Independent redistricting speaks to how district lines are drawn, but it does not change the more important factor of what election method is used to cast and count votes in those districts. Even if typically a worthwhile reform to address the conflict of interest that comes with politicians picking their own voters, independent redistricting alone cannot guarantee competitive elections, partisan fairness, racial fairness, geographic coherence and accountable leadership at the same time. Competition, for example, requires districts with a narrow partisan division, which almost certainly means that racial minorities will not have the power to consistently elect candidates of choice as can be required under the Voting Rights Act. Competitive districts also mean that overall partisan balance can swing wildly based on a small shift in votes statewide. And given that most areas have natural partisan leanings, drawing competitive districts makes it difficult to follow traditional criteria like compactness and maintenance of local political lines. Perhaps due to these inherent conflicts, independent redistricting has had minimal impact on electoral competition, and a mixed impact on fair partisan and racial representation. We need voting system reforms to enhance redistricting reform.
Independent redistricting alone has not increased competitiveness nor representativeness
Arizona and Iowa are often highlighted by redistricting reformers for having adopted independent redistricting for congressional and state legislative elections, but neither state has particularly competitive elections nor the increased accountability and fairer representation sought by reformers. In Arizona Congressional elections from 2002 to 2006, 20 of 24 races have been won by landslide margins of more than 20% (a generous metric) since independent redistricting in 2001, and fully three of the four non-landslides were in 2006, a breakthrough year for Democrats in terms of the popular vote. Moreover, since 2001, only one candidate has defeated an incumbent. No other incumbent has come close to losing. None of its 30 state senate seats were competitive in 2004, and almost half were not even contested.
Even though Iowa is unusually balanced in its partisan division across much of the state, all of its U.S. House incumbents in 2004 were re-elected, with an average margin of victory of 18%. Iowa’s incumbency rate indeed has been nearly 98% since the adoption of independent redistricting. In addition, no woman has been elected to Congress in either state since adoption of independent redistricting, and women candidates have fared less well in state elections after each new redistricting plan. Arizona Latinos are suing their state plan, which has been in continuous litigation since 2002.
Competitiveness and fair representation are mutually exclusive
Because voter choice and fair representation are mutually exclusive in a single-member district, there is no way to achieve both of these critically important goals consistently, even with independent redistricting. First, the imbalance of partisan divisions in nearly all states makes it impossible to make every district competitive. But even if one could make all districts perfectly competitive, that would mean 50% of each district would favor one major party, and 50% would favor the other. The elections might be very competitive (at least in races without popular incumbents), but nearly half of the population would not be represented after each election – and the winners easily could all be in one major party. Current legislative districts are largely uncompetitive, but more voters are represented by someone of similar views. Within our current single-member district system, no set of district lines can satisfy both of these goals. We must look beyond this limited model.
Proportional voting methods as a means to achieve all goals
We need proportional voting methods as part of redistricting reform. Following are two examples of modest proportional voting systems that would greatly enhance redistricting reform.
Superdistricts: Multi-seat districts with a proportional voting method
To achieve increased competition, fair partisan and racial representation, higher quality campaigns, and higher voter turnout, states could elect legislators in multi-seat "superdistricts" with a proportional voting method – ideally a candidate-based system like the choice voting method used in Cambridge city elections. Each district would elect several seats, which would enable representatives to speak for different communities of interest within the superdistrict. Votes would be tabulated using a proportional voting method that establishes a victory threshold based on the number of seats to be elected. In a five-seat district, for example, any candidate winning about one-sixth of the vote would win one seat, and like-minded voters with 51% support would elect three candidates.
Illinois had a widely praised experience with a proportional voting method based on voting for individual candidates called cumulative voting; used in three-seat districts for its state House of Representatives from 1870 until 1980, the system was repealed in an initiative focused on reducing the size of the legislature by a third. Illinois’ experience likely would have been all the better by accommodating increased choice and encouraging coalition-building, through choice voting.
Superdistricts would provide more reliably accurate representation of a community's racial, gender, and political diversity while boosting voter choice and allowing for geographically coherent districts. A criteria-driven independent redistricting commission should map out each superdistrict.
Districts plus: Single-member districts with additional accountability seats
Developed by reformers in Michigan, the district plus proposal involves adding “accountability seats” to the legislature to compensate for partisan unfairness in district elections. For example, in a 100-seat legislative chamber, 20 accountability seats might be added to 80 district seats. Voters would have two votes: one in their local district and one to determine the legislative leadership by using the 20 accountability seats to ensure the party with majority of votes wins a majority of seats.
Voters would elect a representative in the 80 local district races, just as they do now, and then cast an “accountability vote” to determine overall control of the legislature. The 20 accountability seats would be allocated to ensure that a party’s share of seats in the legislature corresponded to its share of the accountability vote. For example, if a party won 54% of the total vote, but only 39 of the 80 district seats, it would get 15 of the 20 accountability seats to ensure an accurate share of 54 seats. A party would need to win a minimum share of accountability votes to win seats, such as 5%. A party could award accountability seats in different ways. One approach would be to have the parties nominate one candidate for each four adjoining districts. A party would publicly declare before the election how it would award any accountability seats it won. It could choose to elect candidates from the four-seat superdistricts where that party won its highest share of accountability votes. It might choose instead to first elect nominees from superdistricts where the party won votes, but no seats; this latter approach would ensure more voters have two-party representation.
In the district plus system, no voters would be left in an election where their participation was effectively meaningless. Even if living in a lopsided partisan district, the prospect of helping one's preferred party would provide (1) voters with an incentive to turnout and (2) parties an incentive to campaign vigorously everywhere. Furthermore, the state legislative leadership would be directly accountable to the voters; if a majority wanted new leadership, they could elect a new speaker.
Superdistricts and district plus show how proportional voting systems can enhance independent redistricting commissions to achieve elections that serve the public interest. At the very least, we suggest that independent commissions be empowered to consider such options. As public dissatisfaction over legislative elections grows and opportunities for serious reform increases, more comprehensive and sweeping proposals must be included in the discussion if reformers are serious about providing voters with competitive choices and fair representation.
July 11, 2007