Can Changing Primary Elections Boost Participation?

Most Americans don’t vote in primary elections—and when they do, those voters tend to be older, wealthier, and more ideologically extreme than the broader electorate. 

That matters because in today’s polarized and heavily gerrymandered system, primary elections often decide who ultimately serves in Congress (in most districts). If only a narrow slice of the public participates, representation suffers.

In a new study, my coauthors (Todd Donovan and Nathan Micatka) and I examine whether changing how primaries work can bring more—and more diverse—voters into the process. Specifically, we ask: Do open, all-candidate primaries increase turnout and reduce demographic gaps in participation?

Why Primary Rules Matter

Turnout in congressional primaries regularly hovers around 20 percent nationwide. Even more concerning, primary voters are not representative of the broader electorate: they are disproportionately older, more affluent, and are strong partisans rather than moderates. 

Because most general elections are uncompetitive, primaries frequently determine who holds office. This makes low and unequal participation in primaries a major—yet often overlooked—source of bias in the US election system.

If only a small, unrepresentative group votes in primaries, the candidates—and policies—that emerge may reflect their preferences, not the public’s.

A New Way to Study Who Votes—and Who Doesn’t

Previous research on primaries has relied heavily on surveys or aggregate turnout statistics, both of which have limitations. Using administrative voter file data covering nearly the entire U.S. adult population, we compare turnout across states with closed primaries, open primaries, and all-candidate primaries and for subsamples of the population based on age, education and income groups.

We analyze turnout in the 2018 and 2022 midterm primaries, comparing states with:

• Closed primaries (only registered party members can vote),
• Open primaries (independents can participate), and
• All-candidate primaries, where all candidates appear on a single ballot and all voters can participate.

This approach allows us to assess turnout not just among registered voters, but across the full voting-age population—and to examine differences by age, income, and education.

With millions of cases in the national voter files, this approach allows us to see not just who votes, but who doesn’t. Survey data isn’t available nationwide for low-turnout primary elections, and if it were, research has found people often overreport turnout. Using administrative records from the 50-state voter files overcomes these problems.

What We Found

The results are striking. All-candidate primaries are consistently associated with higher turnout across age, income, and education groups. Open primaries, by contrast, show little effect on participation.

The biggest gains appear among groups that are typically underrepresented in primaries:

• Young voters (ages 18–34) were more than twice as likely to vote in all-candidate primaries compared to closed primaries.
• Lower-income voters were significantly more likely to participate under all-candidate rules.
• Turnout increased across all education levels, including among those least likely to vote in primaries overall.

Importantly, these patterns hold even after accounting for prior voting history, electoral competitiveness, and state voting laws—suggesting that the design of the primary itself matters.

Younger voters show the biggest gains

Young people are the least likely group to vote in primaries nationwide. But in states with all-candidate primaries, turnout among voters aged 18–34 was more than twice as high as in states with closed primaries.

Income and Education Gaps Narrow - But Don’t Disappear

All-candidate primaries were also linked to higher turnout among lower-income voters, as well as among middle- and higher-income groups. Participation increased across all education levels, suggesting that all-candidate systems raise overall engagement rather than simply shifting who votes. Still, because higher-income and more educated voters remain more likely to participate overall, all-candidate primaries reduce—but do not eliminate—inequality in turnout.

Why It Matters

Primary elections play an outsized role in shaping who governs, yet they routinely exclude large segments of the public. Our findings suggest that all-candidate primaries can meaningfully reduce some of these participation gaps, especially by bringing younger voters into the process.

While all-candidate primaries are not a cure-all for political inequality, they appear to lower barriers to participation in one of the most consequential—and least representative—stages of U.S. elections. As more states debate primary reform, these results offer new evidence that institutional design can shape who shows up—and whose voices are heard.

##

Caroline Tolbert is Professor in the Department of Political Science and Distinguished Chair at the University of Iowa. Her research agenda focuses on campaigns, elections and voting behavior, especially the effects of changes to election policy. She is the coauthor of Accessible Elections: How the States Can Help Americans Vote (2020, Oxford University Press) with Michael Ritter and the Problem with Primary Voters (2026, University of Pennsylvania Press) with Rob Boatright and Nathan Micatka. Her work focuses on measurement and methodological improvements.