Andrew Hall is the Davies Family Professor of Political Economy at the Graduate School of Business and Professor of Political Science at Stanford University. Hall combines large-scale quantitative datasets with tools from economics, statistics, and machine learning to understand how to design democratic systems of governance, with a focus on American elections and legislatures as well as the governance of online communities.
Summary
Promise of the Reform
"There are many potential administrative benefits to voting by mail, including reducing wait-times and enhancing voter satisfaction. Advocates have long claimed that expanding vote-by-mail options, and especially moving to universal vote-by-mail, increases overall political participation. In particular, it is thought to expand the participation of more marginalized political groups like minorities, younger voters, and voters who face accessibility challenges to voting in person."
Key Takeaways from the Research
Expanding access to absentee voting leads many voters to choose to switch from voting in person to voting by mail, indicating at least some preference among an important set of voters for voting by mail.
However, expanding access to absentee voting does not seem to increase turnout rates at all, at least where we are able to measure its impact.
Universal vote-by-mail programs tend to increase turnout, although the effect is modest—leading to approximately a 2-to-4 percentage-point increase in turnout.
Universal vote-by-mail does not seem to cause large increases in Democratic vote share, contrary to popular claims. However, we cannot rule out the possibility that it could advantage Democrats in very close races.
Important Questions that the Research Does Not Answer
What effect does voting by mail have on people’s sense of civic duty, relative to voting together in person, and what factors make this effect larger or smaller?
What effect does voting by mail have on the fairness and security of the election system, and on people’s perceptions of these?
What effect does voting by mail have on how voters aggregate information, if different voters cast their votes at different times, possibly before and after late-breaking information comes out?
Introduction
“Proponents of expanding vote-by-mail argue that it lowers the costs of voting and thereby increases political participation.”
Vote-by-mail is a catch-all term for a variety of methods for allowing voters to cast their votes by mail. It is used in a variety of ways in most states and has been around for many years but has become newly salient during and after the 2020 election. Proponents of expanding vote-by-mail options argue that it is a more-convenient way for people to vote that lowers the costs of voting and thereby increases political participation.
Voting by mail is an important set of election administration policies intended to make voting more convenient by allowing voters to cast ballots by mail rather than in person.
Absentee voting is a very common form of voting by mail that requires voters to “opt in” to voting by mail.
Universal vote-by-mail is a more intensive policy in which the state mails ballots to all registered voters.
Both kinds of voting by mail have, for the moment, become a matter of political conflict, with former President Trump and some Republicans opposing it while most Democrats are in favor of it.
The purpose of this review is to explain how social scientists think about the effect of vote-by-mail on political participation and to review empirical evidence for its effects on participation, and to discuss other areas of the vote-by-mail debate that are less well covered by research to date.
Theoretical Research on Voting by Mail and Political Participation
“Most arguments for why voting by mail increases participation argue that it reduces the cost of voting by making it logistically easier to vote.”
How should we think about the ways that vote-by-mail might influence people’s decisions to turn out to vote?
“The calculus of voting” is an old framework that helps structure how we think about why people might vote.¹ The calculus states that a voter will choose to vote when:
(Probability of your vote changing the outcome x Value of outcome) + Psychological Rewards to Voting > Cost of Voting
The idea is that your decision to vote depends on: how much you care about the outcome of the vote, what chance your vote has of determining the outcome, the costs of voting, and other social or psychological rewards to voting. In most mass elections, your chance of determining the outcome is essentially zero, so researchers spend a lot of time studying the role of the costs of voting in determining turnout.
Political actors also devote substantial attention to the costs of voting. During Jim Crow, southern Democrats intentionally made the costs of voting extremely high for Black voters as part of their sustained efforts to prevent them from voting. More recently, Democrats and Republicans have repeatedly clashed over different election reforms that alter the costs of voting in complicated ways, with many Republicans arguing that these changes increase security while many Democrats argue that they lower the turnout of marginalized groups and especially Black voters.
How does vote-by-mail fit into this? Most arguments for why voting by mail increases participation argue that it reduces the cost of voting by making it logistically easier to vote. Instead of going to a polling place in person to vote, voters can vote whenever they want from the comfort of their own home. By making voting more convenient, vote-by-mail may increase turnout among voters who perceive the costs of in-person voting as high.
However, it is possible that vote-by-mail does not meaningfully reduce the perceived costs to voting if the main costs to voting are actually cognitive rather than logistical. Voters still need to care enough to study the candidates and issues, and they may not want to regardless of how easy or hard it is to carry out the act of voting after deciding.
Moreover, voting by mail could also influence psychological rewards to voting if, for example, you feel better about yourself when you vote in person, seeing and being seen by your neighbors as you vote (and getting your “I Voted” sticker). For all these reasons and others, the effect of voting by mail on turnout is theoretically unclear, and we need to examine empirical data to understand how much it changes turnout in practice.
Empirical Research on Voting by Mail and Political Participation
“It seems likely that universal vote-by-mail has a modest, positive effect on turnout overall.”
Understanding how vote-by-mail affects participation requires comparing elections where vote-by-mail is used to elections where it is not. However, we might worry that the places that have chosen to implement vote-by-mail policies are different in many ways from the places that haven’t—they may be more liberal places, for example, or they might be places with greater resources for election administration.
To address statistical problems like these and to figure out if vote-by-mail itself changes participation, we would like to run an experiment where some elections randomly occur with a vote-by-mail option while others occur without it. This would ensure that, on average, the places with vote-by-mail are otherwise the same as the places without it. We could then compare turnout rates across the two sets of elections, and the difference in their turnout rates would indicate the effect of vote-by-mail.
We can’t run an experiment like that. But we can use natural experiments—in particular, we can use states where some counties rolled out vote-by-mail earlier than other counties in the same state did. This is nice, because all of these counties voted in the same statewide and national electrons, but did so with different election administration rules, which gives us something closer to the ideal experiment we would like to run.
Universal Vote-By-Mail
The first study to pursue this strategy² compared voters in California, who were quasi-randomly assigned to universal vote-by-mail or regular voting. The study found no increase in turnout—in fact, if anything, it found that the all-mail elections reduced turnout somewhat, on average. However, it did find more positive effects for local elections.
A subsequent piece of research³ pursued a different natural experiment. Studying the state of Washington, the study compared turnout rates in counties that implemented universal vote-by-mail to other counties that had not yet implemented it. The paper found that universal vote-by-mail increases participation by roughly 2.5 percentage points.
Since then, several other states have also implemented universal vote-by-mail through a staggered county-by-county rollout, allowing for the same kind of empirical analysis. One paper used the same method as the Washington study to examine the effects on turnout and partisan vote share in California and Utah. Both this paper and similar studies found a remarkably similar estimate to the Washington paper.⁴, ⁵ Although the early effects in California may have been small or negative, it seems likely that universal vote-by-mail has a modest, positive effect on turnout overall.
“We cannot say that switching on or off universal vote-by-mail is irrelevant to partisan politics today.”
In addition to studying effects on turnout, research has also examined potential partisan effects. Some people have argued that universal vote-by-mail is more likely to encourage younger or more marginalized populations to vote. If these groups are more likely to vote for Democrats, the logic goes, then the policy may help Democrats more than Republicans.
One paper found very modest estimates for the effect on Democratic vote share.6 However, the upper bounds of the 95% confidence interval in some specifications do contain meaningful estimates (as much as a 1.3 percentage-point increase in Democratic vote share in the most precise estimate). While a 1.3 percentage-point increase is nothing like popular claims about how universal vote-by-mail would usher in permanent rule by the Democratic party, it could make a big difference in close elections, and some of the most important elections in the U.S. are close. As such, while we can confidently rule out large swings in vote share due to vote-by-mail, we cannot say that switching on or off universal vote-by-mail is irrelevant to partisan politics today.
Absentee Voting
As discussed above, absentee voting is a lighter touch version of vote-by-mail that is more commonly used. Instead of mailing every registered voter a ballot, in absentee voting systems, voters can request to vote by mail. States differ in how stringent the requirements are for having a request granted. In “no-excuse” states, anyone who requests a mail ballot can get one. In “excuse” states, on the other hand, the state requires that voters provide an approved reason for needing an absentee ballot.
“Being able to vote more easily by opting into voting by mail does not lead to higher turnout.”
One study took advantage of an unusual rule in the state of Texas, where people aged 65 and older as of Election Day are able to vote absentee without an excuse, while those under age 64 must provide an excuse from a relatively limited set of options.7 This created a natural experiment in which two sets of very similar voters— 65 year olds and 64 year olds—vote in the same elections with the same candidates while facing different vote-by-mail rules. The study found that being old enough to vote absentee without an excuse dramatically increases rates of voting by mail—indicating that people reveal a preference for voting by mail—but that there are negligible effects on turnout rates. That is, being able to vote more easily by opting into voting by mail does not lead to higher turnout.
A second piece of research extended this design in two ways—first, by re-estimating their design during the pandemic, when the effects of being able to vote by mail might plausibly be higher given some people’s health concerns related to voting in person.8 And second, by including Indiana, another state that uses an age cutoff to allow for no-excuse absentee voting. Despite the pandemic, the paper again found negligible effects on turnout. Making it easier to opt into absentee voting does not seem to move turnout, perhaps consistent with the idea that the convenience of voting by mail is not a very relevant consideration to most people when deciding whether to vote or not.
Because absentee voting has such little impact on turnout, it is not surprising that the study also finds it has no effect on partisan outcomes—and in this case, unlike in the case of universal vote-by-mail, the estimated non effect is quite precise.
Conclusion
“We need to better understand the potential tradeoffs to voting by mail, including how it affects risks in ballot security and faith in the electoral system.”
To date, the literature has made significant progress understanding how much expanding access to voting by mail affects participation and partisan outcomes. Moreover, by documenting that many people avail themselves of vote-by-mail when it’s available, these studies suggest that vote-by-mail is popular. Though there is certainly more work that can be done here, there are other foundational questions about vote-by-mail for which we currently lack credible empirical estimates.
The Impact on Democratic Norms
A substantial number of voters reveal a preference for voting by mail, even if its effects on their turnout are modest. We know less about the collective benefits or costs of voting by mail, though. Philosophically, there is a long tradition of believing that the act of coming together to vote in person could be valuable for cultivating democratic norms.9 On the other hand, the realities of voting in person, and especially the issue of waiting in lines to vote, could mean that people become more positive towards democracy when they vote by mail. The overall effect seems ambiguous, and valuable to investigate further.
Confidence in the Process
Second, there is substantial skepticism towards voting by mail among Republicans. While evidence for large-scale fraud is lacking, that does not imply that vote-by-mail has no security or logistical issues. In a survey that asked voters whether they were confident their vote would be properly counted under different modes of voting, respondents demonstrated lower confidence in all-mail elections than in any other mode other than online voting.10 However, a survey can only gauge voters’ attitudes to the principle of voting by mail. The direct experience of actually doing it could change their attitudes in hard-to-predict ways.
In addition, when vote-by-mail causes large delays in reporting the outcomes of elections, this may exacerbate the sense that it is not a well-organized way to run an election. More research should seek to understand these potential tradeoffs to voting by mail, including how it affects tail risks in ballot security, how it affects people’s perceptions of electoral security, and whether that faith is increased in systems where the results of elections are reported rapidly, as in Florida.
Asynchronicity of Voting
Finally, voting by mail makes elections asynchronous —that is, different voters cast their ballots at different times, potentially before or after late-breaking information about the election. Early research suggests that this asynchronicity could affect outcomes,11,12 but the literature is very thin on this question. On the other side of the margin, the ability to vote asynchronously could allow voters more time to study issues and become more informed, and interesting unpublished research supports this idea for downballot races.13
Endnotes
1 Riker, William H., and Ordeshook, Peter C., “A Theory of the Calculus of Voting,” American Political Science Review 62(1) (1968): 25-42
2 Kousser, Thad, and Mullin, Megan, “Does Voting by Mail Increase Participation? Using Matching to Analyze a Natural Experiment,” Political Analysis 15(4): 428–45
3 Gerber, A.S., Huber, G.A., and Hill, S.J., “Identifying the effect of all-mail elections on turnout: Staggered reform in the evergreen state,” Political Science Research and Methods 1(1) (2013): 91-116
4 Thompson, D.M., Wu, J.A., Yoder, J., and Hall, A.B., “Universal vote-by-mail has no impact on partisan turnout or vote share,” Proceedings of the National Academy of Sciences 117(25) (2020): 14052-14056
5 Barber, M., and Holbein, J.B., “The participatory and partisan impacts of mandatory vote-by-mail,” Science Advances 6(35) (2020): 7685
6 Thompson, Wu, Yoder, and Hall, Universal vote-by-mail, 2020
7 Meredith, M., and Endter, Z., “Aging into absentee voting: Evidence from Texas,” Working Paper, 2015 https://www.sas.upenn.edu/~marcmere/workingpapers/AgingIntoAbsentee.pdf
8 Yoder, J., Handan-Nader, C., Myers, A., Nowacki, T., Thompson, D.M., Wu, J.A., Yorgason, C., and Hall, A.B., “How did absentee voting affect the 2020 US election?” Science advances 7(52) (2021): p.eabk1755
9 Chapman, E.B., Election Day: How We Vote and What It Means for Democracy (Princeton University Press, 2022)
10 Fowler, Anthony, “Promises and Perils of Mobile Voting,” Election Law Journal 19(3) (2020): 418-431
11 Meredith, M., and Malhotra, N., “Convenience voting can affect election outcomes,” Election Law Journal, 10(3) (2011): 227-253
12 McCrary, Kaleigh, “Early Voting and Late-Election Information,” Working Paper, 2023 https://www.dropbox.com/s/68le2psww2j7rlj/jmp.pdf?dl=0
13 Szewczyk, James, “How Electoral Institutions Affect Political Accountability: Evidence from All-Mail Elections,” Working Paper, 2018 https://app.box.com/s/5fldh3it32tq9a7alcfzuj36kbqokmj6
About the Democracy Reform Primer Series
Narrowing the gap between research and public dialogue, the University of Chicago Center for Effective Government's Democracy Reform Primers responsibly advance conversations and strategy about proposed changes to our political institutions. Each Primer focuses on a particular reform, clarifies its intended purposes, and critically evaluates what the best available research has to say about it. The Primers do not serve as a platform for either authors or the Center to advance their own independent views about the reform; to the contrary, they serve as an objective and authoritative guide about what we actually know—and what we still don’t know—about the likely effects of adopting prominent reforms to our political institutions.
In some instances, the available evidence may clearly support the claims of a reform’s advocates. In other instances, it may cut against them. And in still others, the scholarly literature may be mixed, indeterminate, or altogether silent. Without partisan judgment or ideological pretense, and grounded in objective scholarship, these Primers set the record straight by clarifying what can be said about democracy reforms with confidence and what requires further study. This series is produced with support from Democracy Fund and the Democracy Innovation Fund.
About the Series Editor
Anthony Fowler is a Professor at the Harris School of Public Policy at the University of Chicago. His research applies econometric methods for causal inference to questions in political science, with particular emphasis on elections and political representation. Fowler is currently the Co-editor in Chief of the Quarterly Journal of Political Science, and the co-author (with Ethan Bueno de Mesquita) of Thinking Clearly with Data: A Guide to Quantitative Reasoning and Analysis (Princeton University Press, 2021). Fowler earned his Ph.D. in government from Harvard University and completed undergraduate studies at the Massachusetts Institute of Technology.