Research review: turnout effects from voter ID are often small or hard to detect
Annual Review synthesis notes plausible mechanisms for suppression but empirical estimates tend to be small; measurement hurdles are substantial.
All topicsVoter ID laws · 6 peer-reviewed studies for, 6 against
The evidence on voter ID laws is genuinely mixed on aggregate turnout, but the strongest policy-relevant takeaway is distributional: even when overall turnout effects are small or statistically indistinguishable from zero, the burdens can fall unevenly. Large-scale panel evidence in QJE finds no average turnout or registration decline from strict ID requirements, and some analyses report neutral or even positive turnout effects in certain election contexts. Supporters also argue that ID requirements can improve perceived integrity, and experimental evidence suggests information about ID rules can reduce fraud perceptions among registrants. However, validated-turnout studies in major political science outlets find that strict ID laws can differentially reduce minority turnout and shift the participating electorate, and follow-on work emphasizes that “no average effect” can coexist with meaningful harms concentrated in specific groups or places. Given that in-person voter fraud is rare while the right to vote is foundational, the evidentiary burden for imposing a potentially unequal barrier is high. Net: if adopted at all, the case improves only with broad, free, low-friction ID access, generous alternatives, and careful monitoring for disparate impact.
Ranked by citation count.
Annual Review synthesis notes plausible mechanisms for suppression but empirical estimates tend to be small; measurement hurdles are substantial.
Using validated turnout data, finds strict voter ID laws are associated with lower turnout among racial and ethnic minorities and can shift the electorate rightward.
Annual Review synthesis notes plausible mechanisms for suppression but empirical estimates tend to be small; measurement hurdles are substantial.
American Politics Research assessment of Georgia’s statute reports modest to negligible turnout changes at the aggregate level.
Difference-in-differences analysis finds no negative effect of strict voter ID laws on registration or turnout overall or across major demographic groups.
Harden and Campos (2023) examined voter ID laws' effects on electoral outcomes across state legislatures, governorships, Congress, and presidential races from 2003-2020. The study found negligible average effects on election results, with voter ID requirements motivating supporters of both parties equally to comply rather than producing partisan advantages.
Field evidence indicates information about photo-ID requirements is associated with reduced perceptions of electoral fraud among registrants.
Reports heterogeneous estimates and finds turnout increases in some midterm contexts after strict voter ID implementation, suggesting effects can differ by election type and adoption timing.
Using validated turnout data, finds strict voter ID laws are associated with lower turnout among racial and ethnic minorities and can shift the electorate rightward.
Politics, Groups, and Identities article reports that strict ID requirements have disproportionate effects by race/ethnicity after adjusting for covariates.
Synthesizes and tests evidence that strict voter ID laws can reduce minority participation under certain conditions, highlighting mixed results but meaningful risks of disparate impact.
Argues and documents that strict voter ID requirements impose disproportionate burdens on minority voters, emphasizing distributional inequities even when aggregate effects appear small.
Coll & Clark (2024) find that voter ID laws differentially affect voter confidence by race: Black voters experience decreased confidence in elections when restrictive voting policies are enacted, while white voters' confidence increases, reflecting different historical perceptions of these laws based on their communities' experiences with voting restrictions.
Braun's study of Virginia's 2014 voter ID law finds significant and durable declines in both registration and turnout rates. The research shows the decline in turnout is driven almost entirely by declines in registration, suggesting voter ID laws have an important deterrent effect preventing new voters from participating.
Carlos: Let's start with something simple. Showing ID is normal. You need it to board a plane, pick up a prescription, even rent a movie at some places. Why should voting be the exception?
Derek: Because voting isn't a purchase, it's a constitutional right. When you restrict a right, the burden of proof is on the people restricting it.
Carlos: And that burden is met. Voter ID protects election integrity. People need to trust the outcome, and ID is a visible safeguard.
Derek: Trust matters, I'll give you that. But the fraud ID laws target, in-person impersonation, is vanishingly rare. We'd be building a fortress around a door almost nobody tries to open.
Carlos: Rare isn't never, and prevention isn't only about catching crimes. Experiments show that when registrants learn about ID rules, their fraud concerns go down. Confidence is a real public good.
Derek: It is, and that's your strongest point. But perception can be improved in other ways, like audits and transparent counting, without risking anyone's ballot.
Carlos: Here's what really should settle it, though. A major study in the Quarterly Journal of Economics, using large-scale panel data, found strict ID requirements caused no average drop in turnout or registration. Some analyses even show neutral or positive effects. So where's the harm?
Derek: That study is important, and I won't pretend otherwise. But look at the phrase "no average effect." Averages can hide a lot. Validated-turnout studies in top political science journals find strict ID laws can reduce turnout among minority voters and shift who makes up the electorate.
Carlos: Those findings are contested, and other researchers don't see the same patterns.
Derek: Which is why the evidence is mixed overall. But think about what mixed means here. If the best case is "probably no big aggregate effect," and the worse case is "meaningful harm concentrated in specific groups or places," that's not a comfortable bet when the right at stake is foundational.
Carlos: So you'd never accept any ID law?
Derek: I didn't say that. If a state adopts one, it needs broad, free, low-friction ID access, generous alternatives like signed affidavits or vouching, and real monitoring for disparate impact. Fix the access problem first, then talk about requirements.
Carlos: That sounds like conceding my side can work.
Derek: It sounds like conceding that your side can work only if it stops looking like your side. Strict ID with limited access isn't the policy that survives scrutiny. The cushioned version is, and at that point we're debating safeguards for voters, not safeguards against them.
Carlos: But even then, supporters get what they want: integrity, confidence, a clear standard.
Derek: And if the data later shows certain communities are being turned away, who fixes that? A person who loses their vote can't get an election redone. The costs of being wrong are uneven, and they fall on the people with the least cushion.
Carlos: So your bottom line is caution.
Derek: My bottom line is proportionality. Rare fraud, uncertain benefits, possible unequal burdens. When the right to vote is on the line, that math says the case against stays stronger.
Carlos: I'd say the math says trust and security win.
Derek: Then let's make it secure without making it harder to vote. If we can't do both, the right to vote goes first.
The evidence is mixed. Some large studies find no average decline in turnout or registration, and one finds increases in some midterm contexts, while others report declines, including a Virginia study tied to lower registration.
Several studies using validated turnout data find strict ID laws are associated with lower turnout among racial and ethnic minorities. Researchers note that a small or zero average effect can still hide harms concentrated in specific groups or places.
A study of election results across state and federal races found negligible average effects on outcomes, with supporters of both parties complying at similar rates. Another study found strict ID laws could shift the participating electorate rightward.
In-person voter fraud is rare. Information about photo ID rules has been linked to lower fraud perceptions among registrants, but confidence effects differ by race, with Black voters' confidence falling and white voters' rising after restrictive policies.