July 20, 2009

Pildes: The Return of Alternative Voting

Rick Pildes sends along this guest post:

    Many academics, myself included, have been supportive of greater use of alternative-voting systems, particularly in local government elections, where I believe they might be most appropriate. These systems might be worthwhile on their own merits, and they are often worth considering as remedies in cases in which the current design of elections has been found to violate the Voting Rights Act. Cumulative voting (CV) and limited voting (LV) are among these alternatives, and they are particularly attractive in VRA cases because, if designed properly, they can give minority groups a reasonable opportunity to elect candidates of choice. In CV, voters get multiple votes to cast and can spread them among candidates or concentrate them on one or more candidates. If there are five seats to fill, voters might get five votes; they can give all those votes to one candidate or distribute them among several candidates. In LV, voters get fewer votes to cast than there are seats to fill; if three seats are up, voters might have only one vote to cast. Until recently, federal courts had been somewhat reluctant to adopt these remedies in VRA cases. But a major recent District Court decision might change that, which is why I want to bring that decision to a larger audience.

    Typically in local-government cases under the VRA, the United States challenges at at-large election structure -- for a city council, a school board, a county commission, or similar entities -- for diluting the voting power of a minority group. If liability is found, the typical remedy is then to breakup this at-large election structure and create a number of single-member election districts, some of which are intentionally designed to be dominated by a supermajority of minority voters. As with all election structures, there are tradeoffs with at-large elections. They can have the effect of minimizing minority voting power, but they have the benefit of electing representatives who might be more likely to act from the perspective of the entire city or county, as opposed to representing more narrowly-sliced up geographic areas. The attraction of CV and LV is that they preserve the benefits of at-large elections while addressing the problem of minority exclusion.

    The important recent decision came in United States v. Euclid School Board, No. 08-CV-2832 (N.D. Oh. July 13, 2009). Whatever one concludes about the merits of the analysis here, I want to praise District Judge Kathleen O'Malley, about whom I knew nothing before reading this opinion. The opinion shows a superb command and depth concerning the relevant issues. Because Euclid conceded liability under Section 2 of the VRA, the question was how to change its election structures to remedy this violation.

    Euclid's elected school board had five members elected at-large with staggered terms. Every other year, voters elected either two or three board members. This Ohio city was in the midst of rapid demographic changes: the African-American total population was 7.8% in the 1980s; 16.2% in 1990; 30.5% by 2000; and 44.6% in 2005-07. In the latter period, relevant to this litigation, they compose 40.2% of the voting-age population (VAP). Until 2008, no African-American had been elected as mayor, city council, or school board member. As it always does in these contexts, the United States argued that the proper remedy was to draw five-single member districts, with one of those districts a safe, minority-controlled one (this district was designed to have a 60% VAP of African Americans). Euclid proposed cumulative or limited voting instead, while retaining the at-large, staggered-terms structure. The Court accepted the limited-voting remedy. There are four particularly interesting elements in the Court's decision:

    1. Why Limited Voting Over Single-Member Districts? The broad legal issue is how much of an "opportunity to elect" must a minority community when courts are remediating Sec. 2 violations. In the District Court's words, the United States argued that "a legally acceptable remedy is one that will necessarily result in roughly proportional representation, even if minority turnout is substantially lower than non-minority turnout." The District Court characterized Euclid's remedial standard being "that a plan is legally acceptable so long as it would provide representation were minorities to vote at the same rate as non-minorities." The Court rejected both of these standards and held that the correct one was whether a voting system ensured "genuine opportunity for all citizens."

    First, the Court held that prior law required it to accept the governmental body’s proposed remedy if that remedy was legally acceptable, even if some other remedy would be "better." Thus, the Court had to accept an alternative voting system, rather than single-member districts, if the former would correct the violation. With a three-seat election (as was upcoming), and limited voting, the Court noted that the "threshold of exclusion" was 25%, meaning if African Americans cast at least 25% of the votes and all cast their one vote for a minority candidate, that candidate would win. The Court found it reasonable to assume that this would be the case, that is, there would be a "meaningful" opportunity to elect, based on how the Court analyzed potential minority voter turnout, which was the next big issue.

    2. How Should Courts Deal with Turnout Issues in VRA Litigation Today? Because the black population had risen so dramatically in recent years, the Court had more than the usual problems struggling with that issue. Historically, the minority population had turned out at lower rates than whites. Should the Court ignore that? Should it assume that same historical pattern would continue into the future and use these relative turnout figures to project how effective the remedial, limited voting system would be? Data from 1995-2003 indicated a difference of 25 percentage points in white v. black turnout in school board elections (32.0% to 7.4%). Relative turnout was much more even, though, in some other elections not involving school boards. Or should the Court do something in the middle, between VAP and historical turnout? The Court concluded that the cases had not clearly resolved this question before; it noted that four of the five cases in the alternative-voting context had used VAP, as opposed to historical turnout figures.

    The middle option is the one the Court chose. The Court concluded: "[w]hile the effects of long-standing electoral discrimination on voter turnout are undeniable, there is assuredly some point at which potential voters must themselves come to the polls." The Court concluded it would assume minority turnout going forward would be at least 2/3rds that of white voters and that this was a reasonable baseline to use. Thus, since African Americans were currently 40% of the VAP, if they turned out at this rate, they would be above the threshold of exclusion (.67 x 40% = 27%).

    3. Why Limited Voting Over Cumulative Voting? This last question was the easiest to answer. In the cumulative voting option, each voter would have three votes to cast. In limited voting, each voter would be able to cast only one vote, even when more than one seat was up.

    First, the Euclid School Board was indifferent between the two. Second, limited voting was used throughout Ohio, cumulative voting was not used anywhere. Next, the county elections board, which administered voting in Euclid, argued that it would find it easy to implement limited voting but difficult to implement cumulative voting. In addition, the Court judged cumulative voting to be harder for voters to understand and employ, which increased the likely rate of errors.

    4. Why Maintain At-Large Elections At All? Historically, successful Sec. 2 challenges to at-large elections resulted in eliminating the at-large structure and creating single-member districts. But here, the Court explained and endorsed several legitimate reasons the School Board had for wanting to keep the at-large structure, if possible. Again, this analysis is important because it might stimulate other courts to consider maintaining at-large elections and using alternative voting systems, rather than creating single-member districts, when Sec. 2 violations have been found.

    The Court held that at-large school board elections served important values because the issues boards addressed "require district-wide support or accountability." Single-member districts risked creating "geographic partisanship" that would make the kinds of decisions boards had to make "far more difficult, and, at times, even impossible." In addition, because these are such low-turnout elections, at-large elections made it more likely a sufficient number of candidates would actually run.

Posted by Rick Hasen at July 20, 2009 09:21 AM