The public purpose of primary elections — for which all taxpayers foot the bill — is to winnow a potentially wide candidate pool into a fewer number of candidates in order to clarify the choices voters will have in the general election.
That primaries have traditionally served as a method for both major political parties to separately “nominate” a single candidate does not make this more narrow purpose a requirement of any primary process, as the Supreme Court has found.
Neither the Constitution nor Congress prescribes the method of primary elections. This flexibility gives every state the freedom to decide how best to design their electoral system.
As this paper has illuminated, there are several problems with primaries that are designed to serve both major parties, insofar as they disenfranchise voters, distort representation, and fuel political division.
An alternative to partisan primaries are nonpartisan primaries that instead are designed to serve the voters: they can give every citizen an equal voice, produce more representative outcomes, and improve governing incentives by ensuring our elected leaders are accountable to a broader swath of the electorate.
In nonpartisan primaries, voters participate in a single primary election, with all candidates listed on a single ballot regardless of their party affiliation; the top finishers advance to the general election, and whoever earns a majority of votes wins.
States with nonpartisan primaries experience higher voter participation, and evidence suggests that after adoption of nonpartisan primaries states had increased electoral competitiveness, decreased legislative polarization, and elected more moderate members of Congress.
There are currently two variations of nonpartisan primaries in practice, which differ based on whether only two or more than two candidates advance to the general election.
Soon, Alaska will implement a “top-four” nonpartisan primary system, while California, Nebraska, and Washington currently use “top-two” nonpartisan primaries.
In November 2020, Alaskan voters approved a ballot measure to establish top-four nonpartisan primaries for state and federal offices. The ballot measure also established ranked choice voting in general elections, which will guarantee majority winners by giving voters the option to rank candidates in order of preference. If no candidate receives 50% support from the outset, the last place candidate is eliminated, and votes are transferred from that candidate to the voter’s next choice. The process continues until someone receives a majority of the vote.
The Alaska model was based on a proposal made by Katherine Gehl and Michael Porter in their seminal 2017 Harvard Business School report, “Why Competition in the Politics Industry is Failing America.” Gehl, founder of the Institute for Political Innovation, advocates for a top-five nonpartisan primary and ranked choice voting general elections — combined as “Final-Five Voting.”
Gehl argues that “Final-Five Voting” systems will promote the healthiest state of competition and improve governing incentives the most. With more candidates on the general election ballot, it opens the pathway for more competition and gives voters more choice. As Gehl describes:
"The ultimate purpose of Final-Five Voting is not necessarily to change who wins. It is to change what the winners are incentivized to do. Under this system, the message to Congress is do your job or lose your job. Innovate. Reach across the aisle whenever it is helpful. And come up with real solutions to our problems and create new opportunities for progress. Or be guaranteed new and healthy competition in the next election."
Top four or five nonpartisan primaries will make it very likely that general election candidates will need to compete against opponents from their party. In theory, this will require candidates to focus more on policy issues and ideas, rather than just their party affiliation, advantaging candidates with appeal to the broadest possible electorate.
Nonpartisan primaries with ranked choice voting general elections also also help level the playing field for independents and third party candidates, making it easier for them to gain attention during the primary and removing the “spoiler” effect that occurs when votes are split.
The public purpose of primary elections — for which all taxpayers foot the bill — is to winnow a potentially wide candidate pool into a fewer number of candidates in order to clarify the choices voters will have in the general election.
That primaries have traditionally served as a method for both major political parties to separately “nominate” a single candidate does not make this more narrow purpose a requirement of any primary process, as the Supreme Court has found.
Neither the Constitution nor Congress prescribes the method of primary elections. This flexibility gives every state the freedom to decide how best to design their electoral system.
As this paper has illuminated, there are several problems with primaries that are designed to serve both major parties, insofar as they disenfranchise voters, distort representation, and fuel political division.
An alternative to partisan primaries are nonpartisan primaries that instead are designed to serve the voters: they can give every citizen an equal voice, produce more representative outcomes, and improve governing incentives by ensuring our elected leaders are accountable to a broader swath of the electorate.
In nonpartisan primaries, voters participate in a single primary election, with all candidates listed on a single ballot regardless of their party affiliation; the top finishers advance to the general election, and whoever earns a majority of votes wins.
States with nonpartisan primaries experience higher voter participation, and evidence suggests that after adoption of nonpartisan primaries states had increased electoral competitiveness, decreased legislative polarization, and elected more moderate members of Congress.
There are currently two variations of nonpartisan primaries in practice, which differ based on whether only two or more than two candidates advance to the general election.
Soon, Alaska will implement a “top-four” nonpartisan primary system, while California, Nebraska, and Washington currently use “top-two” nonpartisan primaries.
In November 2020, Alaskan voters approved a ballot measure to establish top-four nonpartisan primaries for state and federal offices. The ballot measure also established ranked choice voting in general elections, which will guarantee majority winners by giving voters the option to rank candidates in order of preference. If no candidate receives 50% support from the outset, the last place candidate is eliminated, and votes are transferred from that candidate to the voter’s next choice. The process continues until someone receives a majority of the vote.
The Alaska model was based on a proposal made by Katherine Gehl and Michael Porter in their seminal 2017 Harvard Business School report, “Why Competition in the Politics Industry is Failing America.” Gehl, founder of the Institute for Political Innovation, advocates for a top-five nonpartisan primary and ranked choice voting general elections — combined as “Final-Five Voting.”
Gehl argues that “Final-Five Voting” systems will promote the healthiest state of competition and improve governing incentives the most. With more candidates on the general election ballot, it opens the pathway for more competition and gives voters more choice. As Gehl describes:
"The ultimate purpose of Final-Five Voting is not necessarily to change who wins. It is to change what the winners are incentivized to do. Under this system, the message to Congress is do your job or lose your job. Innovate. Reach across the aisle whenever it is helpful. And come up with real solutions to our problems and create new opportunities for progress. Or be guaranteed new and healthy competition in the next election."
Top four or five nonpartisan primaries will make it very likely that general election candidates will need to compete against opponents from their party. In theory, this will require candidates to focus more on policy issues and ideas, rather than just their party affiliation, advantaging candidates with appeal to the broadest possible electorate.
Nonpartisan primaries with ranked choice voting general elections also also help level the playing field for independents and third party candidates, making it easier for them to gain attention during the primary and removing the “spoiler” effect that occurs when votes are split.
In 2020, “Alaskans for Better Elections launched a Final-Four Voting reform campaign in direct response to the Primary Problem. First, lawmakers had been previously “primaried” for working across the aisle. In 2020, five incumbent Republican legislators lost their seats for being too willing to compromise: in one of the Republican races, only 7.5% of the district’s eligible voters decided the winner. Second, the majority (63%) of the state’s voters are unaffiliated with either major party, but had historically been prohibited from participating in partisan primaries. Third, there is a history of independent candidates winning in Alaska, but also of independent candidates coming up short because of the “spoiler effect.”
As former state representative Chuck Kopp (R) — one of the primaried incumbents — remarked: “In primaries, the way the system is set up, you only have to message to and appeal to the extreme edge of the parties, left and right, and activate those to vote.”
Nonpartisan primaries in Alaska have the potential to allow leaders, including members of the bipartisan governing coalition in the state house, to act in the interest of what’s best for the state without the fear of being “primaried.” Commentators are already documenting how the reform, which will take effect in 2022 for all elections, is changing governing incentives, enfranchising voters, and creating new pathways for electoral competition.
In 2020, “Alaskans for Better Elections launched a Final-Four Voting reform campaign in direct response to the Primary Problem. First, lawmakers had been previously “primaried” for working across the aisle. In 2020, five incumbent Republican legislators lost their seats for being too willing to compromise: in one of the Republican races, only 7.5% of the district’s eligible voters decided the winner. Second, the majority (63%) of the state’s voters are unaffiliated with either major party, but had historically been prohibited from participating in partisan primaries. Third, there is a history of independent candidates winning in Alaska, but also of independent candidates coming up short because of the “spoiler effect.”
As former state representative Chuck Kopp (R) — one of the primaried incumbents — remarked: “In primaries, the way the system is set up, you only have to message to and appeal to the extreme edge of the parties, left and right, and activate those to vote.”
Nonpartisan primaries in Alaska have the potential to allow leaders, including members of the bipartisan governing coalition in the state house, to act in the interest of what’s best for the state without the fear of being “primaried.” Commentators are already documenting how the reform, which will take effect in 2022 for all elections, is changing governing incentives, enfranchising voters, and creating new pathways for electoral competition.
Three states — California, Nebraska, and Washington — use top-two nonpartisan primaries. The congressional delegations from California and Washington, 14% of Congress, are elected in nonpartisan primaries. In Nebraska, the system is only used for the state legislature. Voter participation is higher in states with nonpartisan primaries and some research demonstrates the system has improved electoral competition while decreasing legislative polarization.
Data collected for this report found, on average, 2020 voter participation in California (34.4%) and Washington’s (44.8%) nonpartisan congressional primaries to be much higher than an estimated average 18.3% turnout rate in partisan primaries. A Bipartisan Policy Center report also found that total ballots cast in Washington and California in the three most recent midterm elections exceeded national averages (see Figure 11). In his 2020 study, which controlled for other factors, scholar Seth Hill found nonpartisan primaries responsible for a 6% voter turnout rate in primary elections.
Voter participation may be higher, in part, because nonpartisan primaries promote healthier competition between candidates. According to Ballotpedia, the two states with nonpartisan primaries for congressional elections far outpace the national average, both in the share of contested primaries and the share of primaries in which incumbents face a challenger: they also had more primary candidates running per district (see Figure 12).
Additionally, more than 40% of districts (151) nationwide had no competition in the dominant party’s primary, while just seven of the 63 Congressional districts (11%) with nonpartisan primaries had only two candidates. In more than half of the nonpartisan primaries, voters had a choice between at least five candidates.
Further, the system improves competition by increasing accountability for incumbents. In a report published by the Schwarzenegger Institute, Charles Munger Jr., an advocate of the top-two system, found that electoral accountability for California’s state legislative incumbents improved after implementation of nonpartisan primaries noting that “under pre-reform partisan primaries, an average of 0.6 incumbents per year lost in primaries, while under the top-two, an average of 3.3 incumbents per year lost in same-party general elections, a factor of 5.6 more.” Other research also found increased incumbent accountability. Following the 2012 redistricting cycle and after the implementation of top-two primaries, 27.6% of incumbents were defeated or decided not to run again, up 10% from 2002 following the 2002 cycle.
Evidence also shows nonpartisan primaries can help promote moderation in Congress. In a 2020 study, University of Southern California political scientist Christian Grose used ideology scores (DW-Nominate) based on voting record to compare U.S. Representatives who were elected from different primary systems. Over a period from 2003-2018, Grose found that among newly elected members, “those elected in top-two primaries are more than 18 percentage points less extreme than closed primary legislators."
Grose theorizes that the threat and/or reality of two candidates from the same party appearing on the general election ballot has a moderating influence on candidates. He argues this feature of the system requires candidates not only rely on partisan heuristics, but instead to appeal to a broader cross section of the electorate with more nuanced, issue-based views. The aforementioned report by Charles Munger Jr. found that 20 of 58 (34%) competitive same-party general election contests for state legislature and U.S. House in California between 2012-2016 resulted in a winner who would have otherwise lost their partisan party.
In 2020, eight districts in California and one district in Washington advanced top-two winners from the same party. Instead of being primaried by in the primary, the top-two primary structure allowed moderate candidates with cross-over appeal to compete in both the primary and general elections. For example, in Washington’s 10th District, Marilyn Strickland (D) and Beth Doglio (D) both advanced out of the state’s top-two primary. Strickland won in the general election after, notably, campaigning as a moderate candidate willing to work across the aisle.
An earlier study by Grose found that after the implementation of California’s nonpartisan primary system and their independent redistricting commission, the median member of the state legislature became more moderate over the course of the first two years of implementation; polarization between the two parties had decreased by 15% in the state assembly and by 10% in the state senate on the ideological polarization scale.
Oklahoma University professor Steven Sparks found candidates running in nonpartisan primaries were less likely to mention partisan affiliation or use ideological statements on their websites, but more likely to issue bipartisan statements. Jane Junn and Sara Sadhwani, also of USC, attributes the nonpartisan primary system alongside other reforms (term limits and redistricting) to a California governing system that “enhances the prospect for normatively ‘good’ representation” and does less to reward entrenched partisan interests.
There are, however, limitations of the top-two system. First, other political scientists who have studied top-two primaries have found less or no impact on polarization. Second, the top-two primary could remain the election of consequence in “safe” districts to determine who is elected, limiting voter choice in general elections; competitive primary election fields get narrowed to just two candidates by a relatively small portion of the electorate (top-two nonpartisan primaries traditionally have about half the voter participation as in general elections). Third, it is possible that neither of the top two finishers in a particular district are from the dominant party in that district, if many run and the vote is split. Fourth, because vote splitting can still occur in the primary and only two candidates to the general election, large barriers remain for third party and independent candidates; in fact, only one such candidate (an incumbent legislator — Chad Mayes — who left the Republican Party) has ever been elected under the system.
Advancing additional candidates from the primary and using ranked choice voting in the general election may help solve the latter two challenges, and advocates of nonpartisan primaries suggest more time is needed for these reforms to demonstrate their full impact — including to shift who runs and who wins as incumbents retire.
Three states — California, Nebraska, and Washington — use top-two nonpartisan primaries. The congressional delegations from California and Washington, 14% of Congress, are elected in nonpartisan primaries. In Nebraska, the system is only used for the state legislature. Voter participation is higher in states with nonpartisan primaries and some research demonstrates the system has improved electoral competition while decreasing legislative polarization.
Data collected for this report found, on average, 2020 voter participation in California (34.4%) and Washington’s (44.8%) nonpartisan congressional primaries to be much higher than an estimated average 18.3% turnout rate in partisan primaries. A Bipartisan Policy Center report also found that total ballots cast in Washington and California in the three most recent midterm elections exceeded national averages (see Figure 11). In his 2020 study, which controlled for other factors, scholar Seth Hill found nonpartisan primaries responsible for a 6% voter turnout rate in primary elections.
Voter participation may be higher, in part, because nonpartisan primaries promote healthier competition between candidates. According to Ballotpedia, the two states with nonpartisan primaries for congressional elections far outpace the national average, both in the share of contested primaries and the share of primaries in which incumbents face a challenger: they also had more primary candidates running per district (see Figure 12).
Additionally, more than 40% of districts (151) nationwide had no competition in the dominant party’s primary, while just seven of the 63 Congressional districts (11%) with nonpartisan primaries had only two candidates. In more than half of the nonpartisan primaries, voters had a choice between at least five candidates.
Further, the system improves competition by increasing accountability for incumbents. In a report published by the Schwarzenegger Institute, Charles Munger Jr., an advocate of the top-two system, found that electoral accountability for California’s state legislative incumbents improved after implementation of nonpartisan primaries noting that “under pre-reform partisan primaries, an average of 0.6 incumbents per year lost in primaries, while under the top-two, an average of 3.3 incumbents per year lost in same-party general elections, a factor of 5.6 more.” Other research also found increased incumbent accountability. Following the 2012 redistricting cycle and after the implementation of top-two primaries, 27.6% of incumbents were defeated or decided not to run again, up 10% from 2002 following the 2002 cycle.
Evidence also shows nonpartisan primaries can help promote moderation in Congress. In a 2020 study, University of Southern California political scientist Christian Grose used ideology scores (DW-Nominate) based on voting record to compare U.S. Representatives who were elected from different primary systems. Over a period from 2003-2018, Grose found that among newly elected members, “those elected in top-two primaries are more than 18 percentage points less extreme than closed primary legislators."
Grose theorizes that the threat and/or reality of two candidates from the same party appearing on the general election ballot has a moderating influence on candidates. He argues this feature of the system requires candidates not only rely on partisan heuristics, but instead to appeal to a broader cross section of the electorate with more nuanced, issue-based views. The aforementioned report by Charles Munger Jr. found that 20 of 58 (34%) competitive same-party general election contests for state legislature and U.S. House in California between 2012-2016 resulted in a winner who would have otherwise lost their partisan party.
In 2020, eight districts in California and one district in Washington advanced top-two winners from the same party. Instead of being primaried by in the primary, the top-two primary structure allowed moderate candidates with cross-over appeal to compete in both the primary and general elections. For example, in Washington’s 10th District, Marilyn Strickland (D) and Beth Doglio (D) both advanced out of the state’s top-two primary. Strickland won in the general election after, notably, campaigning as a moderate candidate willing to work across the aisle.
An earlier study by Grose found that after the implementation of California’s nonpartisan primary system and their independent redistricting commission, the median member of the state legislature became more moderate over the course of the first two years of implementation; polarization between the two parties had decreased by 15% in the state assembly and by 10% in the state senate on the ideological polarization scale.
Oklahoma University professor Steven Sparks found candidates running in nonpartisan primaries were less likely to mention partisan affiliation or use ideological statements on their websites, but more likely to issue bipartisan statements. Jane Junn and Sara Sadhwani, also of USC, attributes the nonpartisan primary system alongside other reforms (term limits and redistricting) to a California governing system that “enhances the prospect for normatively ‘good’ representation” and does less to reward entrenched partisan interests.
There are, however, limitations of the top-two system. First, other political scientists who have studied top-two primaries have found less or no impact on polarization. Second, the top-two primary could remain the election of consequence in “safe” districts to determine who is elected, limiting voter choice in general elections; competitive primary election fields get narrowed to just two candidates by a relatively small portion of the electorate (top-two nonpartisan primaries traditionally have about half the voter participation as in general elections). Third, it is possible that neither of the top two finishers in a particular district are from the dominant party in that district, if many run and the vote is split. Fourth, because vote splitting can still occur in the primary and only two candidates to the general election, large barriers remain for third party and independent candidates; in fact, only one such candidate (an incumbent legislator — Chad Mayes — who left the Republican Party) has ever been elected under the system.
Advancing additional candidates from the primary and using ranked choice voting in the general election may help solve the latter two challenges, and advocates of nonpartisan primaries suggest more time is needed for these reforms to demonstrate their full impact — including to shift who runs and who wins as incumbents retire.
i. Parties can still endorse a preferred candidate, and candidate names can still include party affiliation, just as done today.
ii. Nebraska conducts a top-two nonpartisan primary for the state’s nonpartisan, unicameral legislature. The system does not apply to primary elections for Congress. Louisiana is unique in that it does not have a traditional party primary election, but is often referred to as holding a nonpartisan election. All candidates run on the same ballot in the general election, and if no candidate receives a majority, the top two finishers participate in a runoff six weeks later.
iii. Without ranked choice voting, the winning candidate could potentially receive a plurality of the vote share. For more information about the merits of ranked choice voting, see Appendix A.
33 Brooks, James. “Ballot Measure 2 would change the way Alaskans vote for statewide candidates and those running for the Legislature. Here's How.” Anchorage Daily News (October 2020).
34 Ibid.
35 Berman, Russell. “The Political Reform Movement Scores Its Biggest Win Yet,” Atlantic (January 2021). Also see: Drutman, Lee. “There’s a reason why Lisa Murkowski can threaten to leave the Republicans,” Washington Post (January 2021).
36 “2018 Primary Election Turnout and Reforms,” Bipartisan Policy Center (November 2018).
37 “2020 Primary Election Competitiveness in State and Federal Government,” Ballotpedia (February 2021).
38 Grose, Christian. “The Adoption of Electoral Reforms and Ideological Change in the California State Legislature,” USC Schwarzenegger Institute (2014).
39 Gutman, David. “Marilyn Strickland wins Washington’s 10th Congressional District; Kim Schrier still leads in 8th,” Seattle Times (November 2020).
40 Grose, Christian. “The Adoption of Electoral Reforms and Ideological Change in the California State Legislature,” USC Schwarzenegger Institute
41 Sparks, Steven. “Polarization and the Top-Two Pirmary: Moderating Candidate Rhetoric in One-Party Contests,” Political Communication 36, no. 4 (March 27, 2019): 565-585.
42 Sadhwani, Sara and Jane Junn. “Structuring Good Representation: Institutional Design and Elections in California,” American Political Science Association (April 2018): 318-322.
43 Santucci, Jack. “Nonpartisan Elections Don’t Reduce Polarization,” Medium (blog) (February 2021).
