Recent research on California’s experience with the top-two primary.

To: Special Joint Committee on Initiative Petitions 

Fr: J. Andrew Sinclair, PhD, Claremont McKenna College 

Re: Ballot Initiative No. 25-12, All-Party Primaries 

Dt: 31 March, 2026 

Dear committee members, 

This brief memo provides a summary of my recent research on California’s experience with  the top-two primary. I would be happy to follow up with any of you, or your staff, to provide  additional information.  

I have studied the top-two since it was first used in California in 2012, including writing a  coauthored book (Alvarez and Sinclair 2015) and academic articles. I am an assistant  professor at Claremont McKenna College; I previously taught at NYU’s Wagner School and  finished my PhD at Caltech in 2013. My current research involves the history of electoral and  administrative reforms in New York and California, so this continues to be an area of active  interest for me. 

My most recent paper (J. A. Sinclair et al. 2026) compared California’s first five elections with  the top-two (2012-2020) to its last five using partisan primaries (2002-2010). The results  would be very similar if we had included the data from 2022 and 2024 as well. California’s  experiment with the top-two primary can certainly help you consider what it would mean  for Massachusetts to adopt a similar rule.  

California’s Top-Two 

In the early 2000s, California appeared ungovernable. In 2003, Californians recalled their  governor, but the dysfunction continued. Approval of the state legislature sunk to around  10%, with Republicans and Democrats failing to reach the necessary common ground even  to pass a budget. In 2009, California’s embarrassed government had to resort to printing  “I.O.U.’s” to pay bills (Steinhauer 2009). In 2010, the state budget was 100 days late  (Nagourney 2010). Californians responded by approving several reforms around that time,  including: the citizens redistricting commission (2008, 2010), a simple majority budget  (2010), and the top-two primary (2010).  

The top-two is a simple nonpartisan election procedure, used for the state legislature,  Congress, and most statewide offices. Any voter can choose any candidate in the primary  election. The two candidates with the most votes advance to the general election. Most of  the time, those two candidates are from different parties, but they can be from the same  party. The candidate with the most general election votes wins.  

There are other kinds of nonpartisan election procedures. The California top-two for these  offices includes candidate party preference on the ballot: it is nonpartisan in the structure of  candidate competition, but not in the information available. 

Which is very similar to the rule in Washington (state) too. 

The top-four (with ranked-choice voting) procedure used in Alaska: both are nonpartisan  systems, in the sense that every voter can choose among all of the candidates and the final  choice may involve candidates of the same party. 

Evaluating the unique impact of the top-two primary on California has been difficult, in part  because California adopted so many reforms at once. As I argued a few years ago, if the main  problem the top-two was supposed to solve was the inability to pass a budget, removing the  supermajority requirement accomplished by other means much of what the top-two was  supposed to do as well (J. A. Sinclair 2017). It is remarkable, though, that by the end of 2025  approval of the legislature was over 50% (Baldassare et al. 2025), given where it had been a  decade and a half earlier. 

Much of the scholarly literature has focused on whether the top-two has helped elect  moderate candidates (compare: McGhee and Shor 2017; Grose 2020). This reflects both one  common way of understanding California’s dysfunction in 2010 (that is: polarization was the  problem) and the intellectual history of research on types of primary elections (Gerber and  Morton 1998). This debate is interesting and important while also nuanced and complicated.  

My paper contributes to this scholarly discussion by reframing some of the debate. Several  aspects of the top-two reform were not receiving adequate attention in the literature. While  some advocates for these rules emphasize moderation, others are more concerned about the  rights of Independent voters, participation more broadly, and the existence of competitive  elections in which voters get to make meaningful choices.  

Increasing Competition and Participation 

In my recent paper (J. A. Sinclair et al. 2026), my coauthors and I focus on two issues:  competition and participation. California is a Democratic-leaning state. Many areas of Los  Angeles and San Francisco vote so overwhelmingly for Democrats that it would be difficult  to draw competitive legislative districts. Under the old system of partisan primaries, all one  had to do in those parts of the state was to win the low-turnout Democratic primary in order  to win the election.  

Same-party general elections add competitiveness in areas where it otherwise  would not exist in November.  

A leading study of direct primaries—of all kinds—observes that they “are contributing to the  US electoral system especially in uncompetitive areas, where they are most needed” (Hirano  and Snyder 2019, 307). Without primaries of any kind, if parties nominated candidates in  conventions and so on, in these districts with uncompetitive general elections there would  be no effective democracy at all. What the top-two does, then, is sometimes move this  competition from the primary election to the general election.  

This is different than using ranked choice voting in a partisan primary, like the recent high-profile  Democratic primary for NYC mayor. 

The paper uses California’s 50th Assembly District (2012-2020) as an example. President  Obama won about three-quarters of the vote there. Yet, in 2012, the primary vote split evenly  between three Democrats and one Republican candidate, with two Democrats advancing to  the general election. The same-party general election was very competitive: fewer than  2,000 votes separated the winner from the loser in a district where cross-party elections  were foregone conclusions.  

Competition matters because the candidates have to try hard to win the support of every  possible voter. Competition in these safe districts between two candidates of the same party  also means that voters have to look beyond just the party label—and some voters do search  for information to help guide their choice (B. Sinclair and Wray 2015). 

(2). Nonpartisan elections mean more people matter.  

Partisan primaries exclude—in practice, if not by law—many voters. Nonpartisan rules bring  them back in. 

In California before the top-two, for example, many Independent voters did not participate  in the partisan primaries, even if formally they could have done so. Parties also had little  incentive to invest in turning out such voters. In 2010 in CA, the last partisan primary: 

Total registered Independents: — — — — 3,423,750 Total voting Independents: — — — — — 805,977 23.5% Total Independents voting on a Democratic ballot: — 164,198 4.8% 

That is, only 4.8% of the registered Independents voted on the primary ballot of the state’s  majority party in a year Democrats would win every statewide office. A similar number voted  on the Republican ballot, and the largest group did not select a party ballot at all.  

With the top-two, every Independent voter’s ballot by default includes the state legislative,  congressional, and statewide executive offices. Just for the sake of argument, look at the 2022  numbers. There were now 4,983,013 Independent voters. Of that group, 1,122,032  participated in the primary. Assuredly, some voted for Republican candidates, and some  skipped the race, but every single voter had the choice in front of them to vote for whichever  candidates they preferred in the primary for every office.  

In the paper, we examine safe Democratic legislative seats and look at the winner’s vote share, out of the total number of registered voters, when facing the last Democratic opponent.  Under the old rules, this was always in the primary. With the top-two, it was still often in the  primary: many districts went on to have general elections featuring a Republican and in both years, a midterm election with a Democrat in the White House.  

The 2010 numbers come from the official Statement of Vote. Since the top-two makes many of  the old categories irrelevant, this no longer appears in the 2022 Statement of Vote, but  can be found here: https://elections.cdn.sos.ca.gov/vca/2022-vca-report/vca-2022- primary-election.pdf.

Democrat, even in safe Democratic seats. In some contests, though, the final choice between  Democrats took place in the general election. 

We find that participation was marginally higher in the last meaningful choice under the top two rules, even when the election effectively ended in the primary. This is likely the result of  the automatic inclusion of Independent voters. When two Democrats advanced to the  general election ballot, of course, overall participation in the final choice was much greater than it would have been in a Democratic party primary. 

The combination of these effects—the potential for meaningful general election competition,  in places that otherwise would never have it, and more voter participation—means that the  winners have to obtain support from more of the people they intend to govern. It is certainly  true that with two Democrats on the general election ballot sometimes Republicans will skip  the election (Fisk 2020), but no Republicans are participating in closed or semi-closed  partisan primaries. In a close contest between two Democrats, the votes of Independents  and Republicans can really matter.  

Choosing an Election System 

Political scientists have known an awkward truth for decades: there is no such thing as a  theoretically perfect election system.5 Choices between election systems involve political  decisions about the costs and benefits, in practical terms, about what is likely to happen most  of the time. The impact of a primary law will also depend on its political context and the  state’s other institutions.  

The Role of Parties  

In most states, most of the time, with most electoral institutions: the majority party candidate  favored by party leaders, donors, and volunteers is likely to win the election (Hassell 2018).  Reform advocates sometimes find this disappointing, and have since the advent of the direct  primary (Boatright 2024). This is not necessarily a bad thing, though. “It must be rather  obvious that a political boss can survive only so long as he wins elections,” wrote New York’s  Edward J. Flynn, “and equally obvious the only way to win elections year after year is to know  what the voters want and give it to them” (Flynn 1947, 221). 

The top-two is a potential outlet for voters if parties and candidates fail to take Flynn’s advice.  If a party really drifts too far away from policies acceptable to most voters, or leaves voters  feeling like candidates are unresponsive to their concerns, the top-two makes sure that no  district is truly safe. It is less what same party elections have done, and more what they might  do, that provides the incentive for performance. The significance of the participation and  

A conclusion reached by scholars studying theories of social choice. As William Riker succinctly  puts it: “it is […] the case that no method of voting can simultaneously satisfy several  elementary conditions of fairness and also produce results that always satisfy elementary  conditions of logical arrangement. Hence, not only may the results of voting fail to be  fair, they may also fail to make sense” (Riker 1982, 115). 

competition results, described above, is that voters are better positioned to express  displeasure, should they really feel it.  

Overall, though, we are not looking at a moment in which California Democratic voters are  anywhere near as frustrated with their own party as they were when the top-two passed in  2010. At the moment, California Democrats tend to disapprove of Republicans or  Republican-controlled organizations; in contrast, they express some approval of the  Democratic Party and its officeholders (Baldassare et al. 2025).  

Turning satisfaction into votes requires leadership and information. If anything, the  California Democratic Party should have taken steps to make its pre-primary endorsement  procedures more decisive: no candidate won the 2026 gubernatorial endorsement, for  example, because it required a threshold of 60% at the pre-primary nominating convention  (Kuang 2026). This rule, not formally a part of the top-two, means that the party is the least  likely to issue an endorsement in the precise circumstances in which endorsements would  most help voters.  

Other Institutions 

The core of the primary law—who gets to vote for whom, and how do those votes turn into  outcomes—is also only one part of a state’s institutions. Other rules such as voter  registration procedures, convenience voting alternatives, ballot access for candidates, and so  on can all influence how the election laws actually work. Furthermore, it also matters what  winners can do in office once they get there. If Massachusetts does change its primary  election law, it would make sense to invest in figuring out how to make it successful, too.  

Further Information 

For additional information, please contact: asinclair@cmc.edu. I would be happy to speak to  legislators supporting, opposing, or uncertain about the proposed change.  

References 

Alvarez, R. Michael, and J. Andrew Sinclair. 2015. Nonpartisan Primary Election Reform: Mitigating  Mischief. Cambridge University Press. 

Baldassare, Mark, Dean Bonner, Lauren Mora, and Deja Thomas. 2025. “PPIC Statewide Survey:  Californians and Their Government.” Public Policy Institute of California, December.  https://www.ppic.org/publication/ppic-statewide-survey-californians-and-their government-december-2025/. 

Boatright, Robert G. 2024. Reform and Retrenchment: A Century of Efforts to Fix Primary Elections.  Oxford University Press. 

Fisk, Colin A. 2020. “No Republican, No Vote: Undervoting and Consequences of the Top-Two Primary  System.” State Politics & Policy Quarterly 20 (3): 292–312.  https://doi.org/10.1177/1532440019893688.

Flynn, Edward J. 1947. You’re the Boss. Viking Press. https://tinyurl.com/ys6ray6f. 

Gerber, Elisabeth R., and Rebecca B. Morton. 1998. “Primary Election Systems and Representation.”  Journal of Law, Economics, & Organization 14 (2): 304–24. 

Grose, Christian R. 2020. “Reducing Legislative Polarization: Top-Two and Open Primaries Are  Associated with More Moderate Legislators.” Journal of Political Institutions and Political  Economy 1 (2): 267–87. https://doi.org/10.1561/113.00000012. 

Hassell, Hans J. G. 2018. The Party’s Primary: Control of Congressional Nominations. Cambridge  University Press. 

Hirano, Shigeo, and James M. Snyder Jr. 2019. Primary Elections in the United States. Political Economy  of Institutions and Decisions. Cambridge University Press.  https://doi.org/10.1017/9781139946537. 

Kuang, Jeanne. 2026. “Swalwell Leads Delegate Vote but California Democrats Divided in Crowded  Primary.” Politics. CalMatters, February 22.  https://calmatters.org/politics/2026/02/democratic-convention-crowded-governors race/. 

McGhee, Eric, and Boris Shor. 2017. “Has the Top Two Primary Elected More Moderates?” Perspectives  on Politics 15 (4): 1053–66. https://doi.org/10.1017/S1537592717002158. 

Nagourney, Adam. 2010. “California Lawmakers Pass Overdue Budget.” U.S. The New York Times,  October 7. https://www.nytimes.com/2010/10/08/us/08california.html. 

Riker, William H. 1982. Liberalism against Populism: A Confrontation between the Theory of Democracy  and the Theory of Social Choice. Waveland Press. 

Sinclair, Betsy, and Michael Wray. 2015. “Googling the Top Two: Information Search in California’s Top  Two Primary.” California Journal of Politics and Policy 7 (1).  https://doi.org/10.5070/P2cjpp7125443. 

Sinclair, J. Andrew. 2017. “Democratic Accountability and Institutional Reform: Lessons from  California’s 2010 Angry Electorate.” California Journal of Politics and Policy 9 (2).  https://doi.org/10.5070/P2cjpp9234856. 

Sinclair, J. Andrew, Ian O’Grady, Bryn Miller, and Catherine M. Murphy. 2026. “Participation and  Competition in Top-Two Elections: Trade-Offs in Election Reform.” State Politics & Policy  Quarterly 26 (1): 103–15. https://doi.org/10.1017/spq.2025.10007. 

Steinhauer, Jennifer. 2009. “Coffers Empty, California Pays With I.O.U.’s.” U.S. The New York Times, July  2. https://www.nytimes.com/2009/07/03/us/03calif.html. 

Ware, Alan. 2002. The American Direct Primary: Party Institutionalization and Transformation in the  North. Cambridge University Press.