Arizona Defeats Top Two Primary: What's Next for Reformers?

by Drew Spencer // Published November 9, 2012


As the nation eagerly followed the incoming results of the Presidential election on Tuesday, we at FairVote also kept a keen eye on the results of a handful of electoral reform ballot measures, including Arizona's vote on Proposition 121, the Top Two primary law. We recently conducted a thorough analysis of Washington State's use of Top Two and were concerned about the impact that this proposed form of Top Two might have in Arizona. But Prop 121's defeat became apparent early in the evening, with over two-thirds of Arizona voting against it.

Currently being used in Louisiana, Washington, and California, Top Two eliminates partisan primaries in favor of having all candidates run against each other in a preliminary election, irrespective of party preference. Then, the general election in November features only the two candidates who received the most votes in the preliminary, again irrespective of political party preference.

Top Two supporters believed that taking the candidates out of partisan primaries would help take the partisanship out of the candidates, resulting in more moderate representatives responsive to all voters instead of just those within their parties. With that goal in mind, it comes as no surprise that supporters wanted the system in Arizona, which has been criticized for party polarization in its state legislature.

However, we now have clear results from three election cycles in Washington and one in California suggesting that Top Two falls well short of its stated goals. Our Washington state findings included:

  • The system limits voter choice, both by restricting the general election to only two candidates and by creating incentives to start with fewer candidates in the preliminary round.
  • Higher profile races tend to feature exactly two competitive candidates in the preliminary round - usually one Democrat and one Republican - suggesting that the parties may be narrowing the field artificially behind closed doors, while lower profile state legislative races rarely have more than two or three candidates in the preliminary at all.
  • Where larger numbers of candidates participate in the preliminary, many races (more than half in Washington) end up with voters splitting their votes among similar candidates, rendering questionable results in the general election.
  • Partisan voting patterns appear stable under Top Two: voting patterns between the preliminary and general rounds suggest that voters treat the system much like a partisan primary, with little crossover voting or support for moderates.

As far as easing partisanship, consider Washington elected governor and seven other statewide offices along with the presidential vote this year. All races were familiar Democratic vs. Republican races. Barack Obama won the state with 55%, and the Democrats won seven of the remaining eight offices, all with comparable margins less than 60% - the only apparent Republican winner just ahead with less than 51% of the vote. California had similar results, with Democrats taking over two-thirds of the state legislature, the highest margin any party has had in California for decades.

Top Two also suffers from irrational results. Consider:

  • The infamous race in California Congressional District 31: a district that is less than 30% white and only 46% Republican, but advanced two white Republicans to the general election, denying any voice in the general election to a majority of its residents. In fact, CD 31 is the most Democratic district in the nation represented by a Republican - and its winner, incumbent Gary Miller, was the candidate who apparently pushed more to the right in winning the district.
  • The California Congressional District 8 race, where the two advancing candidates each earned less than 16% of the vote, advancing instead of other candidates that may well have been more mainstream candidates.
  • The State Senate race in Washington's 38th district in 2010 where special interest groups supported both the liberal Democratic candidate and a conservative Independent to ensure that the more centrist Democratic incumbent could not advance to the general election at all.

FairVote has proposed changes to Top Two that we think would make it work far better for voters. Unfortunately, those suggestions were not incorporated into Arizona Prop 121.

Arizona has been open to election law reform in the past. They passed an aggressive form of campaign finance reform in 1998 and created an independent redistricting commission in 2000. Their rejection of Prop 121 should therefore be understood as a rejection of Top Two, not of change through electoral reform in general.

In that respect, Arizona's rejection of Top Two resembles the rejection by Ohio this year of redistricting reform and redistricting plan referenda in California and Maryland. These are reforms that have been tried elsewhere, but not to great effect, and they do not seem to be giving voters the sorts of elections they want - with real competition, better representation and more candidates responsive to all of their constituents, not just funders and party insiders.

But electoral reform can address these issues. For state legislatures, FairVote recommends American, candidate-based forms of proportional representation we call "fair voting." By expanding current districts to proportionally elected multimember districts, we can make redistricting obsolete while virtually guaranteeing to voters the election of candidates fairly representing the right, left, and center of those larger geographic areas. Our model of fair voting for the United States Congress demonstrates how such systems can elect moderate rural Democrats and urban Republicans while giving voters of all stripes real representation.

This has been tried at the state level in Illinois, where for over a century state legislative candidates were elected from three-seat districts. The result, according to a 2001 report by a commission headed by former Republican governor Jim Edgar and former federal judge Abner Mikva, was greater voter choice, easier access by candidates, more proportional representation by party, more candidate independence from party leaders, and better efforts at statewide consensus. Fair voting at a city level, including in New York City during the La Guardia era, also had positive impacts.

As states grow tired of the same reforms, which in many ways continue to disappoint, fair voting plans offer something new that really works - and has the potential to pull together a coalition that can win and sustain it.