It Takes a Black Candidate

• Author: Paru Shah
• DOI: 10.1177/1065912913498827
• Date: 01 June 2014
• Published date: 01 June 2014
• Subject Matter: Articles

498827PRQXXX10.1177/1065912913498827shah
research-article2013
Article
Political Research Quarterly
2014, Vol. 67(2) 266 –279
It Takes a Black Candidate: A Supply-Side © 2013 University of Utah
Reprints and permissions:
sagepub.com/journalsPermissions.nav
Theory of Minority Representation
DOI: 10.1177/1065912913498827
prq.sagepub.com
Paru Shah1
Abstract
The ongoing underrepresentation of racial and ethnic minorities in most levels of office continues to warrant our
attention. However, scholars have only focused on what factors contribute to the electoral success of minorities,
without attention to a vital precursor—the supply of minority candidates. Using data from the Local Elections in America
Project, this study provides one of the first glimpses into the supply side of minority representation, detailing how
demographic, electoral, and political factors affect the likelihood a black candidate is on the ballot, and the subsequent
impact on the likelihood of a black candidate winning.
Keywords
descriptive representation, black elected officials, minority candidate supply
Introduction
close to 20 percent, with the majority of racial minorities
serving in subnational and substate legislative offices.
By some accounts, the glass is half full. The racial and
What explains this gap in representation? Since pas-
ethnic complexion of elected officials in the United States
sage of the Voting Rights Act, social scientists have
has changed dramatically over the past fifty years, with
investigated the question of why racial and ethnic minori-
increasing numbers of blacks, Latinos, and Asians taking
ties fail to achieve representational parity from a demand
office. For example, the 112th U.S. Congress includes two
perspective—how do racial attitudes and behaviors
Asian American and two Latino Senators, as well as forty-
among black and white voters restrict minority office-
four black and thirty Latino House members. According
holding? With few exceptions (see, for example Canon
to the National Conference of State Legislatures most
1999), none of this previous work has focused specifi-
recent data, nonwhite state legislators numbered above
cally on the question of minority candidate supply. That
1,000 in January 2009, comprising more than 13 percent
is, how often is lack of representational parity due to the
of all state legislators. And although the first black elected
defeat of the minority candidate(s) versus the absence of
to the mayor’s office of a major city did not occur until
minority candidates in the first place? Answering this
1967, today there at least 500 black mayors and 300 Latino
question is of central importance to our understanding of
mayors. In short, there is ample evidence that minority
minority representation for a number of reasons.
candidates are winning more elections in more places than
First, to date the vast majority of research has exam-
ever before.
ined descriptive representation focusing exclusively on
However, by other accounts, the glass is still empty.
inequities in representation based solely on the out-
These same fifty years have witnessed tremendous
come—did a minority candidate win office, why or why
growth in racial and ethnic minority populations, and
not? (see, for example, Cameron, Epstein, and Halloran
thus, these gains in representation are interpreted as insuf-
1996; Casellas 2009; Marschall, Ruhil, and Shah 2010;
ficient, reflecting the continued underrepresentation and
Trounstine and Valdini 2008). Importantly, of the many
perhaps disenfranchisement of racial and ethnic minori-
answers offered up in response to these questions one
ties from the political process. Indeed, as the racial and
ethnic minority population edges closer to 50 percent of
the U.S. population, the gap in representational disparity
1University of Wisconsin–Milwaukee, Milwaukee, WI, USA
grows. Although accurate numbers at all levels of gov-
ernment are hard to ascertain,1 data from the Gender and
Corresponding Author:
Paru Shah, Department of Political Science, University of Wisconsin–
Multicultural Leadership project estimate the percent of
Milwaukee, P.O. Box 413, Milwaukee, WI 53201, USA.
black, Latino, Asian, and Native American legislators at
Email: shahp@uwm.edu

---

Shah
267
rarely, if ever, hears “because there were no minority can-
supply of minority candidates affects the outcome.
didates running for office.” But clearly, if no minority
Starting with what I argue are the building blocks for a
candidate is running then the likelihood of a minority
comprehensive theory of the supply-side of minority rep-
candidate winning is zero. Thus, the typical explanations
resentation, I develop a number of hypotheses about how
of why minorities fail to achieve representational parity
electoral, political, and demographic factors affect the
point to demographic, political, and electoral variables
supply of minority candidates, and the electoral out-
within a jurisdiction as the main culprits, although we
comes. Using data from the Local Elections in America
cannot discern from these studies if these variables affect
Project (Marschall and Shah 2013) for the state of
the likelihood of a minority candidate running or win-
Louisiana that includes candidate race information, elec-
ning. Moreover, because these studies lack information
tion outcomes, and voter registration rates, I compare a
on the racial composition of the candidates running for
traditional single-stage model of black representation
office, they tell us only part of the story—what factors
with the results from a two-stage Heckman’s selection
other than candidacy change the likelihood of a minority
model on the probability of a black candidate running for
candidate winning.
office, and the subsequent likelihood of a black candidate
Second, while recent theoretical advances in political
winning for school board, municipal, and county elec-
ambition and candidate emergence (see, for example,
tions between 2000 and 2010. I conclude with implica-
Lawless 2011; Lawless and Fox 2005; 2010) have
tions of these findings for the more general scholarship
renewed interest in what motivates likely candidates from
on minority descriptive representation in the United
entering the political arena, “almost no research specifi-
States.
cally addresses race or ethnicity in the candidate emer-
gence process at all” (Lawless 2011, 59) This research
Minority Descriptive Representation:
often relies on surveys of potential candidates, who can
Theories and Empirical Evidence
relate their desire to run for office and win but cannot tell
us how often this is accomplished. Nor can this research
Much of the work on minority descriptive representation
answer the question of how minority candidate supply
begins with three premises: (1) when given the choice,
and electoral outcomes are related, or how demographic
minorities will select a co-ethnic candidate, (2) institu-
and institutional features of a jurisdiction affect the likeli-
tional and structural features of a legislative body can
hood of this supply or outcomes of interest.
shape this voter strength, and (3) qualified and eligible
This paper advances our understanding of minority
minority candidates are running for office. First, if there
representation in two ways. First, I explicitly model the
is a single “truth” in studies of minority representation, it
decision to enter the electoral arena by black candidates,
is that demographics are destiny. Of the more than thirty
and, conditional on this, the likelihood of a black candi-
well-known studies of minority descriptive representa-
date winning an election. By untying the two components
tion, everyone has concluded that size of the minority
of representation, I am better able to distinguish between
population matters (see Marschall, Ruhil, and Shah 2010
the hurdles that affect black electoral success at the can-
for a review of articles), and that the likelihood of a
didate entry stage and the electoral competition stage of
minority securing office is positively related to the size of
the process. Second, I reevaluate the traditional relation-
the co-ethnic population (see, for example, Branton 2009;
ships between demographic, political, and electoral deter-
Krebs 1999; Meier et al. 2005). These studies suggest that
minants of minority representation across these two
the likelihood of minority representation is greatest in
stages, and demonstrate that distinct mechanisms operate
“majority–minority” places, or cities/districts that have a
at each stage. The results strongly support the contention
greater than 50 percent minority voting population.
that the underrepresentation of blacks is driven by con-
Second, studies implicate at-large and multimember dis-
straints on their entry onto the ballot—it takes a black
tricting practices as the cause of minority underrepresen-
candidate. The analysis in sum suggests much of what we
tation in elected office (see, for example, Canon 1999;
have considered to be benchmarks in theories of minority
Davidson and Korbel 1981; Engstrom 1987; Engstrom
representation, including size of the minority population
and McDonald 1986; Karnig and Welch 1982). In partic-
and electoral structures, are more crucial at the candidate
ular, at-large elections increase the costs of campaigns,
entry stage, and that we in fact know less about the...