Article | Fall 2005
Rethinking Voting Rights: Emerging Tradeoffs
On August 6, 2005, thousands of marchers flooded the streets of Atlanta to commemorate the 40th anniversary of the Voting Rights Act. When the bill was signed into law by President Lyndon B. Johnson in 1965, it was intended to redress patterns of disfranchisement that continued to persist after the ratification of the 15th Amendment. Several of the Act's key provisions are slated to expire in 2007 and are at the center of heated debates among the legislature, scholars, and the general public.
The issue is not whether these provisions should be renewed, but how they should be implemented to match their stated intent, according to political science professors David Epstein and Sharyn O'Halloran. As the directors of the Center on Political Economy and Comparative Institutional Analysis (COPECIA), Epstein and O'Halloran carry out research on the origin and evolution of political institutions and their impact on economic performance and policy outcomes. COPECIA is one of 15 centers and major projects housed at ISERP.
Their current research examines the legal, political, and racial implications of Section 5, the most controversial component of the landmark law. Under Section 5, areas of the country with the worst history of discriminatory electoral practices—states like Alabama, Georgia, Mississippi, Texas, and municipalities like New York City—are barred from passing any new election laws until the plan is "precleared" by the federal government. "Preclearance" is granted by the Attorney General or District Court only after they determine that the proposed plan will not have a "retrogressive," or discriminatory, effect on the voting power of the minority electorate.
For decades, interpretation of retrogression and the application of Section 5 were relatively straightforward. The Justice Department considered the viability of new election laws with respect to descriptive representation alone-that is, solely in terms of the opportunity for minority voters to elect candidates of their choice. But when the Supreme Court approved Georgia's plan to redraw its 56 state senate districts in Georgia vs. Ashcroft (2003), the ruling challenged the traditional bounds of Section 5 in fundamental ways.
The Georgia plan was first devised following the 2000 census, during a time when the Assembly and Senate both possessed Democratic majorities. "The governor, Roy Barnes, was a Democrat as well," Epstein and O'Halloran explain, "and he led the charge to construct a districting plan that would advantage his party in the upcoming 2002 elections." The plan proposed to "unpack" many of the heavily-democratic districts and create more districts with 25-40% black voter age population, so called "influence districts." As a result, some districts with black populations above 55% or 60% were brought down close to the 50% mark.
According to many observers, the Ashcroft ruling violated Section 5 because it diluted minority voting power in districts where blacks constituted the majority. One scholar denounced the decision as a step towards "gutting" Section 5 preclearance altogether; others, including an ACLU official, claimed the decision would make black citizens second-class voters who could influence the election of white candidates but would be unable to elect candidates of their own race. Yet others like Harvard professor Henry Louis Gates viewed the situation more favorably, arguing that the loss of descriptive representation would be compensated for by a growth in substantive representation, with lawmakers as a whole representing black constituents' interests better overall. "Blacks were winning battles but losing the war as conservative Republicans beat white moderate Democrats," he explains.
Figure 1: Probability of Election vs. BVAP with high electoral polarization
Ashcroft was controversial because it hinged on a number of tradeoffs that broke from historical interpretations of Section 5. One such tradeoff accompanied a change in Section 5's scope. The previous district-by-district assessment of retrogression was replaced with a statewide approach, where losses in one part of the state could be recouped by gains elsewhere.
A second tradeoff mediated between descriptive and substantive representation. Was it more important for minorities to be able to choose their candidates, or for their substantive interests to be best represented regardless of whether they could elect the specific candidate they preferred?
Professors Epstein and O'Halloran demonstrate that this choice was not always so stark. "Electability is a good proxy for retrogression only if it correlates highly with other types of representation," they argue. Section 5 was originally devised under the assumption that descriptive representation is the surest, least risky way to ensure minority voters of substantive representation. And up until roughly the mid-1980s, districting schemes that maximized descriptive representation were indeed identical to those that maximized substantive representation. It is only over the past decade-and-a-half that these two objectives have diverged.
One reason for this shift is changing political conditions. With increased white crossover voting and the rise of the Republican Party in the South, it has become harder to test retrogression based on issues of descriptive representation alone. Prior to the mid-1980s, voting was highly polarized and admitted little tradeoff between districts. Consequently, black voting possibilities were easier to pin down, and retrogression easier to determine. Figure 1 shows the key threshold under this condition as 57.5% black voting age population, where minority control of an election is certain above it and extremely unlikely below it. In this world, equal opportunity (the point at which a minority candidate has an even chance of winning), electability, and minority control coincide perfectly.
However, in the current political climate, where white crossover voting is a potent factor, increases in black voters produce gradual, rather than abrupt, changes in electoral probability. Depicted in Figure 2, this model is more of a menagerie, with less certainty and greater range of voting possibilities. For instance, in "coalitional districts" (possessing between 40% and 50% BVAP), it is relatively likely that a minority-supported candidate will be elected, but it will require some white crossover voting to do so. In Ashcroft, the Justice Department complicated these designations even further with the introduction of "safe minority districts," where the likelihood of a minority candidate attaining office is considerably greater than 50%.
This menagerie of district types raises difficult questions about the tradeoff between descriptive and substantive representation and the tradeoff between districts overall. How many coalitional districts are necessary to outweigh one safely controlled district? Or what combination of coalitional and unsafe districts is needed to outweigh one safe district? Or, as the Justice Department argued, perhaps no combination of other district types could outweigh the loss of a safe one.
Furthermore, emphasizing descriptive representation could give minority voters a share of legislative seats greater than population proportion. Yet maximizing substantive representation could result in electing no minorities at all to office. Ashcroft settles on a compromise between the two, allowing for a more nuanced and evolving interpretation of retrogression.
Figure 2: Key Points and District Types
But by itself, Ashcroft does not provide a clear picture of what this compromise should look like. "Maybe a 50% chance of winning 20 districts is as good as certainty of winning 10," Epstein and O'Halloran muse. "But without a full set of election probabilities, retrogression determinations are arbitrary." The lack of transparency in measuring substantive representation presents a major impediment to the accurate assessment of retrogression. Epstein and O'Halloran propose a solution. With the use of Congressional roll call data, they are able to determine an elected representative's minority support score by evaluating how closely his or her voting record correlates with the voting records of minority legislators. Consequently, the impact of redistricting can be adequately assessed before a proposed plan is implemented.
"The election of minority-preferred representatives has ceased to be, so to speak, a black and white issue," conclude Epstein and O'Halloran. "There is no longer a magic number above which a minority can control the electoral outcomes of a district and below which they do not. Electability is now a matter of degree." These are the new conditions under which legislators will have to contemplate the renewal of Section 5 in 2007.
The complete paper, "The Evolution of Descriptive and Substantive Representation, 1974-2004" by David Epstein and Sharyn O'Halloran was presented at a conference organized by the authors and colleagues at the Russell Sage Foundation on June 24-25, including ISERP Faculty Fellows Rodolfo de la Garza, Robert Lieberman, and Robert Erikson. The report will appear in a book published by the Russell Sage Foundation.
See Also
