The Determinants of Hispanic Representation in Municipal Government.

• Author: Sass, Tim R.

Tim R. Sass [*]

In recent years Voting Rights Act cases brought by Hispanic litigants challenging local election systems have proliferated. This article analyzes the effects of election systems and demographic factors on the electoral success of Hispanics in municipal elections. Ability to speak English well and educational attainment of non-Hispanics are important factors that boost Hispanic representation. On average district elections enhance Hispanic representation, though the impact of district elections varies directly with the level of residential segregation. Although some Blacks may support Hispanic candidates, I do not find any significant effects of increases in Black population on the electoral success of Hispanics.

1. Introduction

   Voting rights and fair representation of minorities have been the subject of legislative and judicial concern for over 30 years. While Black representation still holds a prominent place in the policy debate, attention has begun to focus on representation of Hispanics as well. Over the last decade, voting rights lawsuits involving Hispanic litigants have mushroomed, approaching the level of Black litigative activity. [1] Along with this upswing in litigation has come an increase in research efforts by economists and other social scientists to measure the factors affecting the representation level of Hispanics.

   This paper sheds light on three important issues involving Hispanic representation and minority voting rights litigation. [2] First, what is the level of Hispanic representation, and how is it affected by the economic and demographic characteristics of Hispanics? Second, how do different electoral systems affect Hispanic representation? In particular, do Hispanics fare better in a system of district elections as opposed to jurisdiction-wide at-large elections? Third, how does the potential voting strength of other minority groups, especially Blacks, affect the ability of Hispanics to gain office? Are Blacks and Hispanics likely to vote as a bloc in support of Hispanic candidates?

   This analysis of Hispanic representation has potentially important implications for the current policy debate over minority voting rights and the creation of race- or ethnic-based election districts. To anticipate, I find that previous studies, which ignore citizenship, have significantly understated the degree of Hispanic representation. The results also indicate that mandating district elections is not a panacea for promoting the election of Hispanics. Although I do find evidence that district elections currently increase Hispanic representation on average, the efficacy of district elections varies directly with the level of residential segregation. Furthermore, the empirical results show that increases in the Black population of a city do not have a significant effect on Hispanic representation. This suggests that tenting Hispanics and other racial or ethnic minorities as a single cohesive voting group is problematic.

2. A Brief History of Hispanic Voting Rights Litigation [3]

   When the Voting Rights Act [4] was passed in 1965, the focus was on enfranchising southern Blacks. The original Act banned literacy tests and required preclearance of election system changes in seven southern states. Following pressure from Hispanics, the Act was modified in 1975 to add the heavily Hispanic states of Texas and Arizona to the list of "targeted" states.

   As was the case for Blacks, the emphasis of Hispanic voting rights litigation soon became the dismantling of at-large election systems that allegedly "diluted" minority votes. It was claimed that at-large elections, which submerge minorities within a majority Anglo jurisdiction, effectively block the election of minority candidates since voting is racially and ethnically polarized. The proposed remedy was the establishment of district elections with district boundaries drawn to ensure one or more "majority-minority" districts.

   The first major vote dilution case involving Hispanics was Graves v Barnes (343 F. Supp. 704 [1972]), where the plaintiffs successfully challenged multimember countywide legislative districts in two Texas counties. This victory led to a surge in voting rights litigation [5] and a shift from citywide to district-based municipal elections. [6]

   The successful campaign to eliminate at-large voting systems was temporarily derailed in the case of City of Mobile v Bolden (446 U.S. 55 [1980]). In Bolden, a divided Court ruled that proof of discriminatory intent was required to find at-large voting systems illegal. In response to the Bolden decision, Congress amended the Voting Rights Act in 1982 and established a "results test" whereby an at-large voting scheme could be found illegal if it had the effect of reducing the opportunity for minorities to elect representatives of their choice.

   The Supreme Court first interpreted the 1982 amendments in the case of Thornburg v Gingles (478 U.S. 30 [1986]). In Gingles the Court laid down three necessary conditions for establishing a voting scheme violates the law: (i) the minority group must be sufficiently large and geographically compact that it would constitute a majority in at least one single-member district, (ii) the minority group must show that it is politically cohesive, and (iii) the Anglo majority must vote sufficiently as a bloc so that candidates preferred by the minority are usually defeated.

   Given the first Gingles criterion of sufficient size, an important issue in voting rights cases has been whether Hispanics can be combined with other minorities when analyzing claims of vote dilution. While, conceptually, the law allows for different minority groups to be combined into a single protected group, [7] empirically, there must be evidence that different minority groups vote as a bloc to pass the second Gingles test of "political cohesion." [8] The third Gingles test, consistent defeat of candidates preferred by the minority, is often evaluated by comparing the proportion of Hispanic elected officials to the proportion of Hispanics in the population. These key issues, the representation of Hispanics and whether Hispanics and Blacks typically vote as a cohesive bloc, will be explored in the empirical analysis below.

   Most recently, the Court has taken a dim view of efforts to establish district election boundaries based on race or ethnicity. In the most recent Supreme Court decision involving Hispanic voting rights, Bush v Vera (116 S. Ct. 1894; 1996 U.S. LEXIS 3882), the Court found that race and ethnicity was the predominate factor in creating oddly shaped congressional districts in Texas. The Court ruled that absent a compelling state interest, the districts violated the equal protection clause of the fourteenth amendment. What effect this decision will have on efforts to promote district elections with majority Hispanic districts remains to be seen.

3. Summary and Critique of Previous Research

   The timing and focus of research on Hispanic representation parallels the litigative efforts of Hispanic voting rights advocates. The earliest studies date back to the mid-1970s, when the Voting Rights Act was extended to language minorities and the pace of Hispanic voting rights litigation accelerated. Given the litigative emphasis on vote dilution, research has focused on the effect of district elections on Hispanic representation.

   The first set of studies (MacManus 1978; Taebel 1978; Karnig and Welch 1979) analyze cross-sectional data on Hispanic representation in the 1970s. Using correlation analysis, all three of the studies find a weak positive relationship between district elections and Hispanic representation. In addition to the use of district elections, Taebel finds Hispanic representation to have a small positive correlation with the number of seats on the city council.

   A trio of more recent studies (Davidson and Korbel 1981; Polinard, Wrinkle, and Longoria 1991; Brischetto et al. 1994) conduct longitudinal analyses of Texas cities that experienced changes in their election systems. All three studies find that shifting from at-large or mixed election systems to district elections increased Hispanic representation. However, each of the studies utilize relatively small samples and report only comparisons of means with no tests of significance.

   The most sophisticated analyses of district elections and Hispanic representation to date are four recent studies by Bullock and MacManus (1990), Welch (1990), Zax (1990), and Alozie and Manganaro (1993). Each of these studies employs regression analysis on a cross-section of cities to estimate the determinants of Hispanic representation. Their basic methodology is to regress the percentage of Hispanics on the city council against a constant and the percentage of Hispanics in the population. The coefficient on population percentage is then the marginal effect of population on council membership percentage. If the constant term is near zero, the coefficient on the Hispanic population percentage can also be interpreted as the average effect of population on council membership or the "representation ratio."

   Results of the regression-based studies of Hispanic representation are summarized in Table 1. Theory would suggest that if preferences of both candidates and voters are correlated with ethnicity, then district elections, which typically include districts with above-average proportions of Hispanics, would increase the probability of Hispanic candidates being elected to office. However, the consensus of extant empirical studies is that district elections have little if any effect on Hispanic representation at the municipal level. The authors almost uniformly attribute the apparent ineffectiveness of district elections to the intracity geographic dispersion of Hispanics (which would diminish the difference between the Hispanic population of districts vs. the city as a whole). However, none of the studies directly measures...