overcomes concerns about liberal

assumptions and also removes the subjectivity of researchers selecting which issues

really reflect black interests.1* It is to this aim that we turn in the following discussion of

subjective black interests.

3.1.2 The Black Opinion Index

To examine subjective black interests, I created a Black Opinion Index (BOI) that

measures the congruence between blackopinion on issues found the 1996 in NBES and

roll-call votes taken during the 104th Congress (1995-96). This measure adds additional

depth to the picture o f black representation painted by the analysis o f objective black

interests by I) considering both race-specific votes such as those on affirmative action,

and race-neutral votes such as those on defense-spending; and, 2) by taking into account

black Americans' social conservatism.

“worthless as indicators o f behavior that address constituents’ racial wants." ■*One caveat is that use o f survey data requires researchers to assume that black opinion is constant across districts although it is possible that opinion on any given issue may vary due to regional or even district-specific concerns. The current state o f survey data research makes interviewing sufficient numbers of blacks in each district to draw inferences about district opinion too costly an endeavor and thus remains a question for future research- 47

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Evidence o f social conservatism in black: attitudes is not a new phenomenon and

neither is the observation that the issue positions o f black leadership are sometimes

contrary to the policy preferences expressed by black Americans (Nardsse 1997, Tate

1994, Swam 1993,12-13). For instance, black representatives in Congress seem out of

step with black public opinion regarding homosexuality. Although 90 percent o f blacks

believe that homosexuality is “always wrong” (Tate 1994), during the 104tfa Congress,

two-thirds o f black Representatives voted against a bill that proposed a federal definition

o f marriage that excluded same-sex unions.

Data from the 1996 NBES show that on the issue o f welfare reform, nearly 70%

o f African Americans are supportive o f a five-year time limit on access to benefits for

poor women. A majority o f those with a high school diploma or less (56.6%), as well as

a majority of those who have enrolled in or completed advanced degree programs

(68.6%) support the time limit. African American leadership in Congress, on the other

hand, opposed welfare reform in House roll-call votes. In fect, 74 percent of white

Democrats (28 out o f 38) in the 1996 NBES sample, and 97 percent o f black Democrats

(32 out o f 33) were opposed to welfare reform. Representative Sanford Bishop (D-GA)

was the lone welfare reform supporter among black Democrats. Additionally, although

54 percent o fthe black public supports the death penalty in murder cases, black leaders in

Congress as well as those from major national black organizations like the National

Urban League and the National Association for the Advancement of Colored People

(NAACP) are uniformly opposed to easing restrictions that slow down death penalty

enforcement.

48

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Survey data does not suggest that a majority of blacks ate supportive of

conservative policies, however. Public opinion data consistently show that blacks

support more, not less, federal action in civil rights advocacy, programs that assist the

poor, and initiatives that improve education and healthcare opportunities for blacks

(Gurin, Hatchett, and Jackson 1989, Tate 1994, Dawson 1994). In fact, evidence of a

black-white racial divide on conservative policy issues persists despite gains made by the

black middle class in income and education achievement- Across the board, blacks

express more liberal policy preferences when compared to whites on a range of policy

issues, especially those regarding federal assistance for minorities and job creation

programs (Tate 1994). However, on issues like federal spending on crime and national

defense, the racial divide lessens (Dawson 1994, Tate 1994, Canon 1999).

The BOI measures congruence between black opinion and representatives’ votes

on six issues: increasing spending on crime, support for the death penalty, support for

affirmative action, decreasing defense spending, support for welfare reform, and limiting

illegal immigration. Opinions on issues from the NBES were used to determine the

selection of roll-call votes cast during the 104tth Congress. Using this method, I did not

have to assume liberal dispositions among blacks across issues since congruence between

actual black responses (not liberal voting indices) and legislators’ votes create the

dependent variable. I also did not need to determine whether a roll-call vote reflected a

black interest issue because the BOI is concerned with black preferences for public

policies and not with the racial content o f the policies.

4 9

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1 2 D ata and M ethods

In this analysis, I examine the roll-call voting behavior o f House members during

the 104th Congress (1995-96) whose districts had populations that were at least 15 percent

black. After excluding three black Democrats who did not complete their terms, and the

lone Latino legislator (to simplify the race variable), the sample consists of 96 House

representatives.19 O f this number, 69 percent (n=66) are white representatives and 31

percent (n=33) are black. Nearly 75 percent o f the sample are Democrats, while the

remaining are Republicans. Dummy variables for the race (black=l, else 0) and party

(Democrat=l, else 0) o f the member are included in the analysis.20 The number o f terms

served by each legislator is included in the models to control for the influence o f seniority

on responsiveness. Representatives who have served more terms in the House may have

garnered sufficient loyalty from constituents to depart from strictly acting as delegates in

Washington (Bianco 1994, Fexmo 1978).

The percentage o f black constituents in the district is included in the model as a

measure o f constituency influence.21 Chapter 2 demonstrates the mixed influence o f the

19 See Appendix B . 20 Using an interaction term to account for the relationship between race and party that is likely affecting liberal voting behavior would be more theoretically satisfying. However, this model is directly comparable to most other models ofblack representation in the field. Additionally, because there are no black Republicans m the sample, the two dichotomous variables for race and party account for two o fthe available three degrees o f freedom and thus, serve essentially the same purpose as interaction terms would. 21 Because o fthe high muiticollinearity between the race o fthe member and the district's black population, many researchers do not include both variables in the same multivariate model. Including both variables biases the results toward accepting the null hypothesis that no relationship exists when rt should be rejected. When I ran the models without the percent black in the district, the strength and direction o fthe coefficients remained stable. 50

...... -.gPOtjtf S-,, «- Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. racial composition o f the district on legislator behavior. Greater levels o f black: political

participation since the 1960s have changed the political landscape in the House in a *> ■ * number of ways. Increasing percentages o f blacks in House districts have had a

liberalizing effect on representatives* voting behavior, particularly in the South (Lublin

1997). Once the black percent o f the district reaches 45 percent or more, the likelihood

of a district electing a Democratic or black representative increases dramatically

(Cameron, Epstein, and O’Halloran 1996, but see Lublin 1997). Variables controlling for

percentage o f the district that is urban and regional differences are included in the model

as well. A dummy variable for region is coded as “1” for southern districts, “0”

otherwise. Southern white representatives are not as responsive to blacks as non-

southerners (Lublin 1997) and urbanization is thought to have a liberalizing effect on

representatives’ voting behavior (Whitby 1987).

I compare and contrast multivariate analysis results of the BOI measure of

subjective black interests (defined by black public opinion) with House members’

LCCR22 scores, a conventional measure of objective black interests (defined by

observable phenomena like poverty statistics). Data on LCCR scores were collected

from the Civil Rights Voting Record for the10fh Congress annually produced by the

Leadership Conference on Civil Rights, and data on roll-call votes for the BOI were

collected from various editions o f the Congressional Quarterly Almanac, 1996. Both o f

To maintain the theoretical integrity o f the models, I include both variables in the results presented here. 22 The Leadership Conference on Civil Rights (LCCR) is a coalition of more than 180 national organizations who advocate the interests o f African Americans and other racial minorities, senior citizens, women, and the poor. They issue a biennial report card ranking members o f Congress on them support o f issues o fconcern to the organization. 51

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. the dependent variables ace bound by 0 and 100, where 0 indicates that the

Representative never voted in accord with black interests and 100 indicates a perfect

record of voting support for black interests, hi addition, district descriptions and

representative characteristics were compiled from a variety o f sources including the

Almanac o f American Politics* Congressional Districts in the 1990s, and Politics in

America 1996.

Voting in support o f objective black interests is modeled using Ordinary Least

Squares Regression. Because of limitations in the data, the BOI is more appropriately

modeled using Ordered Probit Regression.23 Voting scores are a function o f two sets of

factors: Representative characteristics (race, party, and seniority) and district

characteristics (percent black population, region, and percent urban).24

33 Results and Discussion

Table 3.1 presents the estimated parameters and summary statistics for the

regression model outlined above. Parameter estimates that are statistically significantly

° The BOI is a percent, not a points, score and manifested only six real values (0,16.67, 3333,50,66.67, and 8333 percent). For that reason, an ordered probit model may be more appropriate. 24 Note that this model includes both the race o f the legislator and percent black in the district Many scholars do not include both variables due to the high multicollinearity between the two (see Swain 1993, Lublin 1997, and Whitby 1997). When high multicollinearity exists standard errors are likely to be unreliable and the risk increases that the null hypothesis will be accepted when there is actually a statistically significant relationship between the dependent and independent variables (see Whitby 1997). However, Berry and Feldman (1985, 46-51) argue that deleting variables that are indicators of unique theoretical concepts (like the role of representatives' race in representation) due to multicollinearity could result in model misspecification and biased coefficient estimators, a more serious problem- 52

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. different from zero provide evidence that the independent variable is associated with

change in the voting score, while controlling for the influence o f other variables. Results

from the QLS regression analysis o f LCCR and Ordered Probit analysis o f BOI scores for

the 104th Congress reveal differences in the way that race influences roll-call voting

behavior on objective versus subjective black interests. Race is a statistically significant

predictor o f congressional responsiveness in both models, though its effect is negative in

the subjective black interests model and positive in the objective interests model. African

American representatives are less likely to represent subjective black interests across

issues than their white counterparts. Liberal voting indices overestimate responsiveness

to black interests by dismissing the growing conservative attitudes among blacks on

certain issues such as those included in the BOI and thus do not account for the apparent

disconnect between black leadership and black opinion on some policy positions.25

INSERT TABLES 3.1 AND 3 3 ABOUT HERE

Consistent with findings by other scholars, the race, party, and seniority26 o f the

representative along with the region and percent urban o f the district are all significant

predictors o f objective black interests as measured by liberal LCCR scores (Swainl993,

25 The BOI is also somewhat sensitive to the way issues are defined during a given Congress. Because it matches black opinion on issues to roll-call votes in the same issue area, it is affected by the liberal or conservative approach Congress takes to addressing the issue during that term. For instance, according to the 1996 NBES, nearly 70% o f blacks want more spending on crime. During the 104th Congress, more spending on crime meant more spending on prisons and police as opposed to more spending on prevention measures like recreation centers and midnight basketball programs. Legislators vote on specific proposals, whereas respondents answer more general questions. Thus, the fit between votes and opinion is imperfect, but manageable. 26 Seniority o f the member, in this case measured as number o f years in office, is often not included in models predicting support for black interests. These results suggest that it should b e. 53

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Whitby 1997, Lublin 1997, Canon 1999). The race o f the representative and the urban

percentage o f the district are both statistically significant predictors o f black interest

responsiveness. Notably, the race variable's impact on congressional votes in favor of

black interests operates differently in each model, predicting behavior in opposite

directions. That race gains explanatory power in the subjective black interests model,

becoming the most important predictor o f members' voting behavior in favor o f black

opinion, is o f added import, hi the objective interests model, a black representative will

have an LCCR score 23 points higher than a white colleague. In the subjective interests

model, having a black representative will reduce the likelihood o f a vote in favor o f black

interests. •

The relative size o f the district's urban population also influences representatives’

responsiveness to black interests across the models, but once again in opposite directions.

Representatives from urban districts are somewhat more likely to represent objective

black interests and somewhat less likely to represent subjective black interests. Seniority,

a variable not usually included in models of policy congruence, attains statistical

significance in the objective black interest model, reducing the likelihood of liberal

voting. Having a representative from a southern district has a negative effect on the

representation o f black interests whether they are objective or subjective, although region

is only statistically significant in the objective black interests modeL Not surprisingly,

the black population percentage in the district is not a statistically significant predictor in

either model. More often th a n not, the percent o f the district that is black fails to reach

54

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. statistical significance when the race o f the representativeis also included in the model

(see Swain 1993, Whitby 1997, Canon 1999).27 f, , : r Although party is the most important predictor for representatives* support o f

objective black interests, contributing more than double the explanatory power o fthe race

variable, partisanship is conspicuously absent from among the statistically significant

predictors o f subjective black interests though it is in the expected direction. This failure

to reject the null hypothesis that Democrats and Republicans do not behave differently

when it comes to representing black opinions is contrary to my expectations that liberal

Democrats would be consistently less likely than their Republican colleagues to represent

the conservative opinions of blacks in the BOI measure and thus warrants further

investigation.

33.1 Com paring LCCR and BO I Scores

What are the differences between what the LCCR and BOI scores measure?

Theoretically, the LCCR score is used to measure House members’ responsiveness to

objective black interests, i.e. those defined by observable phenomena. For instance,

members with high LCCR scores are expected to vote in favor o f a range of liberal

policies in the objective interests o f blacks, like supporting affirmative action, inner-city

enterprise zones, and increasing the minimum wage (see Swam 1993, Whitby 1997).

27 The inclusion o f both the race o f the member and the percentage o f blacks in district are important to maintain the underlying theory o f the model. I agree with Canon (1999, 180) that both o f these key explanatory variables must be included in order to truly understand the nature o f racial representation. 55

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Hie BOI, on the other hand, measures how well legislators represent black public

opinion, or subjective interests; when they cast roll-call votes.2*

Caution is key when matching public opinion positions to roll-call votes. For

instance, 70 percent of respondents in the 1996 NBES support laws that protect

homosexuals from job discrimination. The two major issues regarding gay civil rights

that came up for roll-call votes in the 104th Congress were federal recognition of same-

sex marriages and rights for domestic partners. Two different sorts o f questions emerge.

First, does black majority support for gay rights have any implication for what blacks

might believe about the touchy issue of same-sex marriage? Recall the Tate (1994)

finding that a substantial majority o f blacks think that homosexuality is always wrong

and couple that with the fact that many black religious organizations have expressed

opposition to codifying same-sex marriage, and one might conclude that black support

for gay civil rights may not extend to redefining marriage. In the absence of black

opinion survey data directly addressing the issue o f same-sex marriage, the position of

the masses on this question remains unclear although 67 percent o f black and 18 percent

o f white Democratic representatives voted against the federal ban on same-sex marriages

passed by the House in the 104th.

The domestic partners issue demonstrates another way that the waters can be

muddied when using roll-call votes to assess responsiveness to public opinion. The

Prohibition on Domestic Partners Act was actually subsumed in the 1996 District of

“ The BOI could benefit from the inclusion o f more roll-calls (six are included in the analysis). However, the researcher is limited by the availability o f roll-calls that match issues cm which black have been polled and vice versa. For instance, while black opinion 56

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Columbia Appropriations Act. Without additional information on why members voted as

they did, researchers are hard-pressed to know what motivated members to vote as they

did. Was the member voting for D.C. funding or against the Domestic Partners Act, or

both?

Despite these cautionary statements, the BOI provides a reliable measure of

subjective black interests® The BOI and LCCR scores clearly measure two different

aspects o f black representation, but how do members' BOI and LCCR scores compare?

Table 33 presents mean LCCR and BOI scores by race and party.

INSERT TABLE 33 ABOUT HERE

333 Difference of Means for LCCR and BOI Scores

A difference o f means test for members’ LCCR and BOI scores provides baseline

evidence o f variations in voting behavior among black and white Democrats and white

Republicans. While this simple analysis does not allow us to draw conclusions about

why representatives o f different races and parties scored differently on the two measures,

it does demonstrate that statistically significant differences occurred. There is a 29 point

difference between LCCR scores for black and white Democrats, and a 78 point

difference between black Democrats and white Republicans, demonstrating how solidly

liberal African American legislators are as a group compared to their white colleagues. It

is interesting to note that the minimum score attainted by black representatives was 65,

®An ideal study o f subjective black interests would use survey questions that asked respondents about specific policies that had been voted on a previous Congress, rather than general policy preferences unrelated to actual votes. That way, direct comparisons could be made between how legislators voted and black preferences. 57

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. indicating that every black member voted in support o f the LCCR position at least 13 out

of 20 times. Roll-call voting behavior among white representatives was more variable;

with a minimum o f three favorable LCCR votes for white Democrats and as few as one

for white Republicans.

The BOI shows a 27-point difference between mean scores for black and white

Democrats and a 38 point difference between black Democrats and white Republicans. A

look at the m in im u m scores received by each group shows that white Republicans voted

in accordance with black public opinion at least 50% o f the time. A closer look at the

how members voted on the specific issues in the BOI will shed light on why party was

not a significant predictor of support for subjective black interests.

INSERT TABLE 3.4 ABOUT HERE

3 3 3 Support for Subjective Black Interests by Race and Party

Table 3.4 presents representatives' support for individual BOI issues by race and

party. African American representatives? support for black opinion was strongest on the

affirmative action vote and only slightly positive for defense spending. On the issues of

increased crime spending, expediting death penalty cases,, restricting illegal immigration,

and welfare reform, black leaders were almost uniformly at odds with black opinion, but

very much in line with a liberal agenda. White Democrats also strongly supported black

opinion m favor of affirmative action, whereas every white Republican was on the

opposing side. Most white Democrats voted in opposition to black opinion on defense

spending and welfare reform, however their support was more evenly spread on crime

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. spending and the death penalty than that of their black counterparts. Nearly three-

quarters o f white Democrats supported the BOI position limiting illegal immigration to

the United States.

As a testament to the party control exhibited by the Republican Party during the

104th Congress, white Republicans voted in tandem on five out o f the six issues. On two

issues, they uniformly opposed black opinion - voting against protecting affirmative

action benefits for minority broadcasters and for a substantial increase in defense

spending- The fact that white Democrats were more willing to support conservative

positions on issues and voted less often as a bloc provides a good explanation for why

partisanship did not reach statistical significance in the BOI model of support for

subjective black interests. Unlike white Democrats, African American members rarely

veered from a solidly liberal agenda and thus found themselves in opposition to aggregate

black opinion on a number of issues where blacks expressed conservative preferences.

3.4 Conclusion

When it comes to representing black interests in Congress, is the liberal vote

enough? The short answer is, no. When examining the substantive representation of

black interests, scholars should take into account both objective and subjective interests.

The analysis in this chapter reveals several things: 1) that race is indeed an important

factor m the representation of black interests, 2) that when black interests are defined

objectively, the race o f the member is an important predictor, but not as important as

party, and 3) that when black interests are defined subjectively, having a black

59

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. representative lessens the likelihood that black opinion will be supported in roll-call

votes.

It also makes the case that black interests are not always analogous to liberal

policy positions and urges scholars to take black subjective interests — the self expressed

interests o f African Americans— into account. This finding points out the caution that

must be exercised when scholars seek to understand black interests representation solely

through the analysis o f roll-call votes.30 Congruence between representatives' votes and

the needs and wants of the represented are important, but incomplete measures of

representation. The coincidental support o f black opinion by Republicans is indicative of

as much.

Despite evidence that white Republicans are more likely than black Democrats to

represent black subjective interests, only IS percent o f blacks believe that Republicans in

Congress work hard for their interests, while nearly 74 percent believe that Democrats

do.31 Blacks do not see themselves as primary constituencies o f the Republican Party

nationally and rarely see their issues promoted by Republicans. This supports the

argument that the liberal agenda that is at the core of black objective interests is central to

how blacks assess who represents their interests. As suggested above, some preferences

are likely more important than others when blacks determine who represents them in

Congress. Blacks recognize that African American representatives and many white

“In other work on the topic o f black representation, I argue that in order to understand black representation scholars must consider representatives behaviors beyond roll-call voting, to include bill sponsorship, floor speaking, and constituency service (Sinclair Chapman 1998). 31 National Black Election Study, 1996. 60

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Democrats count blacks among their strongest electoral constituencies and thus at some

level seek to represent the interests of this group. The congruence between white

Republicans* votes and black opinion on the death penalty, immigration, ami welfare

reform is largely coincidental and not likely due to these representatives actively seeking

to represent black interests. That white Republicans uniformly opposed the affirmative

action proposal included in the BOI supports this assertion.

The finding that African American representatives often fail to support black

policy preferences on issues where black opinion is conservative suggests that these

representatives are acting as trustees rather than delegates in their representative roles.

For example, the majority o f black constituents may very well have been un-represented

by black legislators on the welfare issue. With the exception o f Representative Sanford

Bishop (D-GA), black Democratic House members consistently voted against welfare

reform.

An alternative perspective argues that black legislators in the 104th Congress may

have been willing to support a “kindler, gentler" version o f welfare reform than the one

presented for floor vote by House Republicans. After ail, many liberals have taken up the

cause o f welfare reform, including Marion Wright Edelman, president o f the Children’s

Defense Fund. One reason for their support was that liberals assumed that the welfare

reform bill was part o f a package deal that would include broad new social policies,

in c lu d in g universal health care. Many liberals also became enthusiastic supporters of

welfare reform because they believed that they would remain in control o f Congress and

thus would be its primary architects. As President Clinton’s approval ratings fell along

61

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. with his national healthcare initiative, a more conservative version o f welfare reform

became the centerpiece o f his re-election campaign. After Republicans gained control o f

the House in the midterm elections o f 1994, a more conservative version of welfare

reform gained momentum.

Are blacks being duped by "false consciousness” when black opinion conflicts

with objective black interests and, thus not folly able to recognize their own best interests

(see Zaller 1992 for a discussion of the pitfalls of this line o f thinking, see also Swain

1993, 5-7)? One could point out the irrationality of blacks supporting welfare reform

when poor African American women and children will be disproportionately

disadvantaged. However, commenting on that question goes beyond the scope of this

research. We must take African Americans at their word and trust that the opinions they

express reflect what they truly believe is in their own best interest. Admittedly, the

measure o f subjective black interest used in this paper is limited in that public opinion

polls rarely tell us how respondents rank their preferences for concrete policy proposals.

Would blacks have preferred a less stringent welfare reform measure than the version that

passed the House? In addition how important a signal was the welfare reform vote to

black voters? Was a legislator's support for welfare reform sufficient to make black

respondents believe a particular member might not represent their interests? A better

indicator for constituents about whether a representative supports black interests might be

a legislator's vote against legislation where blacks are explicitly beneficiaries or losers,

such as affirmative action.

62

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Perhaps black constituents tolerate trustee-style behavior from black

representatives on some issues (like welfare reform and the death penalty) because they

can trust them to behave as delegates, strongly advocating then interests, on others that

are more salient (like affirmative action and civil rights policies). Bianco’s (1994)

assertion that as legislators gam the confidence and trust o f their constituents they also

gain voting leeway, or in Fenno’s (1978) terminology, “wiggle room,” may account for

blacks’ apparent acceptance o f black legislators’ liberal voting. Research included in the

larger study of which this is a part suggests that it may also be the case that black

Americans recognize and value some of the other aspects o f representation that African

American legislators engage in, including bill sponsorship and other behaviors that help

place black interests on the national agenda.

The way that blacks identify what their interests are and who represents them is

likely more complex than researchers can capture in empirical models. Examining both

objective and subjective measures of black interests aims to address some of this

complexity by recognizing the altruism and group interests associated with objective

black interests, while at the same time paying attention to black preferences and

individual self-interests beyond the group. This research suggests that scholars should m include measures of both objective and subjective brack interests in future studies of

black substantive representation. In the next chapter, I consider a broader approach to

black representation that includes symbolic representation as well as substantive

representation in the context o f legislative behavior beyond voting, namely, through bill

sponsorship.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Explanatory Variables LCCR Score Coefficient (SE) Representative Characteristics Race 22.8** (7 3 ) Party 4735** (3 3 ) Seniority -0.41* U 8 ) District Characteristics Percent Black -0.08 (-18) Region -10.19** (3 3 ) Percent Urban 033** (.08)

Constant 8 3 3 (9.66) F-statistic 87.17** Adjusted R1 .85 Number o f Cases 96 **p<.Ol *p<05 (two-tailed test) +p<.10 (one-tailed test) Only House Representatives from districts with at least 15% black populations were used in this analysis. Four members* Reps. Nffume (I>MD% Reynolds (D-IL), and Tucker (D-CA) were excluded because tfiey did not complete full terms during the 104 . Rep. Serrano (D-NY) was excluded to simplify the race v ariab le.

Table 3J Ordinary Least Squares Regression Model of Representatives* Support for Objective Black Interests (LCCR Scores), 104* Congress (1995-96)

64

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. BOI Score Explanatory Variables Coefficient (SE> Representative Characteristics Race -2.09** (.62) Party -0 2 8 (3 2 ) Seniority -.002 (0 1 ) District Characteristics Percent Black .005 (0 1 ) Region -0.07 (30> Urban -0.02** (.01)

Log-Iikelihood 218.18 Chi-square (6) 73.51** PPC 54% PRE 2 9 Number o f Cases 96 •*p<.Ol *p<05 (two-tailed test) +p<. 10 (one-tailed test) Only House Representatives from districts with at least 15% black populations w oe used in this analysis. Four members, Reps. Mfume (D-MDX Reynolds (D-ILX and Tucker (D-CA) were excluded because they did not complete full terms during the 104 . Rep. Serrano (I>NY) was excluded to simplify the race v a ria b le .

Table 3.2 Ordered Probit Regression of Representatives* Support for Subjective Black Interests (BOI Scores), 104* Congress (1995-96)

65

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Party and Race LCCR Min Max BOX M in Max Democrat Black 93.78 65.00 100.00 27.76 0.0 50.00 (33) (33)

White 67.11 15.00 100.00 55.70 16.67 8 8 3 3 (38) (38) Republican Black ------

White 15.80 5.00 55.00 64.67 50.00 66.67 (25) (25) F-statistic 16030** 5835** Note: Table entries are LCCR and BOI Support Scores Number o f cases in parentheses **p<.01

Table 33 Mean LCCR and BOI Scores by Race and Party of Representative, 104th Congress (1995-96)

Black Opinion Index (BOI) Issue Areas A . Action Crime D . Penalty Defense Tmm. Welfare Race and Party Democrat Black 87.9 3.0 0.0 57.6 9.1 3.0 (29) (1) (0) (19) (3) 0 )

White 92.1 57.9 52.6 2 6 3 78.9 2 6 3 (35) (22) (20) (10) (30) (10)

Republican Black

White 0.0 88.0 100 0.0 100 100 (0 ) (22) (25) (0 ) (0) (25)

Note: Number o f cases in parentheses. Issues: Affirmative Action, Crime Spending Death Penalty, Defense Spending, Immigration Limits, and W elfare Reform

T able 3A Percent Support lor Black Preferences on BOI Issues by Representatives* Race and Party, 104th Congress (1995-96) 66

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER 4

BLACK REPRESENTATION THROUGH BILL SPONSORSHIP

Symbolism forms a large and important part of political activity. It supports governments, selects leaders, and defines the terms o f debate. Symbols can be used politically to shape political attitudes, build support, persuade to action, or in one widely accepted definition o f political power, to help A get B to do what A wants done. Writers point out that the central battle in a political conflict is often “the struggle over whose symbolic definition will prevail.” Control o f the symbolic actions o f government is as important as the control o f its tangible effects.

Representing black interests through voting for policy proposals that reflect black

concerns is an important measure of substantive representation, or acting on behalf of

black constituents. The previous chapter examined which members o f the 104 Congress

represented objective and subjective black interests. Objective black interests are

determined by social and economic indicators o f the status of blacks compared to whites,

while subjective black interests are determined by the matching black opinion data on

public policy preferences with favorable roll-call voting. Using a liberal interest group

score—the LCCR score for the 104th Congress, I found that black and white Democrats

were more likely to support black objective interests. When black interests were

measured using an index of black policy preferences—(BOI), I found that black

Democrats were considerably less likely than many white Democrats and Republicans to

vote in support o f subjective black interests. It turns out that black Democratic legislators

- - a y— 32 Barbara Hmkley (1990,1 emphasis added) The Symbolic Presidency.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. are more liberal than blacks are generally as a group. While these findings are

instructive, voting on policy initiatives does not account for the range of legislative

activities available for use in representing the interests o f constituents. In this chapter, I

examine bill sponsorship as tool for representing the interests o f black Americans. Bill

sponsorship, as compared to voting, is a relatively unconstrained activity that allows

legislators great freedom in deciding how many bills to sponsor, what issues to tackle,

and what purposes various bills will serve. Representatives’ legislative portfolios

indicate their issue priorities. Thus, one way for legislators to present themselves to

colleagues and constituents is through the use o f bill sponsorship to indicate their support

o f particular issues or constituents’ interests.*3

Does the race or party o f the representative matter in determining who sponsors

bills that directly or indirectly benefit blacks? Are black representatives more frequently

sponsors o f bills that specifically name blacks as the beneficiaries than whites? Further,

do representatives differ along racial or party lines in the substantive or symbolic focus o f

the black interest bills they sponsor? The literature provides evidence that the race o f the

representative is an important factor in black representation at the bill sponsorship stage.

Canon (1996) found that among House members from districts with at least 25% black

populations, nearly half of black legislators included bills having some racial content in r their legislative agendas, compared to nearly none o f the whrtemembers during the 103rd

Congress. Cobb and Jenkins (1996) also found that the race o fthe legislator was the key

“Schiller (1993) provides a good account of senators’ use of bill- sponsorship and possible relationships between bill sponsorship and representation. See Fenno (1978, chapters 3-4) for a discussion o f the many ways representatives “present” themselves to them constituents. 68

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. explanatory variable in an analysis o f black interest bill sponsorship m the Democrat-

controlled 103rf House o f Representatives. At the state level, Bratton and Haynie (1994)

found that the race o f the legislator had a greater effect on black interest bill sponsorship

th a n the size o f the district’s black population.34 In a cross-sectional analysis of state

legislatures in California and Illinois at three different time periods, Bamello and Bratton

(1998) found that having a black legislator was a statistically significant factor in black

interest bill sponsorship, however the impact o f the legislator’s race declined over time.

In a study of five state legislatures, Haynie (2001) found that majorities of black

legislators introduced black interest bills and that they introduced more black interest

bills proportionally th a n did their nonblack colleagues. He found that black legislators

were twice as likely as other legislators to include black interest bills in their legislative

agendas. The one published account o f bill sponsorship by black legislators in the 104th

Congress was largely descriptive. This study found that the majority o f bills sponsored

by black legislators were, in fact, symbolic, neither distributing nor redistributing any

public good in the standard sense (Tate 2002). Previous scholarship by the author

(Sinclair -Chapman 1996), found that liberal legislators, both black and white, were more

lik ely th a n others to sponsor bills indirectly benefiting blacks, but only black

representatives sponsored bills directly addressing black interests in the 103rd and 104th

34 Bratton and Haynie (1994) also find that black state legislators are often less likely to get their legislation passedth a n white counterparts. No such relationship between member race and legislative success has been found at the Congressional IeveL In fact, evidence from the 103rf Congress indicates that black legislators were slightly more successful at getting their proposals enacted than white colleagues (Tate 2001, Canon 1999, Anderson, Box-Steffensmeier, and Sinelair-Chapman2002). 69

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Congresses.35 Based on extant literature, my expectation is that liberal and black

legislators will be more likely than others to use bill sponsorship as a tool of black

representation, and that black legislators in particular w ill sponsor bills directly naming

black beneficiaries and advocating symbolic black policy concerns.

4.1 Bill Sponsorship and Representation

Before embarking on a discussion of the theoretical importance of symbolic

representation, in this section I consider the connection between bill sponsorship and

representation. What leverage does bill sponsorship add to more fully understanding

black representation? What evidence suggests that bill sponsorship is used as a

representational tool by legislators, generally, and more specifically for blacks?

Goal-oriented representatives approach bill sponsorship with the same strategic

objectives as they do other legislative tasks. They craft their legislative portfolios to meet

one or a mix of three goals: reelection, good public policy, or political power. Members,

primarily concerned with reelection, weigh the pros and cons o f bill introduction in light

of the possibility of alienating, maintaining, or winning votes in the next election.

Aiming to lessen the first outcome and improve the second and third, a member pays

careful attention to the district when she makes the decision to draft legislation that will

bear her name. This occurs despite the fact that constituents are largely unfamiliar with

the legislative agendas o f their elected officials.

35These preliminary findings are based on a s m a ll sample and cannot be used for generalizations. 70

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. In an analysis of bill sponsorship in the U.S. Senate, Schiller (1993) notes that

constituents rarely named a specific piece o f legislation as a reason for liking or disliking

a senator. Americans generally demonstrate low levels o f political knowledge in regard

to those elected to represent them and blacks are no exception. Tate (2002) finds that

barely a quarter o f blacks represented by black House members could correctly recall the

name o f their representative and even smaller numbers, 10 and 8 percent respectively,

could correctly recall the names o f their white or white Republican representatives. Yet,

members of Congress anticipate the preferences o f various constituency groups in the

district prior to sponsoring or cosponsoring bills, supporting other colleagues* legislative

efforts, and casting votes. As one Senate aide put it, “[b]ills are a way for senators to

respond to constituent concerns. They can be touted to constituents. There are a lot of

voters who are interested in policy issues—you basically can’t afford to waste time on

issues that don’t reflect constituent concerns” (Schiller 1993,163).

Translating specific policy proposals into messages o f district representation

requires that legislators spend considerable time publicizing their bills and educating

constituents. These messages may be communicated to political elites first, who then

t r a n s m it the messages to various district constituencies. Schiller (1993, 177) speculates

that a senator who has introduced a bill benefiting teachers might inform officials of the

state teachers’ union in hopes that the officials would inform rank and file members o f

the senator’s action. Likewise, representatives wishing tocommunicate their advocacy o f

black concerns to black constituents must find effective ways to disseminate information

about their policy initiatives. The end result is that constituents form a general impression

71

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. of the representative based on cues from trusted elites or other sources (like radio,

television, or pout news outlets) rather than knowledge about specific policy proposals.

The animat Congressional Black Caucus Dinner and Legislative Weekend, first

established as apolitical fund-raiser and issue forum in 1971, is an example o f how black

legislators publicize then legislative agendas to select elites, including black state

legislators, local ministers, business leaders, entertainers, academics, and black media

sources with the objective of generating favorable impressions among various

constituencies. Giving his account o f this event, Representative William Clay (D-MO)

characterized the first CBC dinner as the “most memorable event for black Americans

since the 1963 March on Washington” (Clay 1993,161). The goal of the dinner was

political and “aimed at sensitizing the black electorate...for effectively exercising real

political power” (Clay 1993, 169). According to Representative Clay, “[i]n convening

this meeting, the Caucus took a giant step in advancing the cause of black people by

solidifying broad based, national support for the development of a black agenda” (Clay

1993,161).

In the absence o f organized forums, representatives use their privileges as House

incumbents to relay information about their policy proposals to constituents. House

members use franked-mail newsletters and websites maintained by House staffers to

publicize activities to their constituents. They also make appearances at various venues in

the district, such as black churches or predominantly black schools, to demonstrate their

accessibility to certain groups and to advertise the things they are doing on the group’s

behalf. Representatives thus create reputations for themselves as advocates o f certain

groups or interests and opponents o f others.

72

4

&■% : ■' ■- ■ ■ - '• "‘.■v,;' • * Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. For black members, notably, bill sponsorship provides an important representation

tool in a political body dominated by whites. Bin sponsorship is a means to enter and

recast political debates and shape institutional legislative agendas for previously-

excluded groups. In a statement to a breakfast forum during the 1993 CBC Legislative

Weekend, Representative Maxine Waters noted that “[tjhis discussion is so important

today because there is a concentrated, well-organized effort to keep our voices closed and

shut down on the discussion of race” (Gill 1997). On a separate occasion she also called

attention to how the composition of legislature contributed to a lack o f attention to the

concerns o f racial minorities, the poor, and the disillusioned.

This institution has been the domain o f white males, for the most part. And because we've always had so few African Americans and so very few African American women, most of what ails us is in our communities and most of our concerns were just never debated in this House.36

Members o f Congress, thus, use bill sponsorship as a tool to meet their myriad

goals. Individual members decide how many bills to sponsor and on which issues. More

so than other legislative activities where choice is dependent on the decision-making o f

other colleagues (e.g., strategic voting and committee assignments) bill sponsorship is,

for the most part, an independent choice.37 B ill sponsorship allows House members to

introduce new issues to the political agenda, offer alternatives to established policy

approaches, and support or challenge the status quo on their own terms. For instance,

Representative Maxine Waters, a black Democrat from California, responding to the

36 Representative Maxine Waters as quoted in Gill (1997,126). 37 Schiller (1993) notes that senatorsface certain institutional constraints when deciding to sponsor bills, m cfadm g committee assignments, leadership positions, and staff resources. 73

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. sentiment among blacks in her district that the Los Angeles police department more often

abused than protected blacks, sponsored a House resolution urging the prosecution of

former Los Angeles police detective Mark Furham and the adoption o f reforms in the

police department following the O J. Simpson trial. She also sponsored a resolution

seeking to establish a temporary House committee to investigate the involvement o f the

Central Intelligence Agency in crack-cocaine sales to benefit Contra operations in

Nicaragua based on a series o f reports in theSan Jose Mercury News that were later

found to have some validity (see Duncan and Lawrence 1997). Her actions underscore

the doubts that many blacks (in her district and elsewhere) have regarding the

government’s willingness to sacrifice the needs and concerns of blacks for its own

purposes.” In an effort to bring attention to racial disparities in the criminal justice

system and the detrimental effects on black political participation, Representative John

Conyers, a black representative from Michigan, introduced a bill to secure the voting

rights of former felons. Similarly, in a symbolic gesture of solidarity with people of

Polish descent (who comprise the largest ethnic group in his district), Representative Jack

Quinn, a white representative from New York, sponsored a resolution commemorating

the 205th anniversary o f the adoption o f Poland’s first constitution.

For some, bill sponsorship serves as not only an efficient and effective tool to

communicate to constituents, “I am one o f you,” or “I empathize with you,” it also serves

as a means to expand the legislature’s political agenda. Hence, Representative Waters

“Conspiracy theories regarding government’s role underminingin the stability o f the black community are not uncommon. See Cohen (1999) for an account o f black concerns about the role o f the U.S. government in introducing the AIDS virus to the black c om munity . 74

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. used bin sponsorship to remind the legislature o f an issue that is important in her district

as well as to assure her constituents that she takes the issue seriously enough to expend

political capital on it Representative Conyers* bill on the disenfranchisement o f former

felons directed attention to what he views as an issue of black representation and set the

stage for national discourse. Representative Quinn is not Polish, but he used bill

sponsorship as a tool to communicate to his PoliSh constituents that he recognizes their

heritage and empathizes with them. For all three legislators, these kinds of gestures

through bill sponsorship likely aimed at securing the trust and loyalty of certain

constituents, with the ultimate goal o f gaining a political benefit at the polls.

4.2 Theoretical Considerations: Symbolic Representation

Proposing legislation is a basic legislative function. However, not all bills are

sponsored with the intent of getting policies enacted. The fact that some bills are

sponsored for symbolic reasons is well recognized by congressional scholars. Sometimes

legislators sponsor bills with an intent aimed more at agenda-setting and position-taking

than policy-making (Gross 1953, Mayhew 1974). Mayhew (1974, 62n) recounts a

response from a representative m Clapp’s (1963) The Congressman: His Work as He

Sees It, who says, “I try to introduce bills that illustrate, by and large, my ideas—

legislative, economic, and social. I do like being able to say when I get cornered, ‘yes,

boys, I introduced a bill to try to do that in 1954.’ To me it is the perfect answer.”

Mayhew assigns this type o f activity little value and suspects that voters do the same,

however some members of Congress expect just the opposite and spend considerable

time crafting legislative portfolios that anticipate district needs. In these instances, bill

75

I

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. sponsorship may function symboIicaUy to give a voice to issues, rather than to necessarily

implement new public policy (Edelman 1964; Elder and Cobb 1983). This concept of

symbolic legislation serves as a prime example o f the concept o f symbolic representation

developed in this chapter. Mayhew* s misgivings notwithstanding^ the fact that some

legislation is symbolic does not make it any less important than more substantive

legislation. Indeed, Elder and Cobb (1983:20-22) contend that, [sjince most people are

not very vigilant in their monitoring o f the political process and lack detailed information,

they rely on symbolic cues in making their assessments... It is the making o f policy rather

than its execution that the public is most sensitive to.”' v "

T .inking bill sponsorship to representation requires an acknowledgment o f both

the substantive and symbolic functions that legislation canfu lfill. There is a difference

between substantive bills and bills that have substantive impact. A bill can be substantive

in focus and, in the absence o f enactment, have no impact at all. An example of such a

case is legislation written to ensure fair-lending practices in minority communities. If the

bill were clearly targeted against specified lending practices and clear implementation

mechanisms were outlined, the bill would have a substantive impact if it were enacted.

The point is that the absence o f enactment, and thus substantive impact, does not preclude

it from having a substantive focus.

Symbolic legislation differs from substantive legislation in several ways. The key

distinctions derive from the feasibility of implementation and the likelihood of

substantive impact. Kingdon (1995,200-201) describes five characteristics that separate

feasible policies from those that are likely to fail: I) technical feasibility, 2) congruence

with the values o f policy community members, 3) consideration o f budgetary and other

76

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. constraints, 4) the support o f mass and specialized publics, and 5) receptive politicians in

the legislature. Integrating symbolic responsiveness theories from Edelman (1964),

Pitkin (1972), and Eulau and Karps (1977) with Kingdon’s (1995) theoretical analysis of

agenda-setting and policy formation, I developed a four-criteria measure to identify

symbolic policy proposals.39

First, symbolic bills are often, radical, sweeping departures from the “policy

streams”40 that dominate the House policy agenda and general voter dispositions. These

bills do not fit within the pervading policy objectives of most members o f the political

community — they are policy outliers. Second, symbolic bills target largely powerless

groups with few resources for mobilization and of little threat to other members of

Congress. Third, symbolic legislation is often technically unfeasible. That is, the details

o f action are unspecified and there are no identifiable implementation mechanisms. For

instance, a bill that promised universal employment for the poor would be coded as

symbolic because the targeted group is largely powerless to decisively influence the

decision-making o f most members and the implementation is technically unfeasible. The

final criteria is that symbolic legislation will have little or no tangible impact for the

targeted group even if it is enacted.

My work distinguishes itself from similar research by offering a theory of

symbolic representation that not only captures obviously symbolic activity like

39A t a m inim um , a bill must satisfy only one criterion in order to be classified as symbolic. In general, most symbolic bills will satisfy more than one criterion. For instance, obviously symbolic bills, fike naming post offices, may only satisfy the “little or no tangible impact even if enacted” requirement 40 See Kingdon, John W. 1995.Agendas, Alternatives, and Public Policies, 2nd ed., especially pp. 131-139. 77

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. sponsoring commemoratives or resolutions, bat that also accounts for policy proposals

that appear to be substantive but in reality are only symbolic gestures. Three examples o f

symbolic legislation w ill illustrate these points.

An easily recognized symbolic bill is the kind aimed at naming a local post office

in the district in memory o f a civil rights leader, or a postage stamp after B.B. King.

Although these bills objectively promote black interest and are often enacted, their

tangible impact is negligible. A more sophisticated form of symbolic legislation is the

kind that seeks some lofty benefit for the black community that is absolutely unlikely to

be realized given the dominant policy moods in the country and in Congress.

An example o f this is HJR.40, a bill introduced by Representative John Conyers

(D-MI) early in the first session of the 103rd Congress/1 This bill called for a

commission to study the need for and allocations o f reparations to African Americans for

slavery-related injustices. The first and last official action on the bill was its referral to

the Judiciary Committee on January 5, 1993. Although Conyers recruited 29 co­

sponsors, including one Republican, he did not manage to get the bOl reported out o f the

Democrat dominated sub-committee. In fact, o f the 34 pieces o f legislation originally

sponsored by Conyers during the 103rd Congress, only 2 bills received affirmative votes

in the House. This dismal record o f legislative effectiveness did not prevent Conyers

from touting his legislative agenda during the annual Congressional Black Caucus

Legislative Weekend the following foil. Nor did it prevent him from hosting a legislative

sem inar on the issue o f African American reparations during that event.

41 Representative Conyers has introduced this bill in several Congresses. 78

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Still, symbolic legislation need not be characterized as entirely futile. A bill

introduced in hopeless conditions during one Congress might find its moment o f triumph

in a future one provided that the political climate and public opinion sway in its favor.

For example, Representative Ron DeUums (D-CA) introduced legislation establishing

economic sanctions to protest apartheid in South Africa during every session o f Congress

for almost two decades before it was favorably reported from committee and passed by

the House in 1986. While substantive in its effects for black and white South Africans,

the bill remains largely symbolic with regard to addressing the interests of African

Americans in the United States because it fails to directly or indirectly affect then-

objective policy interests. For the most part, black interest in South African apartheid

represents a symbolic attack on institutionalized racism. Legislative action in this area

does not reflect the fulfillment o f material policy objectives for the black community.

Substantive bills are those that directly or indirectly address black interests with

narrow goals, clear implementation mechanisms, and tangible results if enacted. A bill

sponsored to encourage economic development through the creation of empowerment

zones and enterprise communities is an example o f a bill coded as substantive. Another

type o f bill presents a hybrid mix of symbolic and substantive goals. These hybrid bills

are neither purely symbolic, nor purely substantive. Legislation sponsored to require the

mandatory registration o f handguns is coded as a hybrid bill that indirectly affects black

interests. If gun registration actually results in a reduction of the number of guns in

minority neighborhoods, then the legislation w ill have a substantive affect on black social

well-being. The bill is symbolic, particularly in the political context of the 104*

Republican-controlled Congress, because it was outside the prevailing policy stream.

79

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 42. Data and Methods

Using the same data and model as described in chapter 3 (section 3.2), I use

Ordinary Least Squares regression analysis to examine who sponsors direct, indirect,

symbolic, substantive, and hybrid black interest bills.42 Because o f high multicollinearity

between member race and the racial composition o f congressional districts (r=.91), I ran

the models with and without the percent black.0 The dependent variable is the

percentage o f total bill introductions by each member that reflects black interests with a

direct, indirect, symbolic, substantive, or hybrid focus. The model is specified as

follow s:

BIB = bo+ biRACE + l^PARTY + b3SEN + b4%BLACK + bs%URBAN + beREGION + e where, BIB = percentage of black interest bills by type RACE = race of legislator, I if black, 0 other PARTY = party of legislator, 1 if Democrat, 0 other SEN = seniority of legislator, the number of terms in office %BLACK = percent black in the district %URBAN - percent o f district characterized as urban REGION = region of country, 1 if south, 0 other

42 Using an event count model is probably more appropriate for this data. Because most members did not sponsor any black interest bills, the data is overdispersed at zero. This presents problems for OLS models, whereas a Negative Binomial Model better fits data distributed in this fashion. Canon (1999,194) suggests that using the proportion o f bills sponsored that have varying content helps to correct for the variation and instability in the data and this is the approach that I have taken in this project. 43 As presented in the tables, the standard model includes both the race o f the member and the racial composition o f the district. The amended model includes only the race o f the member. 1 also ran the models without the race of the member. The racial composition o f the district added far less explanatory power than the race o f the member both in terms o f the size o fcoefficient and the amount o f variance explained. 80

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Based on extant literature* I expect that the race o f the legislator* the legislator's

party affiliation* and the racial composition o f the district w ill be important predictors o f

btH sponsorship benefiting blacks. The race o f the member is the key explanatory

variable for black interest bill sponsorship at the state and national level (Bratton 2002*

Whitby 2002* Haynie 2001* Canon 1999* Cobb and Jenkins 1996). I also expect a

positive relationship between party and black interest bill sponsorship. The descriptive

statistics show that* in terms o f the number o f bills sponsored* Republicans are less likely

to represent black interests through bill sponsorship than Democrats. Canon (1999)

found that the racial composition o f the district had a small* but positive effect on the

proportion o f members’ legislative portfolios that contained racial or part-racial content.

I expect that* all else being equal* the percentage ofblacks in the district population will

increase the proportion o f policy proposals House members devote to the interests o f the

group.

Black Democrats in this sample sponsored 335 bills and amendments* introducing

an average of 10.15 bills per member. White members sponsored 620 bills* averaging

9.84 bills each. Democrats averaged 103 bills each* for a total o f 731 bill introductions,

while Republicans — the majority party members— sponsored 224 bills with an average

of 836 bills each. Each bill and amendment was coded for whether it reflected black

interests, whether it identified blacks as direct or indirect beneficiaries, and for its

symbolic or substantive focus.44

44 Through the course of this research I spoke with several legislative directors and assistants from Capitol Hill about my conceptualization of bills in this way. My interviews (not presented here) with them about their member’s bill sponsorship behavior fit nicely into the conceptualization 1 develop. 81

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Based os the coding scheme outlined here, a total o f 190 bills were coded as

beneficial to blacks. I first coded bills based on objective black interests addressing

housing accessibility, job creation and protection, improvement o f the educational

system, healthcare quality and accessibility, and civil rights. For instance, bills aimed at

ending mortgage redlining practices, encouraging urban revitalization, and preventing

crime were coded as black interest bills. I then coded bills in terms o f their direct or

indirect benefits to blacks. Legislation aimed at increasing the number o f black- and

women-owned small businesses was coded as directly beneficial to blacks, while

legislation seeking to raise the minimum wage was coded as indirect.

Finally, I coded bills according to their symbolic or substantive focus. Purely

symbolic bills are mostly commemorative bills. Many o f these are sponsored in the form

of non-binding House resolutions that do not have the force of law, but rather give a

“sense” o f the sentiment o f the House on a particular issue. These resolutions often pass

the House by voice (as opposed to recorded) vote and are thus relatively low cost to

sponsor and vote for. Simple symbolic bills, such as the bill to mint a one-dollar com

with the likeness o f the Reverend Martin L. King, Jr.. are sometimes introduced as public

bills (designated with H.R. rather than an H.Res.) and were also coded as symbolic. Bills

that addressed concerns of people of color and oppressed groups internationally,

including a bill to increase the funding allocation for the African Development Fund,

were coded as indirect and symbolic because they neither directly affect American

blacks, nor do they substantively affect African Americans^ material or social well-being

82

f - ji' •T"’- Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. In the United States.41 Substantive bills, such as a bill to increase the funding resources

for a local historically black university, are those that, if enacted, would have a

substantive impact on black public policy needs. Others, including a bill requiring that a

certain proportion of poor residents sit on public housing policy boards or one reducing

funds for prison grants in favor o f adding funds to crime prevention activities, are coded

as hybrid legislation. Hybrid bills are neither purely symbolic nor purely substantive in

their public policy affects.

4.4 Results and Discussion

During the 104th Congress, most o f the bills introduced in the House did not

address a black policy agenda. Only 20 percent of the bills sponsored by legislators in

the 1996 NBES sample were coded as directly or indirectly benefiting blacks.4* Table 4.1

shows the likelihood o f House members in the sample to introduce at least one black

interest bill. Nearly 82 percent of black representatives introduced at least one black

interest bill during the 104th Congress compared to a little more than one quarter o f the

white legislators. Democrats were more than twice as likely as Republicans to introduce

at least one policy measure related to black interests. Blacks and Democrats were also

considerably more likely than then racial and partisan counterparts to sponsor

substantive, symbolic, and hybrid bills that directly or indirectly benefited blacks as well.

45 The bill’s chief aim is not to benefit American blacks although it is substantive in its effects for citizens of African nations. For that reason, it is coded as symbolically addressing American blacks* interest in and support o f black people and people o f color internationally- ^ *• 46 Canon (1999) reports .sim ilar results for the bill sponsorship in the 103rd Congress (1993-94). Most o f the bills sponsored during the Congress were non-racial in focus. 83 - y: .

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. TABLE 4.1 ABOUT HERE

As Table 42 shows, race and party were extremely important factors in

determining who sponsored black interest bills during the 104th Congress. Black House

members sponsored 82.6 percent o f the black interest bills in the sample and Democrats

sponsored 97.4 percent. The focus o f the majority o f black interest bills was symbolic or

a hybrid mix o f symbolic and substantive. Black members sponsored 158 pieces o f black

interest legislation, the lion’s share of which were symbolic or hybrid bills. White

members introduced substantially fewer black interest bQIs, but half of those were

substantive. Republican House members sponsored only five black interest bills, two

each were symbolic and hybrid.

The majority of bills benefiting blacks did not directly name blacks as the

beneficiaries. The partisan and racial breakdowns in Table 4.2 show that no group

sponsored more direct bills than indirect. Slightly more than a quarter of the black

interest bills directly named blacks. The remaining black interest bills did not name

blacks specifically, but did reflect key elements of the black policy agenda. These

indirect bills targeted disadvantaged populations, the poor, and inner cities, or they aimed

to more generally address housing and job discrimination, healthcare access, or

improving education.

TABLE S 4 2 and 4.3 ABOUT HERE

Table 4 3 reports the percent of their legislative portfolios that legislators devoted

symbolic, substantive, and hybrid black interests bills by House member race and party.

While 22.6 percent o f the black interest bills sponsored by the NBES sample legislators

were substantive in focus, only 17.1 percent o f the black interest bills sponsored by black

84

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. legislators were substantive compared to 50 percent o f the black interest bills sponsored

by white lawmakers. The majority o f black representatives total bill introductions were

symbolic or hybrid, while bill introductions by type for white and Republican members

was more evenly distributed (though the numbers were considerably smaller).

TABLE 4.4 ABOUT HERE

The descriptive statistics presented in Tables 4.1-43 strongly suggest that the race

and party o f the member will be important determinants o f who sponsors legislation that

is beneficial to blacks. Holding all other factors constant, including political party, does

the members race still predict a disproportionate share o f black interest bill sponsorship?

Table 4.4 shows an Ordinary Least Squares regression analysis of black interest bill

sponsorship by the type o f bill: all black interest bills, those that directly benefit blacks,

and those that indirectly benefit blacks.47

The race of the legislatoris the only variable that consistently reaches statistical

significance in the models predicting bill sponsorship that directly and indirectly benefits

blacks. Being a black legislator increases the percentage o f individual legislative agendas

devoted to black interest bills by a whopping 28 percent- Black representatives introduce

three percent more indirect black interest bills and one percent more direct black interest

47 To maintain the theoretical integrity o f the models, I primarily report the results from the model that includes both the racial composition of the district and the race o f the member as I theorize that both of these variables should affect black interest bill sponsorship- In the rare cases where no variables reach statistical significance in the standard model, I report the results of the amended model that does not include the district’s racial composition. With one exception (the sign for party in the model of indirect black interest bill sponsorship), the size and direction o f the coefficients remains the same across the standard and amended models. 85

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. bills than white colleagues of either party.4* Neither the party of the member, nor the

racial composition of the district reached statistical significance. In addition, the size o f

them coefficients was extremely small compared to the race o f the member and the signs

for both of the variables were contrary to my expectations for bill sponsorship that

indirectly benefits blacks in the standard modeL A reasonable amount o f variance is

explained by each o f the models, although the model predicting direct black interest bill

sponsorship does not perform as well as the others.

TABLE 4.5 ABOUT HERE

The analysis of the percentage o f members7 total bill introductions that focused on

symbolic and substantive black interest bills is presented in Table 4.5. Once again, only

the race of the member reaches statistical significance. Black legislators devote 163

times more o f their legislative portfolios to bills that champion symbolic black interests

than white colleagues. Eleven times more of the policy proposals of black

representatives reflect a hybrid focus on black interests than those of others in the

legislature. The signs for member party affiliation and the racial composition in the

district are in the expected direction with the exception of the negative, though not

statistically significant, relationship between Democratic Party membership and hybrid

black interest bill sponsorship. A reasonable amount of the variance in the models of

symbolic and hybrid black interest bill sponsorship is explained. The model of

substantive black interest bill sponsorship is unable to explain any of the variance in the

dependent variable and not a single independent variable reaches statistical significance.

* No variables reached statistical significance in the standard model o f direct black interest bill sponsorship, so results are reported from the amended model. 86

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Apparently, the variables that explain substantive bill sponsorship for blacks are qrrite

different from those that explain other types ofblack interest bill sponsorship.49

The results show that the race o f the member is an important predictor of black

representation. Black representatives are more likely to sponsor legislation that “speaks”

to black concerns than are then colleagues. The majority ofblack interest bills sponsored

by black legislators are symbolic in focus, which may have some implications for the

policy expectations o f African Americans. While black legislators clearly prioritize black

interests higher than other members, their emphasis appears to be one more of symbols

than of substance. White members appear to be more attuned to addressing specific

policy issues rather than broad African American concerns when the introduce black

interest bills. Democratic members of Congress are considerably more likely than

Republicans to sponsor black interest legislation, but the models failed to reject the null

hypothesis that no difference exists between the two parties in terms o f the proportion of

their legislative portfolios that they devote to black policy concerns.

The results o f the brvariate and multivariate analysis indicate that legislators, both

black and white, take a de-racialized approach to black interest bill sponsorship. The

majority o f black interest policy proposals do not directly name blacks as beneficiaries.

Black interest bills that indirectly benefit blacks outnumber those that directly benefit

49 One explanation for the failure o f the substantive bill sponsorship model and for the unexpected null results for the party variable is that party simply masks an ideological component to bill sponsorship. To check for this possibility, I ran each o f the models substituting Poole-Rosenthai Nominate scores for the party variable and, separately, including both party and ideology. The inclusion of member ideology did not add explanatory power to the models, nor did it achieve statistical significance. It also changed the sign o f the party variable in some models. The race of the member maintained its robustness as an explanatory variable in each model. 87

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. them by nearly 3 to 1 regardless o f the race of the legislator. To the extent that black

interest advocacy exists through bin sponsorship, this advocacy is most often broadly

targeted to benefit multiple groups, with blacks being one o f many, including other racial

and ethnic minorities, women, the poor, and the disadvantaged.

The findings presented in this chapter comport nicely with Canon’s (1999)

conclusion that black representatives engage in a “balanced approach” to group

advocacy. Their legislative agendas are more likely to include black interest policy

proposals and larger proportions of the proposals they sponsor are reflective of black

interests, but they are also quite likely to sponsor bills that address the policy needs of

other communities, including whites and other minority groups, as well. White

legislators in districts that are 15 percent or more black are considerably less likely to

take a balanced approach to representing the policy needs of both black and white

constituents, devoting only small portions o f the overall legislative portfolios to the

concerns ofblack Americans.

4.5 Conclusion

The theories and data presented in this paper had the objective o f linking

everyday legislative activities to the representation of interests, particularly black

interests. To some extent, that goal has been accomplished. I find evidence that some

members of Congress actively seek to use bill sponsorship to represent the interests of

blacks. Most of this advocacy is indirect and symbolic in nature and occurs largely

through efforts ofblack representatives. The reasons for this are manifold.

88

fUk&iH t-2. i ? 'IfjT • i ■: i ■* s^ySt’Li-! Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. First, Black legislators may be more likely to sponsor black interest bills because

of shared experiences with other blacks. This “linked-fate” theory of black

representation is supported by research showing that blacks as a group support

government intervention to address economic and social disparities that disadvantage

large numbers of blacks despite growing differences in socioeconomic status (Dawson

1994, Tate 1994).* Whitby (1997, 2002) characterizes this type of representation as

"spontaneous” because policies that are good for the black population are also good for

black representatives. Thus black interest representation comes “naturally” to black

legislators because they are essentially representing themselves.

An alternative explanation is supported in research at the state level that finds

evidence of a “transference phenomena” in black interest bill sponsorship over time

(Bratton 2002). As the percentage of blacks in the legislature increases, white

representatives sponsor fewer black interests bills and black representatives sponsor

more. In essence, white representatives transfer the primary responsibility of black

interest bill sponsorship to black legislators and devote a smaller portion o f then: bills to

black interest advocacy, while black legislators increase their bill sponsorship and as their

numbers increase, widely share the responsibility among other blacks in the legislature.31

* Tate (1994) finds some evidence of class differences in policy preferences among blacks, but also finds that these differences are mitigated by education and race identification, so that highly educated blacks and those with strong race identification generally expressed support for liberal policies. 31 Bratton (2002) also finds evidence of oppositional bill sponsorship behavior from Republican representatives as the percentage o f blacks m the legislature increases. Over time, she finds that Democrats sponsor fewer bills that are contrary to black interests, while Republicans sponsor more. This line of inquiry is beyond the scope of this research, but is an important question for future research on black representation in Congress. 89

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The reasons why the majority ofblack interest bill sponsorship is symbolic and

indirect may also lie in the very nature o f the policymaking system. Astute politicians

understand that the policy process in the House is difficult terrain for bills to survive.

The use o f legislation as a communication tool with constituents is a relatively easy and

time-efficient means of responding to constituency interests and reinforcing the

perception that their representative is “one of them.” Because individual legislators can

claim successful policy outcomes as the result of their own hard work and attribute failed

policy to a wide host of factors beyond their control, introducing legislation can often be

a win-win situation for the initiator regardless of whether real tangible benefits are

awarded. As David Mayhew (1974: 31) notes, “the electoral goal is achieved first and

last, not in Washington, but at home.” Representation through legislation can be a

reasonably effective strategy for some representatives.

It is quite possible that symbolic legislation amounts, on occasion, to “cheap talk”

where representatives make promises they have no intention of fulfilling and, in effect,

mislead constituents about their true intentions to “represent” their interests. According

to game theorists, because cheap talk is unverifiable and cannot be falsified, rational

actors usually discount its usefulness as a measure o f other actors’ true intentions (Bianco

1994). A cheap talk account of representation through legislation fits well with

Mayhew’s (1974,63) conjecture (noted earlier in the chapter) that constituents probably

give this activity “about the value [it] deserves.” However, what game theorists

characterize as cheap talk, may not actually be so “cheap” after all. Symbolic

representation is a means for political actors to change the dimensions o f debate, critique

the status quo, and offer alternative ways o f viewing political possibilities.

90

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Mayhew’s (2000) recent scholarship on the political legacy o f the U.S. Congress

takes a different perspective on members’ activities than his earlier work, arguing that

some tremendously important and lasting political ramifications cannot be understood

properly if scholarship on the activities o f members o f Congress largely consists of

accounts o f its lawmaking function. Many types of non-legislative activities, including

talk, can have a rather large impact on Congress and on the nation. Johnson (1983,81-

82), argues that “speech acts” themselves (the communication of ideas, preferences, and

strategies between actors) have “binding force” that shape actors expectations. Johnson

(2000, 409) further suggests that symbolic actions gain traction as important forms of

communication, not because they convey detailed messages, but rather because they

present alternative views o f political reality and future possibilities. The importance of

symbols in political discourse lies in the power o f symbolic acts to help define social and

political possibilities for group members (Johnson 1993, 410). In this conceptualization,

the communication of symbols can “help reveal options and identities that might go

otherwise unconsidered give palpable existence to as yet unrealized possibility...” and

help actors “redefine their options and identities” (Johnson 1993,410). Thus, symbolic

policy proposals like Representative Conyers’ (D-MI) black slavery reparations bill,

Representative Lewis’ (D-GA) bill to protect the right to vote for the homeless, or

Representative Dellums (D-CA) bill to establish a living wage and full employment for

all Americans are not cheap talk, but rather serve as vehicles to redefine the political

options for their key constituents.

It is still very likely that all constituents place high value on receiving substantive

benefits in their communities. As Whitby (2002, fh. 8) notes, “a strong argument can be

91

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. made that substantive bills are more important [than symbolic bills] because they have a

greater effect on the livelihood o f individuals.” In a choice between a symbolic benefit

and a tangible one, the tangible one would likely win if voters could chose. This only

makes sense. In fact, symbolic benefits may well be perceived as a “means’* to a tangible

“end”—eventually. When symbols come, one expects that substance will follow.

However, in a choice between symbolic benefits and no benefits at all, symbolic benefits

would likely be preferred. Thus, symbolic representation is not unimportant by any

stretch.

Does this emphasis on symbolic representation mean that black constituents are

not being well represented? The data presented here do not allow for a conclusive

answer to this normative question. The black members in this study were more likely to

introduce black interest bills. They were slightly less likely to introduce substantive

black interest legislation than they were to introduce symbolic legislation, but more likely

than their white counterparts to introduce symbolic legislation representing black

interests. However, sponsoring good bills that go nowhere will not alleviate the

problems that disproportionately affect the black community. That said, we must recall

that symbols do matter, just as the Lincoln Memorial matters to the American public or

the Harold Washington Memorial Librarymatters to black residents in Chicago.

92

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Black Members White Members Democrats Republicans Number o f Members 33 63 71 25 Type o f B ill Black Interest (A ll) 81.8% 27.0% 54.9 % 20.0% (27) (17) (39) (5) Direct 48.5 % 12.7% 31.0% 8.0% m (8 ) (22) (2) Indirect 81.8% 17.5 % 4 9 3 % 8.0% (27) ( I t ) (16) (2)

Substantive 30.3 % 12.7% 22.5 % 8.0% (10) (8) (16) (2) Direct 9.1 % 4.8% 7.0% 4.0% (3) (3) (5) (1) Indirect 3 0 3 % 7.9 % 19.7% 4.0% (10) (5) (14) (1)

Symbolic 66.6% 11.1% 65.1 % 4.0% (22) (7) (28) (1) Direct 48.5% 6.4% 26.8% 4.0% (16) (4) (19) (1) Indirect 45.6% 4.8% 25.4 % 0.0% (15) (3) (18) (0)

Hybrid 66.6% 9.5 % 36.6% 8.0% (22) (6) (26) (2) Direct 6.0% 0.0% 2.8% 0.0% (2) (0) (2) (0) Indirect 63.6 9.5% 3 5 3 % 8.0% (21) («) (25) (2) _ . Note: The table presents the percentage of members in each category that sponsored one or more of each type o f bilL The number of members is in parentheses. The percentages should not sum to 100 because biD sponsorship in each category is independent o f bill sponsorship in any other category.

Table 4J Sponsors of Black Interest Bills (One or More) by Member Race and Party, 104th Congress

93

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Num oer o f Black nterest Bills by Type Total Direct Indirect Sym bolic Substantive Hybrid All Members 190 52 138 73 43 74 Race Black 158 43 115 65 27 66 White 32 9 23 8 16 8 Party Democrats 185 50 135 72 41 72 Republicans 5 2 3 I 2 2 Note; The sample included 33 black representatives and 63 whites; 71 Democrats and 25 Republicans.

Table 4 2 Number of Black Interest Bills Sponsored by Member Race and Party, 104th Congress

Members by Race Members by Party BOl Type A ll Black White Democrats Republicans Members Members

Substantive 22.6% 17.1 % 48.5 % 222% 20% (43) (27) (16) (41) (1)

Symbolic 38.4% 41.1 % 24.2 % 38.9 % 40% (73) (65) (8) (72) (2 )

Hybrid 40.0% 41.8% 24.2 % 38.9% 40% (74) (66) (8) (72) (2)

Totals 100% 100% 100% 100% 100% (190) (158) (33) (185) (5)

Note: Number of bills in parentheses.

T able 4 3 Types of Bills Introduced as a Percentage of All Black Interest Bills Sponsored by Member Race and Party, 104th Congress

94

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Explanatory Types o f Bills Sponsored Variables A!I Direct Indirect b b b b b b (SE) (SE) (SE) (SE) (SE) (SE) Representative Characteristics Race 2 8 .2 * 34.8” 0.67 039** 3.0* "2.9” (9.7) (4.6) (0.66) (3 1 ) (-02) (0.59) Party 1.9 2.4 0.0 0.0 -0.1 .01 (52) (5.0) (0 3 5 ) (0 3 5 ) (.68) (0.67) Seniority 0 2 .14 -0.0 -0.0 0.0 0.0 (02) (0.24) (0.02) (0.02) (0.03) (0 3 1 ) District Characteristics % Black 0 2 --- 0.0 --- -0.0 — (025) (0.02) (.03) South -4 2 -4 5 -0.0 . -0.0 -1.4 -1.04 (5.0) (5.1) (0 3 4 ) (0 3 4 ) (0.65) (0.65) Urban 0.0 0.0 0 0 0.0 0.0 0.0 (0.12) (0.12) (0.01) (0.01) (0.02) (0.02)

Constant 15 6.1 -0 3 6 -0 3 4 0.69 0.64 (12.4} 0 0 .9 ) k (0.83) (0.64) (1-6} (0.65) F-statistic 153*’ 1 8 3 3.6* 4 3 7.6 9 22 Adjusted R* .48 .49 .14 .15 29 3 0 Number of cases 93* 93 93 93 93 93 Note: The dependent variable is the percentage of bills sponsored by a House member in each category. “All” is the percent ofblack interest bills sponsored by each member; “Direct” is the percent ofblack interest bills sponsored by each m em ber that directly benefits blacks; and “Indirect” is the percent o f bills sponsored by each member that indirectly benefits blacks. Each category is divided by the total number of bills sponsored by each member. Three members did not introduce a single piece of legislation in the 104*, Representatives Sisisky (D- VA)r Bishop (D-GA), and Ford, Sr. (D-TN) and are not included in the regression analysis. Representative Melvin Watt’s legislative portfolio for the 104“* consisted entirely of amendments to bills from the Republican Contract with America. **p< .01 level * p<.05 level * p<.10 level (one-tailed test)

Table 4.4 Ordinary Least Squares Regression Model of Direct and Indirect Black Interest Bills, 104th Congress

95

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Explanatory Types o f Bills Sponsored Variables Symbolic Substantive Hybrid b b B b b B (SE) (SE) (SE) (SE) (SE) (SE) Representative Characteristics Race 163* 19.0" 1.1 0.53 10.8’ 163** (6.9) (3 3 ) (4.8) (2 3 ) (5.5) (2.6) Party 2.4 2.6 2.4 2.4 -3.1 -2.7 (3.7) (3.6) (2.6) (2.6) (2.9) (2.9) Seniority -0.0 -0.10 0.0 0.0 0 3 0 0.19 I (0.17) (-17) (0.12) (0.12) (0.14) (0.14) District Characteristics % Black 0.0 — -0.0 — 0.15 — (0.17) (0.12) (0.14) South -0.73 -0.84 -3.1 -3.1 0.16 -0.0 (3.6) (3.6) (2.5) (2.5) (2.9) (2.9) Urban -0.0 -0.0 -0.0 -0.0 0.0 0.1 (0.84) (0.08) (0.06) (0.06) (0.07) (0.07)

Constant 43 6.1 4.7 4 3 5 -8.0 -4 3 (5.4) (8.8) (7-

Table 4.5 Ordinary Least Squares Regression Model of Symbolic and Substantive Black Interest Bills, 104lb Congress

96

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTERS

BLACK SUPPORT FOR SUBSTANTIVE AND SYMBOLIC BILLS

In Chapter 4 ,1 found that the race o f the member, rather than party affiliation,

explained most of the variation in bill sponsorship that was beneficial to blacks. I also

found that half or more of the black interest bills introduced by black and white

Democrats were symbolic. This chapter now considers whether black constituents value

symbolic legislation and whether the type o f representative they have— descriptive or

partisan — affects their support of symbolic policy proposals. Using a unique data

source, I determine how black citizens prioritize legislative initiatives that have direct

implications for the black community. This chapter constitutes the first examination of

constituent attitudes toward symbolic versus substantive legislation.

Survey work dating from the post-civil rights era to the present reveals that

African Americans are extremely liberal in their policy positions (Nie, Verba, and

Petrorik 1979; Gurin, Hatchett, and Jackson 1989; Dawson 1994; Tate 1994). The

politically conservative Reagan years of the 1980s, however, witnessed a small but

consistent conservative trend in black political attitudes. For example, whereas a large

majority o f blacks favored government assistance to blacks and minority groups in the

1960s, today only a plurality expresses such support (Tate and Hampton 2000). hr the

1996 NBES, blacks were divided over the issue of welfare reform with 60 percent 97

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. expressing support for limiting welfare to five years over tfae'recipient’s lifetime. As

noted m Chapter 2, the increasing numbers o f blacks expressing conservative viewpoints

has made the matter o f discerning a “black political agenda” all the more problematic.

In this chapter, 1 examine black preferences for symbolic and substantive policy

proposals using data from the 1996 National Black Election Study. I developed a series

of questions for the survey aimed at gauging black support for symbolic and

substantive.52 Respondents were asked about a series of policies related to the needs of

black and minority communities. The results suggest that although general support for

black interest policies crosses socioeconomic boundaries, symbolic representation

resonates with certain blacks more so than others.

5.1 Analyzing Black Policy and Representation Preferences

Black policy preferences are conceivably influenced by several factors. Using

data from the 1984 and 1988 National Black Election Study, Tate (1994) constructed a

model that examines the impact of ideological and party identification, demographic

differences, personal economic circumstances, subjective class identity, and race identity

on black support for the liberal Democratic Party agenda. When she compared the effects

of social class, race identification, ideological identification and party identification on

black policy preferences, she found that, all else being equal, race identification “had the

most consistent effect on black's attitudes toward social services and race-specific

521 am indebted to Katherine Tate, the principal investigator for the 1996 NBES for including this series o f questions in her survey. 98

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. programs’1 (Tate 1994, 40). Ideological and party identification had modest to strong

effects across black policy preferences, while social class had only a weak effect.

Variables measuring socioeconomic and demographic differences had mixed

effects on black preferences as well. Tate (1994) also found evidence o f class divisions

in black policy preferences related to education and income. On four o f the seven policy

issues directly relevant to black interests (guaranteed jobs, jobs for the unemployed, food

stamps, Medicare, crime spending, public schools, and minority aid) higher education

was inversely related to positive preferences. In the three policy areas where education

reached statistical significance, higher educated blacks expressed less support for food

stamps and Medicare programs than other blacks and somewhat more support for public

schools.

Income was also negatively related to support for most of the black interest

policies, though its effect was more often statistically significant. Higher income blacks

expressed positive preferences for black interest policies only on spending for crime

prevention and public schools. Despite finding evidence of socioeconomic motivated

differences in policy preferences, Tate notes that race identification had an indirect

liberalizing effect on education that mitigates its overall trend toward conservative policy

attitudes. She also notes that ideological and partisan ties influenced black preferences

across more issues than did race identification.

Region may also shape blacks policy preferences. Also using survey data from the

1984 and 1988 NBES, Gurin, Hatchett, and Jackson (1989) found that region had a

consistent, though modest effect on several black poKcy preferences. Non-southem

blacks expressed greater support for government spending on food stamps, crime, and for

99

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. a change in U.S. policy toward South Africa than southern blacks (Gurin, Hatchett, and

Jackson 1989,96).

For all that the preceding studies reveal about black policy preferences, they

reveal nothing about black preferences for representation styles or behavior. The rest of

the chapter considers the following questions: Do blacks prefer that then: representatives

pay more attention to national concerns or to district needs? If they could determine how

much time them representatives devoted to certain kinds of legislative proposals, what

arrangement would they prefer? Do blacks of certain income or educational levels, or

those who feel very strong ties to the fate o f the larger black community prioritize

legislators7 time differently? And finally, what does the way blacks want them

representatives to allocate time reveal about their preferences for symbolic and

substantive representation?

5.1 J Black Assessments of the Value of Legislative Activities

Very few studies have asked blacks what tasks they value that legislators do.

When they have asked, some instructive priorities emerge. Fast, blacks highly value

attention to district needs over national needs (Tate 2002). In a question asking whether

blacks wanted their representatives to concentrate on district or national concerns, 46

percent o f blacks in the 1996 NBES wanted representatives to focus them time on

bringing resources to the district while 33 percent wanted helping constituents to be their

representative’s first priority. Only one-fifth o f blacks wanted their representative to

prioritize national concerns over local ones.

100

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. This recent finding substantiates findings from two decades ago when black and

white preferences for delegate or trustee representational styles were examined in The

Personal Vote (Cain, Ferejohn, and Fiorina 1987). Using data from a 1978 NES/CPS

survey, the authors (1987,42) point out that “racial differences in representative priorities

remain” even after taking class, educational, and partisan differences into account. The

authors conclude that blacks placed a higher priority on legislators helping people in the

district with personal problems than any other group. The foremost role of legislators

according black preferences was to protect the interests of the district and to act as a

constituency servant, responding to needs of individual constituents (Cain, et. al., 42).

Blacks regarded legislators work in Congress on bills concerning national issues as the

least important activity representatives could engage in, behind keeping in touch with the

district, bureaucratic oversight, district advocacy, and constituency service. That blacks

apparently would prefer legislators to deliver pork-barrel benefits to the district and

consistently respond to constituents’ individual problems rather than focus on a national

agenda poses some interesting questions for future research on black representation that

are beyond the scope of this project. However, extant research indicates that blacks

prefer representation that directly reflects their needs rather than national priorities and

that blacks have expectations for the representative-constituent relationship that differ

considerably from whites’ expectations.

The survey questions that I developed in the 1996 NBES address black support

for symbolic and substantive policy proposals in terms of how blacks thought

representatives should allocate their time. The questions asked respondents “fh]ow

much time do you think black House representatives should spend working on the

101

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. following policy proposals? None, Some; or Most of the their time?” The she proposals

consisted of the following: 1) a bill to equalize funding for inner-city and suburban

public schools, 2) a ban on alcohol and cigarette billboard ads in inner-city communities,

3) a Constitutional amendment protecting affirmative action, 4) a bill to create an African

American history museum on the Mail in Washington, D .C , 5) a bill banning the display

o f the Confederate flag in public buildings, and 6) a bill acknowledging the fundamental

injustice o f enslaving American blacks. The bill topics were pulled directly from debates

in the black community about government action in key policy areas that are highly

relevant to black, minority, and poor communities. So although these items are

hypothetical, the policy proposals analyzed in this chapter could reasonably have been

bills actually considered or sponsored by House members within the last ten years.

The results of a bivariate analysis ofblack preferences for legislative activity on

the bills are shown in Table 5.1. Proposals for school funding banning alcohol and

cigarette billboards, and affirmative action are coded as substantive because, if enacted,

they would affect the lives of blacks in a tangible way. Each of these three proposals

would directly affect blacks' social and economic well-being. Proposals for a black

history museum, a ban on the Confederate flag, and an acknowledgment of slavery’s

injustice are coded as symbolic because they would not result in tangible social or

economic outcomes for African Americans even if they were enacted.

INSERT TABLE 5.1 ABOUT HERE

There is some indication that blacks intuitively distinguish between symbolic and

substantive policies and prioritize them differently. In general, the symbolic policy

proposals earned less support among blacks than the substantive ones, although

102

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. majorities expressed the view that members should spend,jimosft or “some” o f their time

pursuing symbolic bills. For example^ a majority.of respondents, 58 percent, wanted

black representatives to spend “some”' time on creating a national black history museum

and an additional 35 percent wanted them to devote most o f their time to this effort.

While nearly one-quarter of respondents said that members shouldn’t spend any thne

legislating a ban on the display o f the Confederate flag in pubHc buildings, 43 percent

said that members should devote “some” of their time to the issue, and 31 percent felt

that members should allocate “most” of their time to it. Slightly higher percentages of

blacks felt that members should spend time on a resolution acknowledging the

fundamental injustice of slavery. Sixteen percent felt that members should devote no

time to this matter.

While symbolic bills were enthusiastically supported by meaningful pluralities in

the black community, substantive bills received even higher margins o f support. A solid

majority o f respondents, 64 percent, wanted black legislators to spend “most” of their

time on remedying funding inequities between suburban and urban schools, while only 2

percent wanted them to spend no time on the issue. More than half of the respondents

wanted representatives to spend “most” of their time on securing affirmative action

protections through a constitutional amendment while 4 percent did not want this issue

prioritized at all. And, a plurality o f blacks want legislators to spend “most” o f their time

on banning unhealthy advertisements on billboards in black communities.

Blacks prioritize substantive policies over symbolic ones, but still highly value

symbolic proposals. For example, over 90 percent of black respondents wanted

legislators to spend “some” or “most” o f their time on two proposals, one substantive and

103

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. one symbolic: 93 percent supported creating a black history museum and 98 percent

supported equalizing public school funding. Overall, blacks were generally supportive o f

legislative activity promoting black interests. Only a small minority, at most a little more

than one-quarter o f blacks found black interest issues so irrelevant that black legislators

should spend no time at all on these concerns. How do these initial findings relate to the

literature on members' preferences for time allocation?

5.1.2 Legislator Expectations and Representation Strategies

A 1977 survey by the House Commission on Administrative Review asked House

members to describe their perceptions o f their job expectations. Davidson and Ofeszek

(2000, 130) report the findings o f this survey as a “snapshot of members' views o f their

jobs." The formal aspects o f their legislative roles—bill sponsorship and other legislative

work, bureaucratic oversight, and committee specialization—drew the top-rank from the

majority o f the sample. Behaving as constituency servants by being attentive to casework

and bringing resources to the district ranked second. Staying in touch with the district by

a variety o f means including mailings and town-meetings, as well as ensuring that district

concerns were represented in Washington, D.C., rounded out the top four duties that

House members thought others expected o f them.

What legislators think about how others want them to spend their time and how

they actually allocate this precious resource are somewhat different. Drawing on a 1993

report issued by the Joint Committee on the Organization of Congress, Davidson and

Oleszek (2000) point out the tradeoffs that legislators make in terms o f how they would

ideally spend their tune and what they actually do. While members would like to spend

104

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. more time on strictly formal legislative duties like studying pending legislation, following

floor debates, and attending committee hearings or markups: they actually spend most o f

then: time meeting with constituents- The authors point out that time allocation is very

much reflective o f how legislators manage conflicting role expectations between purely

institutional demands on the one hand, and constituent demands on the other.

In addition to the demands placed on all elected officials, black representatives

also have additional demands placed on them by a constituency that expands beyond

geographical district borders. While expanded constituencies may not be solely the

domain o f black legislators, representing a national constituency in addition to a district

constituency is a task that most black legislators claim as part of their mission (Clay

1993, Bamett 1982). Researchers have debated about the extent to which the

Congressional Black Caucus has achieved this goal of universal representation ofblack

interests. Singh (1998) concludes that despite the extensive coDective effort that CBC

members devote to advocating black interest, the result has meant only modest gains for

blacks in terms o f public policies.

The disparity between how legislators would like to spend their time and their

perceptions about constituent expectations for time allocation motivates legislators to find

creative ways to represent the interests of their constituents. In the previous chapter, I

argued that representatives use bill sponsorship as a representation tool to indicate to

constituents, “I care about this issue too.” I found that black representatives are more

likely to sponsor both substantive and symbolic black interest bills than their white

colleagues and that white Democrats are considerably more likely than white

Republicans to sponsor bills reflective ofblack concerns. It is probable that legislators

105

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. consider which constituents in the district will be most responsive to the bills they

sponsor and advertise their efforts and accomplishments accordingly. The task of

advertising and position-taking is likely easiest in homogenous districts where there is

less uncertainty about “shared” group preferences. Using bill sponsorship as a

representational tool in racially heterogeneous districts may be trickier. In either case,

however, representatives cannot be certain about what segment o f their constituency is

responding positively to them bill sponsorship efforts.

Here, I test whether certain segments of the black community respond to black

! interest policy proposals differently. Do poorer, less educated blacks prioritize policies

differently than wealthier, more educated blacks? Are blacks that feel very connected to

the black community more supportive o f policy proposals across the six issues described

above than blacks that feel less connected, regardless of the symbolic or substantive

nature of the bills? Also, do blacks with descriptive representatives prioritize issues

differently than blacks with white representatives? The structure o f the survey instrument

presents some limitations for answering these questions, however examining the policy

priorities o f blacks in this manner allows us to evaluate the concept of “representation

through legislation” from the perspective o f the potential recipients o f legislative benefits.

The limits posed by question wording and the research design are worth noting,

but do not swamp the usefulness o f the measures. First, the fact that the questions are not

open-ended means that rather than blacks generating a list o f the top three or four policy

issues they think legislators should focus on. they must select from a fixed list that may

not be immediately reflective o f their own priorities. Yet, asking and coding open-ended

questions could be cost-prohibitive and might not gain us that much more ground since

106 i i

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. the survey questions are grounded in real policy debates and thus likely reflect issues

blacks are acquainted with.

Second, because respondents were not asked to rank order their preferences we

are unclear on how blacks would order their preferences in comparison to each other. For

instance, respondents were not asked whether black legislators should spend more time

on creating a black museum than on a campaign against debilitating billboard ads in poor

neighborhoods. Such a question would allow us to directly compare black preferences

between symbolic and substantive proposals rather than resort to inferences. Conversely,

respondents may have imposed a rank order on the proposals as they answered questions

even though they were not instructed to do so. Respondents may have been unwilling to

answer that black legislators should spendmost o ftheir time on every single issue, even i f

every issue was a personal priority.

Third, asking the question only about the way black legislators should prioritize

their time leaves us unable to evaluate what blacks think white legislators should do. Do

blacks believe that black interest proposals should be priorities o f the legislators who

represent them regardless o f the legislator’s race? Finally, and most difficult, these

proposals are not easily divided into symbolic and substantive categories. Still, in the end,

these questions allow us to get at the question as to whether representatives are

“speaking” to certain sub-populations of their constituents when they pursue certain

policies.

107

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 5.2 Bata and Methods a' As noted previously in this chapter, the data tin Black preferences for symbolic or

substantive policies come from survey questions I developed for the 1996 National Black

Election Study. The analysis considers four sets o f factors: I) socioeconomic indicators

(respondents* educational attainment and yearly income); 2) measures of respondents*

self-identification along a 4-point race identification scale” (l=weak, 4=strong), and

seven-point ideology (l=strong conservative, 7=strong liberal) and party identification

scales (l=strong Republican, 7=strong Democrat); 3) district characteristics including

region (G=non-south, l=south), median income (district median income as reported by

the 1990 Census), and black percentage of the population (district black population as

reported by the 1990 Census), and finally, 4) the type of representation respondents

receive defined as partisan (0=if represented by a member o f a different party from their

own, 1= if represented by a member from their own party) or descriptive (0=white

representative, l=black representative). How will these factors influence blacks*

evaluations o f how black legislators should prioritize their time? Will certain segments

o f the black community be more responsive to symbolic rather than substantive policies?

Will support for individual policy priorities vary among certain groupings?

Because Tate (1994) finds education to have a mixed effect and income to have a

generally negative effect on black support of black interest policies, I expect high levels

of education and income to more often depress than elevate the likelihood of highly

53 The race-identification scale is a common measure o f Iinked-fate used in analyses o f group-consciousness and black politics (Dawson 1994, Tate 1994). Respondents in the 1996 NBES were asked how strongly they agreed or disagreed with the following statement, “What happens to black people in this country has a lot to do with what happens to me.” 108

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. prioritizing black interest policy proposals. To the extent that respondents can intuitively

distinguish between symbolic and substantive policies* I expect blacks with higher

socioeconomic status to be more supportive of substantive policy proposals than

symbolic ones.

Strong race identification has a liberalizing effect and thus is expected to

positively influence support for prioritizing black interest proposals across policy types*

whether symbolic or substantive. Because strong race identifiers express strong feelings

of commonly shared fate with other blacks* I expect them to highly prioritize black

interest proposals across issue areas as well. The expectation for liberals and Democrats

is not altogether clear. Unlike strong race identifiers, liberals and Democrats may be

more discriminating in regards to prioritizing proposals across policy types. Rather than

broadly supporting a wide array o f black interest proposals* liberals and Democrats may

pragmatically support policies that generally reflect the liberal Democratic agenda. They

may readily prioritize legislative activity in favor of affirmative action* an issue that

remains part o f the liberal agenda despite Clinton’s “mend it* don’t end it” admonition*

but not as readily prioritize more racially-directed proposals* like a proposal to

acknowledge the injustice o f slavery.

The type of representation that respondents have may influence their policy

priorities. Blacks with descriptive representatives may guard the time o f black legislators

more jealously than blacks with white representatives. By the same token* blacks with

descriptive representatives* because they most often live in districts with large black

populations* may be more responsive to black interests across policy types and issue

areas. Blacks represented by a member o f their own party may prioritize policy

109

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. proposals differently than blacks without partisan representation. District characteristics

are included as controls for the impact o f district poverty, racial makeup, and region on

black interest policy priorities.

53 Results and Discussion

Tables 5.2 and 53 examine black policy priorities controlling for the type of

representation and self-identification by race, ideology, and party. The type of

representation blacks receive appears to have implications for how they prioritize policy

proposals. Across issue areas and policy types, blacks with descriptive representatives

are more likely to highly prioritize black interest policies than are blacks with white or

partisan representatives. Statistically significant differences between blacks with

descriptive representatives and those without are evidenced in two substantive issues:

banning damaging alcohol and cigarette billboard ads in black communities, and

equalizing spending between urban and suburban schools. Priority differences between

blacks with partisan representatives and those^with representatives from an opposing

party occurs only on the issue o f banning the Confederate flag, and in a surprising turn o f

events, a larger proportion o f blacks with representatives from a party different than their

own highly prioritize activity in this area.

INSERT TABLES 52. AND 5 3 ABOUT HERE

Statistically significant differences among liberals and conservatives are

evidenced on affirmative action, creating a black history museum, and the slavery bill.

Democrats and liberals are more supportive than Republicans and conservatives

regarding activity to secure affirmative action benefits for blacks and women. Support

110

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. for the creation o f a black history museum crosses ideological boundaries, but liberals

and conservatives differ slightly on how black legislators should allocate their time,

likewise on the slavery recognition issue. Partisan differences are not statistically

significant with the exception o f affirmative action, where Democrats support giving this

issue the highest priority on black legislators7 agendas by more than 10 percentage points

compared to Republicans. Strong race identifiers express statistically significant

differences fiom low race identifiers regarding prioritizing legislative activity aimed at

banning public display o f the Confederate flag and for equalizing spending between

urban and suburban schools.

Overall, the bivariate analysis has not revealed major differences between groups

of blacks on any measure. When statistically significant differences occur, the

percentage differences are relatively small. While this may portend continued evidence

of common fate influences on black policy preferences, a multivariate will make the

relationships clearer by estimating the likelihood o f black preferences for legislative

priorities based on the individual impact of each independent variable while controlling

for the possible effects o f other variables.

I use the 1996 NBES to estimate ordered probit models for each of the six

hypothetical bills discussed earlier. The dependent variable is the priority that

respondents place on each one of the policy proposals. This variable is ordered fiom I—

indicating very low or no priority, to 3—indicating the highest priority.

The independent variables consist of individual level data on respondents7

education, income, gender, party affiliation, and race consciousness. Most important for

the purposes o f this analysis are the variables accounting for the type o f representation by

111

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. party and race. Cases where the respondent’s party affiliation or race was matched by

that o f the representative are coded as 1, all other cases are coded as 0. Finally, several

variables provide information on district context that should be controlled for because

they may influence support for certain policies, including region, the size of the black

population in the district, and district wealth.

Tables 5.4 and 5.5 present results o f an ordered probit model o f black support for

policy proposals coded as substantive and symbolic, respectively. The central question is

whether blacks with descriptive representatives have different policy priorities than

blacks with partisan representatives. Secondary questions relate to the impact of self-

identification and socioeconomic status on policy preferences. Overall, the models

indicate that the type o f representation blacks receive, their race, ideological, and political

identities, and their socioeconomic status affect black preferences for legislative priorities

in different ways.

INSERT TABLES 5.4 AND 5.5 ABOUT HERE

With one exception, having a black representative is related positively to the

probability o f blacks assigning greater priority across symbolic and substantive policy

types and across issues, although statistical significance is not reached on four of the six

proposals. Descriptive representation is statistically significant and positively related to

blacks highly prioritizing legislative work on a slavery acknowledgment bill and on

securing affirmative action benefits through a Constitutional amendment. Table 5.4

shows that, all else being equal, blacks with descriptive representatives were marginally

more likely to assign working on a Constitutional amendment for affirmative action the

highest priority on a black representatives’ legislative agendas. As shown in Table 5.5.

112

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. having a black representative also sKghtly increases the probability o f blacks assigning

the highest priority to a bill recognizing the injustice o f slavery. In comparison, having a

partisan representative is positively related to the probability that blacks will highly

prioritize substantive legislative initiatives and negatively related to the probability that

blacks highly prioritize symbolic ones. Blacks with partisan representatives are

somewhat less likely than blacks with representatives from a party different than then-

own to highly prioritize legislative activity to ban public display o f the Confederate flag.

With the exception of supporting a ban on the Confederate flag and (for income)

banning alcohol and cigarette ads, higher socioeconomic status is inversely related to

assigning black interest legislation the highest priority on black legislators' agendas,

through generally these variables do not reach statistical significance. Education reaches

conventional levels of statistical significance in determining whether blacks highly

prioritize banning alcohol and cigarette billboards in inner-city communities. Increasing

education from less than a high school education to college education decreases the

probability of highly prioritizing this issue by 20 percent. Education reaches marginal

levels o f statistical significance in determining black priorities on two symbolic bills: the

slavery bill and the African American museum bill. Moving education from its minimum

to its maximum decreases the probability o f blacks highly prioritizing legislative action

on these issues by .17 and .15, respectively. Household income is a statistically

significant predictor o f black support for highly prioritizing work on the museum bill and

the slavery bill, and marginally impacts the likelihood o f support for an affirmative action

bill. Increasing income from its minimum to its maximum decreases support for

prioritizing building an African American museum by .15 and for a national apology for

113

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. jj s - -

‘ *

1 jI slavery by 2.6. The probability of blacks highly prioritizing legislative work on J affirmative action drops by .16 as their incomes Increase-

Strong race identification is generally positively related to prioritizing black

interest legislation. Strong race identification is positively related to four of the she

policy proposals, although it fails to reach traditional levels of statistical significance.

Strong race identification reaches marginal statistical significance in blacks’ decisions to

highly prioritize legislative activity to ban public display of the Confederate flag.

Changing the strength of race identification fiom low to high results in a small, but

positive (.OS) increase in the probability that blacks want this issue to be a legislative

priority for black lawmakers.

In each instance where party identification provides a statistically significant

explanation o f policy priorities—the affirmative action amendment, slavery apology, and

banning public display o f the Confederate flag—it is positive; indicating that black

Democrats prioritize legislative action on these issues differently than do black

Republicans. Blacks who self-identify as Democrats have a .16 greater probability than

black Republicans of wanting black legislators to devote most o f their time to the task of

securing affirmative action benefits. The probability o f blacks expressing a preference

that banning the Confederate flag be the highest priority o f black legislators increases by

.12 as respondents move from strong Republicans to strong Democrats. Similarly, the j

j probability o f wanting legislators to spend most o f their time gaining national recognition

for the injustices of slavery increases by .14 as respondents more strongly identify

themselves as Democrats. Ideology attains marginal statistical significance in

determining black support for affirmative action. Among those who highly prioritize

114

n ..

*

•’Jir Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. securing affirmative action benefits for future generations, being a strong liberal increases

the probability by .10.

Variables controlling for district characteristics show that, despite attaining

statistical significance, the impact of district wealth on differences in black policy

priorities is not much different fiom zero. The policy priorities of southern blacks were

not statistically different fiom those of non-southem blacks. The relative size of the

black population in the district increased the probability that blacks would highly

prioritize legislative action on affirmative action, school funding, and national

recognition of the injustices of slavery. Black females were marginally more likely than

black males to highly prioritize legislative action on affirmative action and a ban on

certain billboards.

5.4 Black Preferences for Symbolic and Substantive Policy Proposals

Ultimately, it appears that the type of representation that blacks receive has

implications for how they prioritize black interest policy proposals. Blacks with

descriptive representatives respond more favorably to both substantive and symbolic ' r policy proposals than blacks with white representatives. In comparison, partisan

representation appears to make little difference in how blacks prioritize black interest

policies beyond depressing theprobability o f blacks assigning symbolic policies high

legislative priority.

There is no evidence that blacks with higher socioeconomic status are more

supportive of substantive black interest policy proposals than symbolic ones. To the

contrary, it seems that better educated and economically well-off blacks are simply less

115

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. inclined to highly prioritize black interest proposals regardless o f the policy type or issue

area. Even on an issue where socially and economically better-off blacks are likely to

benefit most—affirmative action—I find a decreasing probability o f highly prioritizing

this issue on black IegisIatorsr agendas. This indicates that class divisions are relevant to

black preferences for legislative priorities not only when the interests involved are mainly

reflective of the needs of poorer blacks (e.g., the billboard ban and public school

funding), but also on any issues that are directly related to race-specific concerns (e.g.,

the African American history museum.) Perhaps, class interests redirect the policy

priorities o f better-off blacks to economically focused issues rather than race-specific

ones.

None of the self-identification measures performed especially well in explaining

black legislative priorities. Race identification did not have nearly the broad effect one

might have expected. Rather than being supportive of black interest proposals across a

range of policy types and issues, strong race identifiers were no different than weak race

identifiers in terms of how they prioritized most of the policy proposals. Apparently

strong race identifiers distinguish between symbolic and substantive policy proposals,

responding positively to substantive proposals and mostly negatively to symbolic

initiatives. Strong race identifiers were more likely to prioritize legislative activity

banning public display o f the Confederate flag, but not activity aimed at establishing a

museum honoring African American achievements or gaming a national apology for the

injustices of slavery. Blacks who espoused strong feelings of common fate (fid not

blindly support black interest policies, but rather discriminated among them, placing

higher priority on some and less on others.

116

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Party affiliation and ideological identification do not consistently relate to black

preferences for symbolic or substantive bills or particular policy issues. Compared to

black Republicans, black Democrats appear to highly prioritize issues that conform to a

strictly liberal agenda like affirmative action and banning the public display o f arguably

racist political insignia. Their support o f explicitly racial agendas is less certain, although

black Democrats were more likely to support a bill recognizing the injustices o f slavery.

5 i Conclusion

‘■Representation through legislation" resonates differently with certain segments

o f the black community. Descriptive representation increases the likelihood that blacks

highly prioritize black interest policies whether symbolic or substantive and across issue

areas. While the effects might be small, they show a more or less consistent willingness

on the part o f blacks with black representatives to support legislative activity aimed

squarely at promoting and protecting black interests—defined substantively or

symbolically, regardless. Representation by a legislator o f the same party does not have

nearly as broad an impact on black policy preferences.

The willingness of blacks to highly prioritize black interest policies is countered

by the non-racial, class-based preferences expressed by better-educated and wealthier

blacks. Statistically significant differences between the preferences o f higher and lower

status blacks were evidenced in the likelihood o f their prioritizing legislative activity to

ban alcohol and cigarette billboards, create an African American history museum, and

secure a national apology for the atrocious act o f black enslavement. Lower status blacks

were more supportive of legislative activity on each of these issues than their higher

117

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. status brethren. Based on the six issues analyzed in this chapter, it appears that when

black legislators sponsor symbolic legislation, lower status blacks are a more receptive

audience than higher status blacks.

It may be that as larger proportions o f group members become integrated into the e«- — -- **- - dominant political, economic, and social structure their need for descriptive and symbolic

representation lessens. Education and income are directly rerated to voter turnout and as

such, one might expect black legislators to de-radalize their legislative agendas in order

to reflect the priorities of this segment o f then: constituencies. Their doing so would

result in a loss o f representation that is valued, at least marginally, among the least well-

off and most voiceless blacks.

An alternative argument may be that black constituents with black representatives

give than more latitude than would otherwise be expected to pursue race-specific

agendas, given countervailing class interests. Thus, most black legislators continue to

pursue racial agendas despite growing class differences in the policy priorities o f their

constituents. The leeway that higher status blacks grant their black representatives to

devote time and effort to racialized, highly symbolic policies that are contrary to their

own preferences is reflective o f the importance o f trust in the representation relationship.

Trust gives representatives leeway to deviate from the strict preferences, or demands, o f

their constituents (Bianco 1994,86-89).

Trust arises when one of two circumstances occurs. First, constituents trust their

representatives when they believe that the representative shares the same preferences or

goals in a specific policy area as they do. In this scenario, the representative behaves as a

fiduciary agent— when he acts in his own best interests he is also acting m the best

118

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. interest of those he represents. Second, and more frequently, constituents trust their

representatives conditionally—in one area, but not in another. Although constituents

have some uncertainty about the representative's preferences or goals, they either believe

the representative has considerably more information than they do on the issue and thus

defer to his decision or they believe that the probability o f shared interests is sufficiently

high to warrant trust. As I suggested in chapter 3, blacks may allow black legislators to

pursue their own preferences in some areas (e.g.. voting against welfare reform) in

exchange for their loyalty in other areas. When black representatives sponsor bills that

resonate most with lower status blacks, higher status blacks may conditionally trust black

legislators and grant them leeway on certain issues and in certain legislative contexts

because they trust them to behave as fiduciary agents on other, relatively more important,

issues.

119

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Constitutional Amendment for affirmative action None o fthen time 4% Some o f then time 40% Most oftheir time 56% N = 8I6 Ban on alcohol and cigarette billboard advertisements in inner-city minority communities None ofthen- time 13% Some o f then- time 41% M ost o f their time 46% N =829 Ban on display of Confederate flag in public buildings None of then- time 26% Some o f their time 43% Most o f their time 31% N=79G Bill to create publicly-funded African American History Museum None oftheir time 7% Some o f their time 58% M ost o f their tim e 35% N=813 Bill to equalize funding between inner- city minority schools and suburban schools N one o f them tim e 2% Some o f their time 34% Most of their time 64% N =816 Bill to acknowledge fundamental injustice o f slavery None o f their time 16% Som e o ftheir tim e 41% M ost o f their tim e 43% N=8G4 * Chi-square p<.05 •* p <.01 (one-tailed test) * p < 10 (two-tailed test) Due to rounding error, cell entries may not sum to 100% .

Table 5J Black Support for Hypothetical Black Interest Bills 120

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Affirmative Action Partisan Representative Descriptive Representative YES NO YES NO None o ftheir time 3% 5% 4% 4% Some o f their tune 38 41 36 42 Most of their time 59 54 60 54 N=783 N=816 Billboard Advertisements None oftheir time 14 12 14 12 Some oftheir time 39 45 35 45 Most ofthen time 47 43 50 43 N=795 N=829* Confederate Flag None of then time 24 27 25 26 Some of their time 47 38 44 42 M ost o f their tim e 29 36 30 32 N=760* N=790 African American Museum None o ftheir time 6 7 7 6 Some of their time 58 57 54 61 Most of their time 36 36 39 33 N =782 N =8I3 Urban and Suburban Schools N one o f their time 2 2 3 I Some of their time 32 36 28 37 Most of their time 66 62 69 61 N=783 N=816** Injustice of Slavery None of their time 15 18 15 17 Some o f their time 41 40 38 44 Most of their time 44 41 48 39 N=772 N=783+ * Chi-square p<.05 ** p <.01 (one-tailed test) ~ p <. 10 (two-tailed test). Due to rounding error, cell entries may not sum to 100% . Partisan: same party as respondent. Descriptive: same race as respondent.

Table 5.2 Black Support for Hypothetical Black Interest Bills by the Type of Representation

121

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Affirmative Action Ideological Party Affiliation Race Identification Lib Con Dem Rep Strong Weak None o f their time 2% 5% 3% 14% 3% 4% Some oftheir time 3$ 41 39 39 40 39 M ost o f their tim e 59 54 58 46 57 55 N=734* N=783 ** N=798 Billboard Advertisements None o f them time 12 10 14 11 11% 17% Some o f their time 45 35 42 36 44 34 Most o f them time 43 45 44 52 45 49 N=742 N=795 N=807 Confederate Flag N one o ftheir tim e 25 27 24 36 24% 31% Some o f their time 44 43 44 39 44 40 Most o f their time 31 21 32 25 32 29 N=709 N =760 N=771** African American Museum None o f their time 6 7 6 5 6% 8% Some oftheir time 60 57 59 26 60 54 Most o f their time 35 36 35 31 33 38 N=728* N=782 N=792* i Urban and Suburban Schools N one o f them time 6 2 2 3 2% 3% Some o f their time 31 34 34 29 34 33 M ost o f their tim e 68 63 64 68 63 64 N=733 N=783 N=795* Injustice of Slavery None of their time 19 14 16 25 17% 14% Som e o f them time 43 40 39 34 43 38 Most o f their time 39 45 42 43 40 48 N=724" N=772 N=784+ * Chi-square p<.05 ** p <.01 (one-tailed test) ~p<.IQ (two-tailed test) Due to rounding error, cell entries may not sum to 100%. Ideology: 7-point scale, values 1-3 “conservative^ and values 5-7 “liberal.” Party: 7-point scale, values 1-3 “Republican" and values 5-7 “Democrat-” Race Identification: 4-point scale, values 1-2 “weak” and values 3-4 “strong.” Respondents' were asked how strongly they agreed or disagreed with the following: “What happens to black people in this country has a lot to do with what happens to me.”

T ab le S 3 Black Support for Hypothetical Black Interest Bills by Respondents* Self- Identified Ideological, Party, and Race Identification

122

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Independent Affirmative P(y=3) B illboards P(y=3) School P fr=3) V ariables A ction Funding Socioeconomic Variables Education -.008 -.067’ -2 0 4 -.017 (-03) (.03) (.03) Income -.040* -.162 .001 -.003 (.02) (.02) (.02) Self-Identification V ariables Race Identifier .006 .036 .046 (.05) (.04) (.05) Ideology (liberal) .044* .104 -.010 ' .028 (-02) (.02) (.03) i <

Party (Democrat) o .164 -.033 -.018 © © © CS ^ CS G03) (0 4 ) Representation V ariables Descriptive .233' .089 -.008 .195 (.12) (-11) (.12) Partisan .149 .079 .047 (-11) (-11) (.12) D istrict Characteristics Median Income .000" 311 -.000 .000 (.00) (.00) (.00) South .007 -.114 -.093 (-11) (.10) (.11) O ther Gender (Female) .202* .078 2 0 4 ' .081 .057 (-10) (-02) (.10) N 649 656 664 Log-likelihood -509.89 -639.22 -475.72 C hi-square 33.10 13.61 8.55 * Chi-square p<05 ** p

Table 5.4 The Effects of Descriptive and Partisan Representation on Black Support for Substantive BQIs

123

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Independent P(y=3) *(y=3> P (y=3) Variables Flag M useum Slavery Socioeconomic V ariables Education .027 -.053* -.152 -.056* -.174 (-03) (.(B) (.03) : Income .002 -.042 -.149 -.068" -.258 (-02) (.02) (.02) Self-Identification V ariables Race Identifier .075* .079 -.017 -.060 (.04) (.04) (.04) | Ideology (Liberal) .011 .032 -.001 (-02) (.02) (.02) Party (Democrat) .065' .124 -.008 .062* .135 (.03) (-03) (.03) Representation V ariables Descriptive .050 .148 .223' .088 (.12) (.ID (.11) Partisan -.140 -.038 -.008 (.11) (.11) (.11) D istrict Characteristics Median Income .000 -.000 .000 (-00) (.00) (.00) South .079 -.153 .070 (.10) (.10) (.10) O ther Gender (Female) -.085 -.096 -.114 (.09) (.10) (-10) N 630 645 638 Log-Iikelihood -547.85 -637.82 669.45 Chi-square 13.88 18.24 39.08 * Chi-square p<.05 ** p <.01 (one-tailed test) * p < J0 (two-tailed test) Note: p(y=3) is the probability that respondents want legislators to spend “most” of their tim e on the bilL Ideology: 7-point scale, values 1-3 “conservative” and values 5-7 “liberal.” Party: 7-point scale, values 1-3 “Republican” and values 5-7 “Democrat.” Race Identification: 4-point scale, values 1-2 “weak” and values 3-4 “strong.” Respondents' were asked how strongly they agreed or disagreed with the following: “W hat happens to black people m this country has a lot to do with what happens to me.” Partisan: same party as respondent. Descriptive: sam e ra c e a s respondent

T able S 3 The Effects o f Descriptive and Partisan Representation on Black Support lor Symbolic BiBs 124

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER 6

CONCLUSION

In this dissertation* I have tried to answer the question* “how are black interests

represented in Congress* and by whom?” in chapter I* I presented the politics of

presence versus the politics of parties1* as the theoretical framework for understanding

black representation. This normative debate raises the question o f whether blacks are

best represented in government by descriptive or partisan representatives. After

developing a more comprehensive and broad definition o f political representation than

analyzed previously* I conclude that the race and party of legislators intersects black

representation differently at different stages in the legislative process. The elected

representative’s race is the key explanatory variable for black interest agenda-setting in

the early, pre-deliberatrve stages of the legislative process* while the representative’s

party* and to a lesser extent his or her race* are key to understanding policymaking on

behalf o f black concerns at the level o f roll-call votes.51

M Anne Phillips (1995) develops this framework in considerable detail in her work on gender and racial representation in legislatures. 55 The race o f the legislator gains importance as a predictor of representatives’ roll-call voting behavior when the policy being voted on directly benefits blacks, such as civil rights policies (Whitby and Krause I99S* Whitby 1997).

125

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. I began my analysis with the proposition that black representation cannot be

sufficiently summarized through an analysis o f roll-call votes. In chapter 3 ,1 argue that

although the liberal vote is a key indicator o f responsiveness to black concerns, it is not

enough. Black representation at this final stage in the House legislative process is

constrained by the limited number of bills reported from committee that directly or

indirectly reflect racial concerns. Only a small number of the black interest policy

proposals sponsored during each session survive the filtering activities o f the committee

system. Furthermore, the focus on voting in congressional scholarship ignores the many

other opportunities for black representation that occur earlier in the legislative process.

These other opportunities include holding hearings, cosponsoring bills, writing “dear

colleague” letters to elicit support for certain issues, and most importantly for my

purposes, sponsoring legislation.

Bill sponsorship behavior can be distinguished from roll-call voting behavior by

the level of intensity it indicates. Making a black concern a legislative priority through

initiating legislation is qualitatively different than voting to support or oppose a black

concern. House members do not engage in either activity before considering the potential

costs and benefits o f taking action. Those who sponsor bills that address black concerns

are using political capital to signal to constituents and House colleagues that black issues

are among their personal priorities. As described in chapter 4, bill sponsorship can serve

representatives’ representational goals in several ways. Introducing a bill is a means for

representatives to present themselves to constituents and colleagues, carve out policy

niches, recast the boundaries of policy debates, respond to constituent demands, and

advocate on constituents’ behalf.

126

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Scholars have long-recognized that legislators sponsor hills that are symbolic,

e.g.. Kmgdon 1995, and Schneier and Gross 1993. In chapter 4 ,1 developed a typology

o f black interest bill sponsorship that identified bills as symbolic, part-symbolic (hybrid),

and substantive. Classifying legislation as either substantive, hybrid, or symbolic

constitutes a new and important contribution that my study makes to the field of

congressional studies.

An observation I made while working as a congressional fellow in Representative

Maxine Waters* (D-CA) Washington, D.C. office from the fall o f 1993 through 1994

further illustrates how blacks can be symbolically represented. Representative Waters

gained national attention during the Los Angeles riots in 1992. She is an astute and

experienced politician, having served for several years in the California state assembly

before winning her congressional seat after the retirement o f Representative Augustus

Hawkins in 1990. In the wake of the riots, Representative Waters demanded that

Congress develop an urban agenda to address the ills facing the nation's cities. In March

of 1993, the Congresswoman introduced two bills—H.R. 1020, the Job Training and

Partnership Act and H.R. 1021, the Neighborhood Infrastructure Improvement and Inner-

City Job Creation Act—calling for a total of 15 billion dollars in aid to cities. She

discussed the legislation on the House floor, on television, and in local newspaper

columns— using any venue that would allow her to make her case. During the time that

the bill was being drafted, a member of her legislative staff observed that the

Congresswoman seemed intent on sponsoring legislation that had no chance of being

successfully enacted.

127

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. What explanations account for why an experienced, informed, rational actor

behaved in a way that appeared to undermine serious consideration o f the very issue she

sought to address? One explanation is that she used bill sponsorship as a tool to

transform the political debate about appropriate solutions for the urban crisis. She had

spent her freshman year in the Congress drawing attention to the plight o f urban areas.

Roughly one month after introducing the bills, Representative Waters gave a speech on

the House floor titled, “The Ticking Time Bomb in Our Cities.” She reminded

colleagues o f her advocacy for cities prior to the riots:

Madam Speaker, one year ago today, Los Angeles erupted in flames. Since my election to this House.. J have attempted to create a dialogue in Congress about poverty and despair... 1 spent countless hours defining the roots of our urban crisis. I talked about endemic unemployment and underemployment in our inner cities. I spoke of how companies were closing up shop and moving abroad... I described the damage done by 12 years o f outright abandonment o f our dries by Ronald Reagan and George Bush. 1 talked about how banks had redlined our communities, how the criminal justice system had failed us, and how racism was—alas—alive and well...there are some who say we can’t afford not to help Russia with aid. After all, they still possess nuclear weapons. That’s all well and good, but let me remind you, charity begins at home and there’s a ticking time bomb in our dries as well...W e ignore it at our own peril.56

In this example, Representative Waters used both bill sponsorship and floor

speaking to “speak” on her constituents’ behalf. Based on the typology that I presented

in chapter 4, her bill sponsorship in this area was symbolic—aimed at giving voice to the

concerns o f her constituents and the larger black and urban community, rather than

specifically tailored to win immediate tangible policy objectives. In the context of

56 As printed in the Congressional Record (1993, H2I79). 128

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. uncrystallized interests described by Mansbridge (1999), Representative Waters* actions

were an attempt to reshape the boundaries o f deliberation in the legislature.

Requesting IS billion dollars in urban aid communicated a sense of urgency,

marked a potential course of action, and suggested a comprehensive solution for a

growing urban crisis. The communication operated at two levels, between the bill

sponsor and the legislature, and between the sponsor and constituents.57 This type of

representation is overlooked in studies o f black representation that examine only roll-call

votes because agenda-setting behavior is not accounted for.

Several studies since the mid-1990s have empirically analyzed black

representation through bill sponsorship at the state and national level, but none o f those

have so thoroughly considered the bill sponsorship behavior of legislators as I have in

this dissertation. I developed a theory of symbolic bill sponsorship that accounts for

purely symbolic policy proposals as well as bills, such as Representative Waters*

described above, that serve an immediate symbolic purpose despite them appearances as

substantive. Beyond this theoretical contribution, my empirical findings are also unique

in that they demonstrate how prevalent symbolic bill sponsorship is as a tool in black

representation through legislation.

6J Black Representation through Bill Sponsorship

Establishing that a substantial proportion o f black interest legislation is actually

symbolic or part-symbolic (hybrid) is a second critical contribution that I make in this

57 A third level o f communication could include communication between the bill sponsor and the larger community o f which she is a part- 129

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. dissertation. Chapter 4 shows that half of the black interest bills sponsored by white

legislators and more than two-thirds o f the black interest bills sponsored by black

legislators were symbolic. Bills were coded as symbolically representing black interests

if they were purely symbolic (possessing no tangible impact even if enacted, e.g.,

buildings honoring black heroes), policy outliers (e.g., reparations), targeted unorganized

constituencies (e.g., the homeless), or were technically unfeasible (e.g., full employment).

An additionally important finding is that the race o f the legislator was the key variable in

explaining variation in black interest bill sponsorship by type. Black legislators were

positively associated with the sponsorship o f black interest bills in every category except

substantive. Differences in member race, party, and the racial composition of districts

explained none of the variance in substantive black interest bill sponsorship. This

indicates that symbolic and substantive bill sponsorship are distinctive behaviors that

may not be explained by the same sets o f factors as roll-call voting.

The importance of symbolic bill sponsorship is in its communicative function.

Symbolic bills are vehicles that legislators can use to challenge the status quo, introduce

new issues to the legislative agenda, and offer constituents alternative ways to view

political potentialities. These bills can synthesize grassroots movements, provide a

platform for more tangible policy initiatives, and call attention to individual legislators'

broad policy objectives. In a legislative arena where bill sponsors know that the chances

for any bill are slim, symbolic bill sponsorship offers a low cost way to take a position,

stake out a policy niche, or advocate for group interests. Engaging in symbolic bill

sponsorship is neither a naive behavior, nor a waste of the legislator's time. To the

•t contrary, doing so provides a prime venue for representatives to express support or

130

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. empathy for certain groups in an uncertain, and sometimes hostile, political environment.

Roll-call voting provides blacks with a different land o f representation.

6.2 Black Representation through Roll-call Votes

Chapter 3 examined the roll-call voting behavior of members in support of

objective and subjective black interests. In an analysis o f House members' LCCR scores,

I found that, all else being equal, the party of the member was a more important

explanation of liberal voting behavior than was the race of the member, seniority, the

racial composition in the district, or region. These findings square well with other

research in this area and require no further explication here. The findings regarding how

well members represented subjective black interests require more discussion.

Chapter 3 establishes that race, and not party, accounted for the degree to which

House members represented objective black interests. Using an index that matched black

public opinion to six roll-call votes, I found that black representatives were significantly

less likely than white representatives to vote in a manner consistent with black public

opinion. Black. legislators were found to be more liberal in their voting behavior than

black opinion warranted on four of the six issues (crime, immigration, death penalty,

welfare reform). Only on two issues (affirmative action and defense spending) did the

majority ofblack legislators vote in concert with national black opinion.

However, the finding that black legislators are more liberal than blacks nationally

must not be overstated. Most black representatives were elected during a period o f time

when American political attitudes were liberalizing greatly. These black members still

hold office, and yet as noted in chapter 3, blacks have become more conservative. This

131

!

•s.. ' ^ H . J Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. new political conservatism among blacks does not indicate a sweeping shift in black

political attitudes. Even with this trend favoring conservatism among rank-and-file blacks

since the 1980s. they remain the most liberal group in America.

Some recently elected black members of Congress, e.g., Representative Sanford

Bishop (D-GA), have been more willing to break ranks with senior, solidly liberal black

representatives on key policy proposals, including welfare reform and crime prevention,

and gun control. This suggests that as older representatives retire, they may be replaced

by politically ambitious, post-civil rights era representatives who more readily adapt to

the centrist policies of the Democratic Party. This new generation of representatives

may more naturally reflect the socially conservative preferences of a growing segment of

the black community. Such an outcome will exacerbate the paradox of advocacy

highlighted m chapter 2.

The paradox of advocacy is that blacks are faced with a trade-off between the goal

of increasing black substantive representation through maintaining mostly white

Democratic voting majorities in the legislature, and increasing racial group advocacy

through pushing for more descriptive representation. Chapters 3 and 4 showed that black

representatives assumed the chief responsibility for promoting a black agenda through a

combination of voting for objective black interests and bill sponsorship during the 104th

Congress. If a new generation of black representatives abandons a racialized policy

agenda, black advocacy through agenda-setting will decline leaving the most

marginalized blacks without a voice in Congress.

132

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 63 Black Preferences for Symbolic and Substantive Policies

Chapter 5 analyzed survey data for indications of which blacks were more

supportive of symbolic policy proposals. Respondents were asked to prioritize three

symbolic and three substantive hypothetical policies on black House members’ legislative

agendas. The data show that blacks were generally supportive of both symbolic and

substantive black interest bills. However, the results of the multivariate analysis show

that socioeconomic disparities existed in blacks’ assessments of how high black

legislators should prioritize black interest policies. Blacks with higher socioeconomic

status were less inclined to highly prioritize legislative action on the bills. Strong race-

identifiers who strongly believe that their fate is linked to the experiences of other blacks

did not consistently support symbolic or substantive policy proposals. They were also

less inclined to support symbolic policies than weak race-identifiers, though the

differences were not statistically significant.

6.4 Directions for Future Research

Black political representation is a complex enterprise that should no longer be

limited to the analysis of how members vote. I have shown that bill sponsorship behavior

can be categorized by direct or indirect focus, and type: symbolic, hybrid, and

substantive. Further, I have argued that bill sponsorship behavior is distinct from roll-call

voting behavior to the extent that sponsorship measures the intensity of a member’s

public commitment to black concerns. Twenty percent of bills introduced by the House

members m this study were black interest bills. Black representatives sponsored a

133

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. majority of those bills and a large proportion of all legislators* total black interest bill

introductions were symbolic.

Despite these advances in understanding how, and by whom, blacks are

represented in Congress, my findings are limited to representational behavior in a single

Congress. Are black legislators more or less likely engage in symbolic bill sponsorship

behavior when they are in the minority party and have lost the benefits of majority party

power, e.g., committee and subcommittee chairmanships? The answer turns on whether

symbolic behavior increases when group marginalization increases. Black Democratic

representatives may have been swept into the ranks of the minority party in the wake of

the 1994 Republican takeover, however, as Guinier (1994) points out in her essay, “The

Triumph of Tokenism.” black representatives remain marginalized even when they are

part o f the majority coalition.1*

With regard to voting for black issues, the benefits of descriptive representation

do not consistently translate to black substantive representation when black interests are

defined subjectively (wants) rather than objectively (needs.) Although the liberal vote is

a key feature of maintaining a progressive black policy agenda, it cannot fully account for

the growing social conservatism expressed by many blacks. However, the limited

number of votes used to construct the measure of subjective black interests demands a

cautious interpretation of this finding. Because of the conservative nature of votes in the

104th Congress and the small number of roll-call votes that could be reasonably matched

to black public opinion, it is possible that 1 have overestimated the disconnection between

5* See Smith (1996) for a discussion o f the Congressional Black Caucus alternative budget as an example ofblack marginalization in the House- 134

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. black legislators and black masses. Designing a survey to examine public opinion on

specific legislation or roll-call votes could remedy this uncertainty.

Future research should also consider a larger sample of white House

representatives. While limiting the sample size to representatives from districts with 15

percent black populations allowed for the inclusion of all black Democratic House

representatives, it excluded many white liberals from non-southern districts. These

representatives may be additional sources ot symbolic and substantive black

representation.

6.5 Symbols and Substance in Black Representation

Representing blacks requires a complex, sophisticated, strategic, and innovative

approach from elected representatives. Representing black interests can range from

voting to support black concerns within the boundaries of the status quo political

establishment, to advocating alternatives to politics as usual through introducing of

ambitious policy proposals. This study shows that black and white legislators behave

differently at different stages in the legislative process.

Black representation is a story of balancing approaches. The political importance

of the black presence at the agenda-setting stage is balanced by the importance of party at

the vote-aggregating stage. The tangible benefits gained by narrowly-tailored pragmatic

substantive policy proposals that address specific and immediate black concerns within

the confines of the contemporary political environment are balanced by the broad, far-

reaching, sometimes radical symbolic proposals .that serve to signal new political

possibilities just beyond the horizon.

135

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. This study has uncovered some contradictions in the descriptive representation of

black interests. Black representatives do not consistently represent black policy

preferences, particularly when blacks express conservative opinions. Black

representatives lead in agenda-setting tor black concerns, but most of their bill

sponsorship activity is symbolic. And. finally, socioeconomic differences in black

preferences for symbolic and substantive proposals require black legislators to figure out

how to respond to the needs of a black community that at one level is becoming more

distinctive, while at another level it remains closely linked by common fate perceptions.

A key feature in the relationship between black representatives and the black community

is trust. The political, social, and economic alienation and shared history of

discrimination that reinforce a strong racial group identity among rank-and-ftle blacks

also shapes the relationship that blacks have with their black representatives.

Fundamentally, blacks trust black representatives to represent them and, in certain

circumstances, give them leeway in determining how best to do so.

This study has implications for the study of how group interests are represented

more broadly. Representatives use legislative tools to represent all kinds of groups, from

racial minorities and women, to corporations and organized interest groups. Symbols and

substance are fundamental to political discourse and as such merit further study in other

contexts.

136

*

. W Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. APPENDIX A

DESCRIPTION OF NBES SURVEY DATA

The 1996 National Black Election Study fNBES) was conducted in 1996 by Katherine

Tate at the Ohio State University with tunds from the Ohio State University and the

National Science Foundation (SBR-9507469). Like the previous 1984 and 1988 NBES.

the 1996 NBES was a random-digit dialing telephone survey. It was designed to provide

an ongoing examination of black political attitudes and behavior, particularly behavior

during the 1996 presidential election. The survey was also designed to facilitate

comparisons across time with the previous NBES surveys, and to allow for racial

comparisons with the 1996 Center for Political Studies5, National Election Study. The

pre- and post-election surveys include 1.216 and 854 respondents respectively. The pre­

election survey was conducted from July 19 through November 4, 1996. The post-

election survey is a re-interview of respondents from the pre-election survey and was

carried out from November 4. 1996 through January 6. 1997. Respondents in the 1996

NBES were matched to their congressional districts and asked to evaluate their own

representatives. A total o f252 House districts were included in the survey. Nearly one-

third of the black respondents in the sample were represented by black representatives.1

1 Background information about the 1996 NBES was taken from Tate (2002.211-21 5) Black Faces in the Mirror: African Americans and Their Representatives in the U.S 137

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. For this dissertation. I included only representatives in the NBES sample from

districts with at least 15 percent black populations. Afrer removing three representatives

who failed to serve complete terms during the 104th and one Latino representative to

simplify the race variable, my sample included representatives from 96 of the original

252 districts and 33 of 38 black Democratic House members in the NBES." I

supplemented the NBES with data pertaining to individual members of Congress. This

included adding information on district characteristic:;, such as the racial composition of

the district, region, median household income, and percent urban. I also collected data on

the seniority, party affiliation, and vote margin of each representative. Most importantly.

I collected data on the roll call voting and bill sponsorship behavior of each

representative m the sample. This required creating a measure of black pohc>

preferences on issues covered in the NBES. finding marching roll call votes during the

104th Congress, and comparing the collective preferences of rank-and-file blacks to how

representatives voted. The collection of bill sponsorship data was a time-intensive

endeavor that required reading the bills sponsored by each representative in my sample

and coding them for focus (direct or indirect black focus) and type (symbolic,

substantive, and hybrid).

Congress. See Tate (2002) for more detailed information about sampling methods and response rates for the 1996 NBES. 2 A total of 100 districts in the NBES sample had black populations o f 15 percent or more. Representatives from three districts failed to complete full terms during the 104th Congress and were excluded. A fourth representative was excluded to simplify the race variable. The representatives, by name, were: Reps. Mfume (D-MD). Tucker (D-CA). Reynolds (D-IL). and Serrano (D-NY).

138

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. APPENDIX B

SUMMARY DESCRIPTIVES OF REPRESENTATIVES IN SAMPLE

The units of analysis for chapters 3 and 4 of this project are U.S. House Representatives whose districts were included in the 199o National Black Election Study. Only representatives from districts with at least 13 percent black populations were included in the analysis. The sample is mostly white i65.b percent), heavily Democratic (74 percent), mostly southern <63.5 percent), and mostly urban (86.5 percent). The table below provides some summary statistics on each Representative in the sample

Race 1 State 1Representative 1 P a n ' Region I % Black Black t I AL Earl F. Hilliard 1 Dem South i 68.00 1 CA Ronald Dellums i Dem Non-South 1 32.00 h ' Dem Non-South 1 40.00 ... Julian Dixon 1 t Maxine Waters 1Dem Non-South 1 43.00 1 FL Corrine Brown 1 Dem South 155.00 ! Carrie Meek 1Dem South I 58.00 i Alcee Hastings 1 Dem South I 52.00 I g a Sanford Bishop. Jr. ! Dem South 157.00 1

i John Lewis i Dem South 1 62.00 t r * '“ t I Cvnthia McKinnev I Dem South 164.00 | i flL Bobby Rush ( Dem Non-South 1 70.00 ! i j Cardiss Collins 1 Dem Non-South 1 66.00 I LA William Jefferson 1 Dem South 161.00 Cleo Fields 1 Dem South ! 58.00 ! I MD Albert Wvnn I Dem Non-South 1 58.00 I Ml John Conyers. Jr. i Dem Non-South 1 69.00 i Barbara Rose Collins I Dem Non-South I 70.00 , MS Bennie Thompson I Dem South I 63.00 ! ! MO William Clav I Dem Non-South 1 28.00 t continued)

Table B.l Descriptive Statistics for Representatives from Districts with 15 Percent Black Populations in the 1996 NBES Sample

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table BJ (continued)

INJ (Donald Payne I Dem Non-South 1 60.00 ! (NY I Floyd Flake | Dem Non-South ! 56.00 1Edolphus Towns Dem Non-South 161.00 j 1Major Owens | Dem Non-South 1 74.00 1 i t Charles Rangel \ Dem Non-South 47.00 ! 1 NC 1Eva Clavton 1 Dem South 57.00 1Melvin Wan 1Dem South 57.00 , * OH I Louis Stokes 1 Dem Non-South 59.00 1 PA 1Chaka Fattah |Dem Non-South 52.00 1 s r James CIvbum 1 Dem South 62.00 1 i TN Harold Ford Dem South 59.00 | ; t x Sheila Jackson-Lee 1 Dem South 51.00 . 1 E. Bernice Johnson Dem South 50.00 j I VA Robert Scon Dem South 64.00 1

| White . ! Al Sonnv Callahan Rep South 28.00 i lerry Everen Rep South 24.00 Glen Browder Dem South 26.00 Robert Cramer Dem South 15.00 1 AR Rav Thronton Dem South 18.00 | ; Jav Dicke\ Rep South 27.00 ; i CA George Miller Dem Non-South 17.00 ! 1 DE Al Michael Castle Rep Non-South 17.00 i i FL Pete Peterson Dem South 23.00 I Sam Gibbons Dem South 17.00

! 1 GA Jack Kingston Rep South 23.00 1 Mac Collins Rep South 18.00 I | Saxbv Chambliss Rep South 21.00 Charlie Norwood Rep South 18.00 1 IL Jerrv Costello Dem Non-South 17.00 I ! IN Peter Viscloskv Dem Non-South 21.00 Andrew Jacobs Dem Non-South 30.00 1 KY Mike Ward Dem South 18.00 t LA Billy Tauzm Dem South 21.00 i ; Jim McCren Rep South 27.00 I ! Richard Baker Rep South 18.00

t .» Jimmy Hayes Dem South 18.00 I MD Wavne Gilchrest Rep Non-South 15.00 t Benjamin Cardin Dem Non-South 17.00 Steny Hoyer Dem : Non-South 19.00 1 MI Dale Kildee Dem Non-South 18.00 t continued)

140

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table BJ (continued)

1 MS Roger Wicker i Rep I South | 23.00 j ; G.V. Montsomerv I Dem 1 South 131.00 1 i Mike Parker 1 Dem 1 South I 41.00 1 Gene Tavior 1 Dem 1 South | 20.00 i 1 MO Karen McCarthv I Dem I Non-South 24.00 1 I NJ Robert Andrew- I Dem 1 Non-South 1 16.00 1 NY Daniel Fnsa t Rep I Non-South 16.00 1 1 Eliot Engel I Dem 1 Non-South 42.00 Jack Quinn 1 Rep ! Non-South 17.00 ' I NC David Funderburk f Rep ‘ South 22.00 Walter Jone> 1 Rep ! South 21.00 i Fred Heineman ■ ReP ! South 20.00 ! t 1 Rep ! South 15.00 Charlie Rose 1 Rep i South 19.00 W.G. " B iir Hetner 1 Dem ! South 23.00 1 OH Steve Chabot 1 Rep 1 Non-South 30.00 Tonv Hall I Dem 1 Non-South 18.00 | PA Thomas Fogiietta 1 Dem 1 Non-South 52.00

i William Covne i Dem I Non-South 18.00 . 1 s c Mark Sanford 1 Rep ! South 20.00

, Flovd Spence 1 Rep ! South 25.00 1 : Lindsev Graham 1 Rep ! South 21.00 i Bob Ingiis 1 Rep i South 20.00 John Sprart. Jr. I Dem 1 South 31.00 1 I t n Bob Clement 1 Dem ! South 23.00 ! I John Tanner 1 Dem 1 South 20.00 ! ITX Jim Chapman 1 Dem 1 South 18.00 I John Brvant t Dem 1 South 16.00 1 Steve Stockman 1 Rep 1 South 22.00 i Chet Edwards 1 Dem 1 South 16.00 Martin Frost 1 Dem 1 South 19.00 Ken Bentsen t Dem ! South 27.00 : VA Herbert Bateman 1 Rep 1 South 18.00 Owen Pickett 1 Dem ! South 17.00 i 1 Norman Sisisky I Dem 1 South 32.00 | ! L.F. Payne. Jr. t Dem ! South 25.00 WI 1 Thomas Barren 1 Dem 1 Non-South 35.00

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. APPENDIX C

1996 NBES (POST-ELECTION) SURVEY QUESTIONS USED TO CREATE THE BLACK OPINION INDEX

Crime Should federal spending on crime be increased, decreased- or kept about the same*? (N=844) Valid Percem Increased 63.5 Same 30.9 Decreased 5.b

Death Penalty Do you favor or oppose the death penalty for persons convicted of murder? (N=752 1 Valid Percent Favor 54.0 Oppose 46.0

Defense Spending* Should federal spending on military and defense be increased- decreased- or kept about the same? (N=836) Valid Percem Increased 21.2 Same/Decreased 78.9

Welfare Reform There's a law that puts a five-year limit on how long someone can receive welfare benefits. Do you favor or oppose this five-year limit? (N=807) Valid Percem Favor 68.5 Oppose 31.5

142

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Affirmative Action Because of past discrimination, minorities should be given special consideration when decisions are made about hiring applicants for jobs. (N=&12) Valid Percent Agree 55.4 Disagree 44. h

Immigration** Do you think the number of immigrants from foreign countries who are permined to come to the United States to live should be increased a little, increased a lot. decreased a little, decreased a lot. or left the same as it ts now? (N=802) Valid Percem Increased 9.9 Same 40.4 Decreased 49.o

Note: * A majority of respondents. 52.9%. chose the middle ground of keeping defense spending the same. To make the question more useful for analysis. I collapsed the middle category into the decrease spending category. This position is supported by data that show that most blacks would not increase detense spending (Dawson 1994. Tate 1994). **Opinion on immigration is divided, but only 10 % supported an increase.

143

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. APPENDIX D

ROLL-CALL VOTES USED TO CREATE THE BLACK OPINION INDEX (BO I). 104™ CONGRESS

Issue Area Roll-call Number, and Subject o f Vote NBES Position

Crime 117 Passage of H.R. 667 to increase federal Favor Prison Spending spending on prisons by S 2.6 billion dollars

Death Penain 109 Passage of H.R. 729 to make it more likely Favor Limits on Appeals that the death penalty will be invoked by limiting habeas corpus appeals

Weltare 269 Passage o f H.R. 4 to limit weltare benefits to Favor Welfare Reform Five years (vetoed by President i

Affirmative Action Amendment to support affirmative action Favor Mtnontv Broadcasters policies for minorities who own broadcast properties

National Defense 646 Passage of HJL 2126 to increase spending Oppose Increase spending on national defense

Immigration 89 Passage of HJL 2202 to increase restrictions Oppose Limit illegal on illegal immigration

144

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. APPENDIX E

ISSUES IN LEADERSHIP CONFERENCE ON CIVIL RIGHTS (LCCR) SCORES, HOUSE OF REPRESENTATIVES, 104™ CONGRESS

Issue Area LCCR Position

1. Welfare Reform Opposed 2. Welfare Overhaul/Passage Opposed 3. Illegal Immigration Restrictions/Conference Repor Opposed 4. Immigration Education Restrictions/Passage Opposed 5. Immigration Restrictions/ Legal Immigration Opposed 6. Minimum Wage Increase/ Passage Supported 7. Minimum Wage Increase/ Small Business Exemption Opposed 8. Church Arson Prevention/ Passage Supported 9. Legal Services Corporation/ Appropriations Supported 10. Balanced-Budget Amendment/ Passage Opposed 11. Product Liability/ Conference Report Opposed 12. Education Funding/ Further Continuing Appropriations Supported 13. Lobbying Disclosure for Groups with Federal Grants Opposed 14. Teamwork for Employers and Managers/Passage Opposed 15. Affirmative Action for Broadcast Properties Supported 16. Health Insurance Revisions/ Conference Report Supported 17. English as Official U.S. Language/ Passage Opposed 18. Crack Cocaine Sentencing Guidelines/ Passage Opposed 19. Family Rent Cap Supported 20. Assault Weapons Ban Repeal/Passage Opposed

145

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. APPENDIX F

DESCRIPTION OF VARIABLES

Representative Characteristics Race A Dummy variable coded as I for black re Pam A Dummy variable coded as I tor Democrats and 0 for Republicans. ______I Seniority _ Number of vears in office. I District Characteristics Percent Black The percentage of the district population that is black based on data from the 1990 census. | Region A Dummy variable coded as I for southern districts, and 0 for non-southern districts. Urban The percentage of the district population that is urban. Median Income Median district income Constituent Characteristics Education Highest level o f education achieved Income i Personal income in current dollars (continued)

Table F.I Description of Variables

1 My measure for southern districts included the following 13 states: Alabama. Arkansas. Florida. Georgia. Kentucky. Louisiana. Mississippi. North Carolina. Oklahoma. South Carolina. Tennessee. Texas, and Virginia (see Cameron. Epstein, and O'Halloran 1996). Note that this definition of south is not the traditional measure based on the 11 states of the Old Confederacy that excludes Maryland. Kentucky, and Oklahoma. Nor is it the same measure as used bv the 1990 Census that, in addition to the states included in my measure, also includes Delaware, the District of Columbia, and Maryland. 146

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table F.l (continued)

Race Identification 4-point scale measuring feelings of common fate between respondent and circumstances affecting other blacks. Based on how strongly blacks agreed with the statement, “what happens to other black people m this country affects what happens to me." Coded i for low race-identification and 4 for high race-identification Ideoloev 7-point scale measuring respondents self- identification as a liberal or conservative. Coded I _for strong conservative and 7 for strong liberal. _____ Pam 7-potnt scale measuring respondents self-identified party affiliation. Coded I for strong Republican and 7 for strone Democrat. Descriptive Representative A Dummy variable coded I tor respondents with black representatives and 0 for those with white representatives ______Partisan Representative A Dummy variable coded I tor respondents with representatives from their own party and 0 for those with representatives from a different party than their i own.

147

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. BIBLIOGRAPHY

Anderson. William D.. Janet M. Box-Steffensmeir. and Vaiena Sinclair-Chapman. 2002 “Modeling Member Effectiveness in the U.S. House of Representatives." Unpublished manuscript

Achen. Christopher H. 1975 “Measuring Representation" American Journal of Politics. 22:475-510.

Aldrich. John H. and Forrest D. Nelson. 1984. Linear Probability. Logtt. and Probit Models. Newbury Park. CA: Sage Publications. Inc

Arnold. Douglas R. 1990. The Logic o f Collective Action. New Haven: Yale University Press.

Barker. Lucius J.. and Mack H. Jones. 1994. 3rd ed" African Americans and rht American Political System. New Jersey: Prentice Hall.

Barker. Lucius J. 1988. Our Time Has Come. Chicago: University of Illinois Press.

Bamello. Michelle A. and Kathleen A. Bratton. 1998. “Racial and Gender Differences in Bill Sponsorship and Roll-Call Voting: The Cases of Illinois and California." Paper delivered at the 1998 annual meeting of the Midwest Political Science Association. Chicago. Illinois.

Barnett. Mageurite Ross. 1987. “The Congressional Black Caucus: Illusions and Realities of Power." in The New Black Politics: The Search fo r Political Power. edited by Michael Preston. Lenneal S. Henderson, and Paul Puryear. New York: Longman Press.

Barone. Michael and Grant Ujifsa. w/ Richard Cohen. 1995. The Almanac o f American Politics. Washington. D.C.: National Journal.

Barrett. Edith J. 1995. “The Policy Priorities of African American Women in State Legislatures." Legislative Studies Quarterly. 20:223-247.

148

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Bartels. Larry. 1991. "Constituency Opinion and Congressional Policy Making." American Political Science Review. 85:457-454.

Bernstein. Robert A. Electionsr Representation, and Congressional Voting Behavior: The Myth o f Constituency Influence. Eneiewood Cliffs: Prentice Hall.

Berry. William D. and S. Feldman. 1985. Multiple Regression in Practice. Beverly Hills. CA: Sage Publications. Inc.

Bianco. William T. 1994. Trust: Representatives and Constituents. Ann Arbor University o f Michigan Press.

Black. Earl and Merle Black. 1987. Politics and Society in the South. Cambridge. MA. Harvard University Press.

Black. Merle. 1978. "The Racial Composition of Congressional Districts and Support for Federal Voting Rights in the South." Social Science Quarterly. 59: 435-450.

Bobo. Lawrence, and Frank D. Gilliam. Jr. 1990 "Race. Sociopolitical Participation, and Black Empowerment.” American Political Science Review. 84:2.377-395.

Bositis. David A. 1995. African Americans and the 1994 Midterm: What Happened? Washington. D.C.: Joint Center for Political and Economic Studies.

______. 1994. The Congressional Black Caucus in the I03ra Congress. Washington. D.C.: Joint Center for Political and Economic Studies.

Box-Steffensmeier. Janet M.. and Valeria N. Sinclair. 1996. "Legislative Effectiveness in the U.S. House of Representatives: Struggle. Strategy, and Success.” Paper delivered at the 1996 annual meeting of the Midwest Political Science Association. Chicago. Illinois.

Brace. Kimball. Bernard Grofman. and Lisa Handley. 1987. Does Redistricting Aimed to Help Blacks Necessarily Help Republicans?" Journal o fPolitics. 49:169-185.

Bratton. Kathleen A. 2002. "The Effect of Legislative Diversity on Agenda Setting.” American Politics Research. 30:115-142.

Bratton. Kathleen A., and Kerry L. Haynie. 1994. "Agenda-Setting and Legislative Success in State Legislatures: The Effects o f Gender and Race.” Journal o f Politics. 61:

149

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ' l u * * - ■' Browning. Rufus P.. Dale Rogers Marshall, and David H. Tabb. 1984. Protest is Not Enough: The Struggle o f Blacks and Hispanics for Equality in Urban Politics. Los Angeles: University of California Press.

Bullock. Charles S. 1995. “The Impact of Changing the Racial Composition of Congressional Districts on Legislators* Roll Call Behavior." American Politics Quarterly. 23:141-58.

______1981. “Congressional Voting and the Mobilization o f a Black Electorate in the South.” The Journal o f Politics. 43:662-82

Bullock. Charles and Susan MacManus. 1981. "Policy Responsiveness to the Black Electorate: Programmatic Versus Symbolic Representation.’* American Politics Quarterly. 9:3.357-368

Cain. Bruce. John Ferejohn. and Morris Fiorina. 1987. The Personal Vote. Cambridge. MA: Harvard University Press.

Cain. Bruce. 1992. “Voting Rights and Democratic Theory Toward a Color-blind Society?” In Grofman and Chandler, eds. Controversies in Minority Voting.

Washington. D.C.: The Brookinss* Institutions.

Canon. David T. 1996. “The Politics ol Commonality and the Politics of Difference in the United States House of Representatives.” Paper delivered at the 1996 annual meeting of the American Political Science Association. San Francisco. California.

______. 1999. Race, Redistrichng, and Representation: The Unintended Consequences o f Black-Majority Districts. Chicago: University of Chicago Press.

Cameron. Charles. David Epstein, and Sharyn OTfalloran. 1996. “Do Majority- Minority Districts Maximize Substantive Representation in Congress?” American Political Science Review. 90:794-812.

CarmichaeL Stokely. and Charles V. Hamilton. 1967. Black Power: The Politics of Liberation. NY: Vintage.

Cavanagh, Thomas E-. ed. 1997. Strategies Jor Mobilizing Black Voters: Four Case Studies. Washington. DC: Joint Center for Political Studies.

Champagne. Richard and Leroy N. Rieselbach. 1995. “The Evolving Congressional Black Caucus: The Reagan-Bush Years.” in Huey L. Perry and Wayne Parent.

150

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. eds. Blacks and the American Political System. Gainesville. Florida: University o f Florida Press.

Clapp. Charles L. 1963. The Congressman: His Work as He Sees Jr. Washington. D.C.: Brookings Institution.

Clay. William. 1993. Just Permanent Interests: Black Americans in Congress. 1870- 1992. New York: Amistad Press.

Cobb. Michael D.. and Jeffery A. Jenkins. 1996. “Who Represents Black Interests in Congress? Sponsoring and Voting for Legislation Beneficial to Black Constituents " Paper delivered at the 1996 annual meeting o f the Midwest Political Science Association. Chicago. Illinois.

Cohen. Cathy. 1999. The boundaries o f Blackness. AIDS and the Breakdown o f Black Politics. Chicago: University o f Chicago Press.

Combs. Michael W.. John R. Hibbing. and Susan Welch. 1984. “Black Constituents and Congressional Roll Call Votes "Western Political Quarterly. 37:424-434.

Cook. Timothy E. 1979. “Legislature versus Legislator: A Note on the Paradox of Congressional Support."Legislative Studies Quarterly. 4:43-52.

Congressional Quarterly Almanac. 1996. Washington. D.C.: Congressional Quarterly Press.

Congressional Districts in the 1990s. 1993. Washington. D.C.: Congressional Quarterly Press.

Davidson. Chandler, and Bernard Grofman. eds. 1995. Quiet Revolution in the South. Princeton: Princeton University Press.

Davidson. Chandler. 1984. “Minority Vote Dilution: An Overview." in Chandler Davidson, ed.. Minority Vote Dilution. Washington. D.C.: Howard University Press.

Davidson. Roger H. and Waher J. Oleszek. 2000. 6th ed.Congress and its Members. Washington- D.C.: Congressional Quarterly Press.

Dawson. Michael C. 1994. Behind the Mule: Race and Class in African-American Politics. Princeton: Princeton University Press.

151

Reproduced with permissionof the copyright owner. Further reproduction prohibited without permission. Deen. Rebecca E.. and Zoe M. Oxlev. 1995- “The Representation o f Minority Group Interests: Women and African Americans in Ohio “ Paper presented at the 1995 annual meeting o f the American Political Science Association. Chicago. IL.

Dexter. Lewis Anthony. 1970. Elite and Specialized Interviewing. Evanston. IL: Northwestern University Press.

Duncan* Philip D„ and Christine C. Lawrence. 1997. Politics in America. 1998: The 105th. Washington. D.C.: Congressional Quarterly. Inc.

______. 1995. Politics in America. 1996: The 104th. Washington. D.C.: Congressional Quarterly. Inc.

Duncan. Philip D. 1993. Politics inAmerica. 1994: The 103rd. Washington. D.C.: Congressional Quarterly. Inc.

Elder. Charles D. and Roger W. Cobb. 1983. Tne Political Uses o f Symbols. New York: Longman Press.

Edeiman, Murray. 1960. “Symbols and Political Quiescence." American Political Science Review. 54:695-704.

______. 1964. The Symbolic Uses o f Politics. Urbana. Illinois: University of Illinois Press.

Eisenger. Peter K. 1982. “Black Employment in Municipal Jobs: The Impact o f Black Political Power." American Political Science Review. 76:330-92.

Epstein. David and Sharon O’Halloran. 1999 “A Social Science Approach to Race. Redistricting. and Representation." American Political Science Review. 93: 187- 191.

Eulau. Heinz, and Paul D. Karps. 1977. “Tne Puzzle o f Representation: Specifying Components o f Responsiveness." Legislative Studies Quarterly. 2:3.233-255.

Fenno. Richard F.. Jr. 1990. Watching Politicians: Essays on Participant Observation. Berkeley: IGS Press.

______1978. Home Style: House Members in Their Districts. Glenview, IL: Scott. Foresman and Company.

1973. Congressmen in Committees. Boston: Little Brown.

152

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Fiorina. Morris P. 1974. Representatives, Roll Calls and Constituencies. New Haven: Yale University Press.

Fleisher. Richard. 1993. "Explaining the Change in Roll-Call Voting Behavior of Southern Democrats." Journal o f Politics 55:327-341.

Frymer. Paul. 1999. Uneasy Alliances: Race and Party Competition in America. Princeton: Princeton University Press.

Giles. Micheal W. and Kaenan Hertz. 1994 "Racial Threat and Partisan Identification " American Political Science Review.

Gill. LeVeme McCain. 1997. African American Women in Congress: Forming and 7 runsforming History. New Brunswick. NJ: Rutgers University Press.

Gilliam. Franklin D.. Jr. 1996. ‘‘Exploring Minority Empowerment: Symbolic Politics. Governing Coalitions, and Traces o f Political Style in Los Angeles."American Journal o f Political Science. 40:1.56-81.

Glaser. James M. 1996. Race. Campaign Politics, and the Realignment in the South. New Haven: Yale University Press.

Grofman. Bernard. Robert Griffin, and Amihai Glazer. 1992 "The Effect o f Black Population on Electing Democrats and Liberals to the House o f Representatives." Legislative Studies Quarterly. 17:365-379.

Grofman. Bernard. Lisa Handley, and Richard Nienri. 1992. Minority Representation and the Quest fo r Voting Equality. NY: Cambridge University Press.

Grofman. Bernard. 1982. “Should Representatives Be Typical o f Then: Constituents? in Grofman. Bernard. Arend Lijphart. Robert McKay, and Howard Scarrow. eds.. Representation and Redistricting Issues. Lexington. MA: Lexington Books.

Guinier. Lani. 1994. The Tyranny o f the Majority. NY: The Free Press.

______1992. "Voting Rights and Democratic Theory: Where Do We Go From Here?" In Grofman and Chandler, eds. Controversies in Minority Voting. Washington. D.C.: The Brookings Institutions.

Gurin. Patricia. Shirley Hatchett, and James S. Jackson. 1989. Hope and Independence: Blacks’ Response to Electoral and Party Politics. New York: The Russell Sage Foundation.

153

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Hacker. Andrew. 1992. Two Nations: Black and White. Separate, Hostile, and Unequal. New York: Scribner Press.

Hall. Richard L. 1996. Participation in Congress. New Haven: Yale University Press.

Hall. Richard L., and Colleen Heflin. 1994 “The Importance o f Color in Congress: Minority Members. Minority Constituencies, and the Representation o f Race in the U.S. House." Paper presented at the 1994 annual meeting o f the Midwest Political Science Association. Chicago.

Havnie, Kerry L. 2001. African American Legislators in the American States. New York: Columbia University Press

Henry. Charles. 1977. “Legitimizing Race in Congressional Politics."American Politics Quarterly. 5 : 149-17t>. ^

Hero. Rodney E. and Caroline J. Tolbert. 1995. “Latinos and Substantive Representation in the U.S. House o f Representatives: Direct. Indirect, or Nonexistent0" American Journal o f Political Science. 39:640-651

Hero. Rodney E. 1992. Latinos and the U.S. Political System: Two-tiered Pluralism. Philadelphia: Temple University Press.

Hill. Kevin. 1995. “Does Creation o f Majority Black Districts Aid Republicans?" Journal o fPolitics. 57:384-401.

Hill. Kim Quaile and Patricia M. Hurley. 2002. “Symbolic Speeches in the U.S. Senate and their Representational Implications." Journal o f Politics. 64:219-231.

Hinkley. Barbara. 1990. The Symbolic Presidency. New York: Routledge. Chapman, and Hall, Inc.

Hochsch3d. Jennifer. 1995. Facing Up to the American Dream: Race. Class, and the Soul o f a Nation. Princeton: Princeton University Press.

Hutchings. Vincent L.. Harwood K. McClercking. and Guy-Uriel Charles. 2001. “Congressional Representation o f Black Interests: Ideological Commitment or the Electoral Connection.” Unpublished manuscript.

Jewell. Malcolm E. 1985. “Legislators and Constituents in the Representative Process." in Loewenberg. Gerhard. Samuel C. Patterson, and Malcolm E. Jewell, eds.The Handbook o f Legislative Research. Cambridge: Harvard University Press.

154

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Johnson. James. 2000. “Wbv Respect Culture?^American Journal o f Political Science. 3: 405-418.

______. 1993. 'Is Cheap Talk Really Cheap? Prompting Conversation Between Critical Theorv and Rational Choice.- American Political Science Review. 87: 74-86.

Johnson. James B. and Philip E. Secret. 1996. “'Focus and Style o f Representational Roles of Congressional Black and Hispanic Caucus Members.-Journal o f Black Studies. 26: 245-273

Jones. Charles E. 1987a “Testing a Legislative Strategy: The CBC's Action-Alert Communication Network.- Legislative Studies Quarterly. 12: 521-535

Jones. Mack. 19/6 "Black Office-holding and Political Development m the Rural South. ” Review o f Black Political Economy. Summer.

______1978. "Black Political Empowerment in Atlanta: Myth and Reality. ’’ Annals of American Acaaemy o fPolitical and Social Science. 439: 90-117.

kamig Albert, and Susan Welch. 1980. Black Representation and Urban Policy. Chicago: University o f Chicago Press.

Key. V.O. 1949. Southern Politics. New York: Vintage Press.

Donald Kinder and Nicholas Winter. 2001. “Exploring the Racial Divide: Blacks. Whites, and Opinion on National Policy.- American Journal o fPolitics. 45:439- 453.

King. Gary. 1989. Unifying Political Methodology: The Likelihood Theory o f Statistical Inference. Cambridge: Cambridge Univarsity Press.

Kingdom John. 1981. (1989. 3rd ed.) Congressmen's Voting Decisions. Ann Arbor: University of Michigan Press.

______.1984. (1995.2nd ed) Agendas. Alternatives, and Public Policies. Boston: Little. Brown.

KimbaIIT David C.. and Samuel C. Patterson. 1995. “Living Up to Expectations: Public Attitudes Toward Congress.- Presented at the 1995 annual meeting o f the American Political Science Association. Chicago. IL.

155

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Kousser. J. Morgan. 1992. “The Voting Rights Act and the Two Reconstructions.” In Controversies in Minority Voting: The Voting Rights in Perspective, ed. Bernard Grofman and Davidson Chandler. Washington. D.C.: Brookings Institution.

Kuklinski. James H. 1979. “Representame-Lonstituency Linkages: A Review Article.” Legislative Studies Quarterly. 4:1.121-140.

Leadership Conference on Civil Rights. 1996. "Civil Rights Voting Record for the 104th Congress.” (November.)

Lublin, David Ian. 1997. The Paradox o f Representation: Racial Gerrymander and Minority Interests in Congress.. Pnnceton: Princeton University Press.

Lusane. Clarence. 1994. “Unity and Struggle The Political Behavior o f African American Members of Congress." Black Scholar . 24:16-29.

Mansbridge. Jane. 1999. “Should Blacks Represent Blacks and Women Represent Women? A Contingent. ‘Y es’"Journal o f Politics. 61: 628-657.

Marable. Manning. 1985. Black American Politics. London: Verso Press.

Mavhew. David R. 1974. Congress: Tne Electoral Connection. New Haven: Yale University Press.

Mavhew. David R. 2000. America s Congress. Activism in the Public Sphere, James Madison through Newt Gingrich. New Haven: Yale University Press.

McAdam. Doug. 1982. Political Process and the Development o f Black Insurgency, 1930-1970. Chicago, University of Chicago Press.

McClain, Paula D. 1996. “Black Politics at the Crossroads or in the Cross-Hairs?" American Political Science Review. 90:867-873.

McClain. Paula D. and John A. Garcia. 1993. “Expanding Disciplinary Boundaries: Black. Latino, and Racial Minority Groups in Political Science.” InPolitical Science: The State o f the Discipline, ed. Ada Finifter- Washington, D.C.: American Political Science Association.

McDermott, Monika. 1996. “Vonng for Women and Black Candidates: Race and Gender Cues in Low-Information Electoral Contests.” Paper delivered at the 1996 annual meeting o f the American Political Science Association. San Francisco. California.

156

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Mfller. Cheryl- 1990- “Agenda Setting by State Legislative Black Caucuses: Policy Priorities and Factors o f Success." Policy Studies Review. 9: 339-354.

Miller. Warren E-. and David Stokes. 1963. "Constituency Influence on Congress.” American Political Science Review. 57:56.

Moms. Alrion. 1992. “The Future o f Black Politics: Substance versus Process and Formality." National Political Studies Review. 3:168-174.

Narcisse. Eric John. 1997. “Reanalyzing Black Substantive Representation: Whither the Interest Group Scores?- Paper presented at the 1997 Annual Meeting o f the American Political Science Association. Washington. D.C

Nie. Norman. Sidney Verba, and John Petrocik. 1979. The Changing American Voter Cambridge: Harvard University Press.

Nye. M ary Alice and Charles S. Bullock III. 1992. "Civil Rights Support: A Comparison of Southern and Border State Representatives." Legislative Studies Quarterly 17:81-94.

Overby. Marvin I and Kenneth M. Cosgrove. 1996. "Unintended Consequences? Racial Redistrictine and the Representation o f Minority Interests." Journal or Politics. 58:540-550.

Parker. Glenn R. and Roger H. Davidson. 1979. “Why do Americans Love Their Congressmen So Much More Than Their Congress0- Legislative Studies Quarterly. 4: 53-60.

Parker. Frank R. 1990. Black Votes Count. Chapel Hill: University o f North Carolina Press.

Patterson. Samuel C.. Randall Ripley, and Stephen V. Quinlan. 1992. “Citizens Orientations Toward Legislatures: Congress and the State Legislature."Western Political Quarterly. 45(2): 315-38.

Phillips. Anne. 1995. The Politics o fPresence. Oxford. England: Clarendon Press.

Pinderhughes. Dianne. 1987. Race and Ethnicity in Chicago Politics. Chicago: University or Illinois Press.

Pitkin- Hannah F. 1967. The Concept o f Representation. Berkeley: University of California Press.

157

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. PowelL G. Bingham. 2001. “Democratic Representation: Two Contributions trom Comparative Politics." Paper presented at the 2001 annual meeting of the American Political Science Association. San Francisco.

Schiller. Wendy J. 2000. Farmers and Rivals Representation in U.S. Senate Delegations. Princeton: Princeton University Press

______. 1993. Bill Sponsorship in the United States Senate: Opportunity and Constraint. Ph.D. Dissertation. University o f Rochester.

______. 1995a. “Senators as Political Entrepreneurs: Using Bill Sponsorship to Shape Legislative Agendas.”American Political Science Review. 39:186-203.

______. 1995b. “Constituent Expectations and Agenda Formation in the United States Senate." Presented at the 1994 annual Meeting o f the Midwest Political Science Association. Chicago. Illinois.

Schneier. Edward V. and Bertram Gross. 1993. Legislative Strategy: Shaping Public Policy. New York: St. Martin's Press.

Sipeiman. Carol K„ Lee Sigeiman. Barbara Walkosz. and Michael Nitz. 1995. “Black Candidates and White Voters: Understanding Racial Bias in Political Perceptions.” American Journal o f Political Science. 39:243-65.

Sinclair. Barbara and David Brady. 1987. “Studymg Members o f the United States Congress.” in Movser. George, and Margaret Wagstaffe. eds.Research Methods for Elite Studies. London: Allen&Unwin.

Sinclair-Chapman. Valeria. 1998. “Promises or Progress^ Representing Black Interests in Congress.” Paper presented at the 1998 annual meeting o f the Midwest Political Science Association. Chicago. Illinois.

______1996. “Symbols and Substance: Constituents. Legislators, and Legislative Work.” Papa: presented at the 1996 annual meeting o f the American Political Science Association. San Francisco. California.

Singh. Robert. 1998. The Congressional Black Caucus: Racial Politics in the United Slates Congress. Thousand Oaks. CA: Sage Publications.

Smith. Robert C. 1996. We Have No Leaders. African Americans in the Post-Civil Rights Era. Albany: State University o f New York Press.

______. 1990. "Recent Elections and Black Politics: The Maturation or Death o f Black Politics?" Political Science and Politics. 22:1.160-163. 158

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Smith, Steven S. 1989. Call to Order: Floor Politics in the House and Senate. Washington. D.C.: Brookings Institution.

Sniderman, Paul M . and Piazza. Thomas. 1995. The Scar o f Race. Cambridge: The Belknap Press

Sniderman. Paul M„ Philip Tetlock, Edward Carmines, and Randall Peterson. 1993. “The Politics o f the American Dilemma: Issue Pluralism." InPrejudice. Politics, and the American Dilemma, edited by Paul M. Sniderman. Philip Tetlock. and Edward Carmines. Stanford, CA: Stanford University Press.

Sonenshein. Raphael J. 1995. Politics in Black and White: Race. and Power in Los Angeles. Princeton. NJ: Princeton University Press.

Swain. Carol M. 1993. Black Faces. Black Interests: The Representation of African Americans in Congress Cambridge:. Harvard University Press

______1992. "Changing Patterns o f Africaiv American Representation in Congress. In Tne Atomistic Congress: An Interpretation of Congressional Change. Allen D. Hertzke and Ronald M. Peters. Jr.. eds. Armonk. NY: M.E. Sharpe. Inc.

Tate. Katherine. 1994. From Protest to Politics: The New Black Voters in American Elections. NY: Russell Sage Foundation.

Tate. Katherine and Gloria J. Hampton. 2000. “Changing Hearts and Minds: Racial Attitudes and the 1964 Civil Rights Act." InLegacies o f the 1964 Civil Rights Act. Bernard Grofman. ed. Charlottesville: University o f Virginia Press.

Tate, Katherine. 2002. Black Faces in the Mirror: African Americans and Their Representatives in Congress. Princeton: Princeton University Press.

Themstrom. Abigail M. 1987. Whose Votes Count? Affirmative Action and Minority Voting Rights. Cambridge: Harvard University Press.

Thomas. Sue. 1994. How Women Legislate. Oxford: Oxford University Press.

Tien, Charles and Dena Levy. 1998. “Asian-American and Hispanic Representation in the U.S. House o f Representatives." Paper presented at the 1998 annual meeting o f the Midwest Political Science Associauon. Chicago. Illinois.

Verba, Sidney, and Norman Nie. 1972. Participating in America: Political Democracy and Social Equality. NY: Harper and Row. 159

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Vega. Auturo and Juanita M. Firestone. 1995. “The Effects o f Gender on Congressional Behavior and the Substantive Representation o f Women." Legislative Studies Quarterly- 20:213-222.

Walters. Ronald. 1992. “Two Political Traditions: Black Politics in the 1990s.”The National Political Science Review. 3:198-207.

Walton. Hanes. 1985. Invisible Politics. Black Political Behavior. Albany: State University o f New York Press

Weissberg. Robert. 1978. “Collective versus Dyadic Representation in Congress." American Political Science Review. 72: 535-547.

Welch. Susan, and John R. Hibbinc 1984 “Hispanic Representation in the U.S. Congress.”Social Science Quarterly. 65:328-35.

Williams. Linda F. 1989. “White-Black Perception o f the Electabilitv o f Black Political Candidaies.” National Political Science Review. 2:45-64.

Williams. Melissa S. 1998. Voice, Irusi. and Memory: Marginalized Groups and the Failings Liberal Representation. Pnnceton: Princeton University Press.

Whitby. Kenny J.. and George A. Krause. 1998. “Heterogeneity in Civil Rights Voting Behavior: Representation o f Minority Policy Interests in Congress.” Paper presented at the 1998 annual meeting o f the Midwest Political Science Association.

Whitby. Kenny J„ and Franklin D. Gilliam Jr. 1991. “A Longitudinal Analysis of Competing Explanations for the Transformation o f Southern Congressional Politics.” Journal o f Politics. 53:504-518.

Whitby. Kenny J. 2002. “Bill Sponsorship and Intraracial Voting Among African American Representatives.” American Politics Research. 30: 93-109.

______1997. The Color o f Representation: Congressional Behavior and Black Interests. Michigan: University o f Michigan Press.

______1987. "Measuring Congressional Responsiveness to the Policy Interests of Black Constituents." Social Science Quarterly. 68:1367-377.

______1985. “Voting Behavior o f Southern Congressmen: The Interaction o f Race and Urbanization.” Legislative Studies Quarterly. 10:505-17.

160

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ZaHer. John. 1992_ The Nature and Origins o f Mass Opinion* New York: Cambridge University Press.

161

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.