Do Voter ID Laws Strengthen or Undercut Democratic Elections?
Putting the US Voter ID Law Debate in Comparative Perspective

Summary  Voter ID laws require showing specific identification at a polling place or proving citizenship when registering. These laws can be evaluated by asking what risk they address, whether they reduce it, and what they require of eligible voters. Debates over voter ID laws in the United States largely overlook the experiences of the many other countries that have experimented with similar regulations.  Experiences in France, Mexico, the United Kingdom, and Albania show that the burden depends on how readily voters can obtain the required documents and how the rule is administered. Even though the laws are slightly different, those lessons help assess proposed U.S. citizenship documentation requirements.  While the effects on electoral fraud are hard to judge, these cases show there are concrete ways to reduce the burden on citizens, such as by using documents citizens already possess or creating a required document that is universally held, and by ensuring citizens have the time and resources to acquire the document well in advance of the election.

Every democracy must protect election integrity while ensuring that eligible citizens can vote. In the U.S. voter ID debate, proponents argue that documentation requirements prevent illegal voting and strengthen public confidence in elections. Opponents argue that the requirements may disenfranchise eligible voters who have difficulty obtaining the required documents.  Both sides tend to also disagree on the magnitude of the problem they seek to address; for example, opponents critique voter ID laws as a disproportionate response to a small problem given the cost to eligible voters, while proponents focus on the importance of eliminating all possible instances of fraud.

The Safeguard American Voter Eligibility (SAVE) Act, passed by the House in February 2026, would require documentary proof of citizenship to register for federal elections and presentation of photo identification in order to vote.¹ The SAVE Act proposal raises questions about how much additional protection of electoral integrity the requirement would provide, how many eligible citizens would face difficulties obtaining or presenting documents, and how election officials would administer it in practice.

Some states in the U.S. have enacted polling-place voter ID laws, which require showing certain types of identification at the polls, not proving citizenship at registration.²  In terms of the benefits of the laws, so far no research has shown conclusively that the laws have reduced fraud.³  Citizens’ perception of fraud also seems to be unchanged, as residents of states with such laws are no more likely to trust elections or to perceive a lower incidence of electoral fraud than those who live in states without the regulations.⁴

In terms of costs, voter ID laws tend to accept more readily available documents, like driver’s licenses or other types of state IDs.  In contrast, the requirements of the SAVE Act are more stringent than some of the state laws, suggesting the cost to voters will be greater than those in state laws. An estimated 12% of registered voters do not have the documents they need to register under the proposed SAVE Act.⁵ This suggests a potentially high burden on citizens and a large administrative effort needed to issue IDs and register citizens to ensure enfranchisement is not affected, even above the burden from the state voter ID laws.  The evidence on turnout of the state laws is mixed, meaning it’s hard to assess the potential cost of the proposed national law.⁶

Many countries have enacted national-level regulations related to voter ID,⁷ so we can also use those international experiences to help us understand how we should think about proposed voter ID laws in the U.S.  Experiences in France, Mexico, the United Kingdom, and Albania help clarify how identification requirements operate in practice. Although their rules differ from the SAVE Act, the cases show how document availability, administrative capacity and independence, and the time allowed for implementation can shape the burdens a requirement imposes. For each case, we consider the problem the rule sought to address, what eligible voters had to do to comply, and the known or suggested consequences of the reform for outcomes like election integrity, public confidence, and voting access.

Assessing whether a certain type of law reduces fraud is incredibly difficult. Such fraud is hard to measure, precisely because those who engage in fraud are trying to hide it.  This creates an inherent difficulty because both proponents and opponents of these laws are basing their arguments on a phenomenon that is largely hidden.

The available evidence, however, does suggest that such fraud in the United States is rare, looking at both academic research as well as the incredibly low numbers of successful prosecutions of illicit voting.⁸ Because illicit voting is rare, it is also hard to detect changes in fraud that are big enough to make it clear those changes were meaningful and not random blips.  Even if such changes could be identified, it would be difficult to know for sure that a new law was the cause.  This creates a particular difficulty in assessing these laws because we don’t know whether the benefit in terms of reducing fraud is worth the cost of potentially disenfranchising eligible voters.

It’s somewhat easier to study the costs of voter ID laws, for example, by tracking how many people lack the documents or by studying voter turnout before and after the law.  In the analysis we present here, we discuss likely benefits and evidence if it exists, but concentrate on the costs, which are more easily observed.

What can we learn from voter ID laws in other countries?

The effects of an identification requirement depend on the problem it addresses and how voters obtain the required documents. France, Mexico, the United Kingdom, and Albania illustrate different approaches to providing identification and administering a new rule.

France — Broad access to identification with little impact on disenfranchisement.

• The context: France abolished mail-in voting in 1975 amid concerns about fraud.⁹ Its polling-place identification requirement has a separate and longer history: voters in larger municipalities have been required to verify their identity since at least 1976, with legislation starting in 1999 requiring that documents include photos.
• What voters must do: Voters in municipalities with at least 1,000 residents must present one of several accepted forms of photo identification.¹⁰ Several features seem to ease compliance. The national identity card is free, multiple alternative documents are accepted, and citizens who turn 18 are generally registered automatically.¹¹ Obtaining a national identity card nevertheless requires an application, a photograph, proof of citizenship, and a visit to a municipal office.¹²
• What happened? The requirement’s effects on fraud, turnout, and public confidence have not been established. While France cannot show whether the benefits of voter identification outweigh its burdens, it does provide an example of a policy whose features may limit those burdens by making identification free, accepting multiple documents, and using automatic registration.

Mexico — Voter identification as part of broader electoral reform.

• The context: Mexico entered the 1990s following disputed elections, with a voter roll that lacked credibility and an election administration system controlled by the government. Reforms created the Federal Electoral Institute in 1990, rebuilt the voter roll, introduced a photographic voter ID, and made the electoral institute fully independent of the executive in 1996.¹³
• What voters must do: Citizens must register, obtain the free credential from an electoral office, and present it to vote. By 1994, approximately 92 percent of voting-age citizens had obtained one.¹⁴
• What happened? Mexico’s 2000 election, which produced the country’s first transfer of presidential power between parties in more than seven decades, was judged generally free and fair.¹⁵ That outcome followed the introduction of the new voter roll and credential as well as broader reforms to election administration, campaign finance, and party competition. The separate contribution of the identification requirement on reduced fraud therefore cannot be fully established. Mexico instead shows how voter identification can form one part of a broader effort to build a more credible electoral system.

The United Kingdom — Addressing a potential vulnerability but introducing significant costs.

• The context: Personation—casting a ballot in someone else’s name—was already rare. The Electoral Commission recommended polling-place identification in response to the potential vulnerability created by the absence of an ID requirement, rather than in response to evidence of widespread fraud.¹⁶ Voters must present one of several accepted forms of identification or obtain a free Voter Authority Certificate. Uptake of the free certificate has remained low relative to estimates of the number of voters without other accepted identification.¹⁷
• What happened? The government’s evaluation found insufficient evidence to determine whether the requirement deterred personation, which remains rare.¹⁸ Public confidence was already high, suggesting trust was unlikely to increase: in 2022, before the reform was implemented, 87 percent of respondents believed that voting at polling stations was safe from fraud and abuse.¹⁹ In terms of cost, the requirement created measurable barriers. Approximately 14,000 people who went to polling stations in May 2023 and 16,000 at the 2024 general election were unable to vote because they lacked accepted identification.²⁰ These estimates do not include people who decided not to attempt to vote because of the requirement.²¹ The evidence therefore establishes some access costs with little evidence of benefits.

Albania — Targeting likely fraud with a compressed rollout and uneven access.

• The context: Before the 2009 parliamentary election, voters without passports could use birth certificates with photographs as identification. These documents were considered insecure and had routinely generated allegations of misuse and politically motivated manipulation.²²
• What voters must do: Under the 2008 Electoral Code, voters had to present either a valid passport or a new biometric identity card. Issuance of the cards began in January 2009, less than six months before the election.²³ Applicants had to visit an application center and pay a fee, although subsidies were available for some groups. Delays in opening centers outside the capital, Tirana, as well as infrastructure failures and inconsistent fee assistance created additional obstacles.²⁴
• What happened? International observers described the distribution campaign as broadly successful despite the constraints: all but 3,321 of more than 1.4 million applications were processed before election day. Yet nearly 78,000 completed cards remained uncollected, and approximately 257,000 citizens recorded as lacking passports did not apply. These figures do not indicate how many of those citizens were in Albania and intended to vote; government estimates suggested that many were living abroad. Observers nevertheless documented concrete access problems, including application centers that were inaccessible to some voters with disabilities and inconsistent implementation of fee exemptions for Roma applicants. They also verified instances of governing-party activists becoming improperly involved in the process.²³ Albania therefore shows how a short implementation period and limited administrative capacity can create unequal barriers and opportunities for political involvement, even when most applications are processed successfully.

Implications for the United States

Direct burdens on eligible citizens. The international cases show that these burdens depend on more than whether a free document is theoretically available. France accepts numerous forms of identification and provides a free national identity card, features that may make its requirement easier to meet. Mexico achieved broad coverage through a free voter credential administered as part of a national electoral registry. By contrast, some British voters were unable to vote despite the availability of a free Voter Authority Certificate, and Albania’s compressed rollout and fee-based design created obstacles related to costs, travel, infrastructure, and administrative capacity. Free documents and alternative procedures can reduce burdens, but their availability does not ensure that every eligible voter can use them.  Still the international cases suggest that free documents given universally, alongside automatic registration, can reduce the burden on voters.

In the United States, the immediate documentation burden could be substantial, particularly compared with systems that accept many common forms of identification or provide a widely held voter credential. The United States, in contrast to most countries, lacks a free, universally held national document that establishes citizenship.²⁵ Under the SAVE Act, applicants would generally have to produce a passport or combine documents establishing citizenship and identity.²⁶ Name discrepancies may require additional documentation.²⁷ A national survey estimated that 21.3 million voting-age citizens lacked ready access to common citizenship documents, including 3.8 million who reported possessing none of them.²⁸ These figures do not estimate how many people would ultimately be unable to register, but they indicate how many could face an additional documentation task. The requirement would also create new responsibilities for state and local election officials responsible for receiving and verifying the documents.

Potential and realized benefits. The international cases provide limited evidence about the laws’ effectiveness in preventing or deterring illegal voting, because it’s both hard to detect fraud and hard to be sure any changes are attributed to a specific policy.  Mexico’s more credible elections followed the introduction of a new voter roll and credential alongside broader reforms to election administration, campaign finance, and party competition. The contribution of voter identification therefore cannot be separated from the larger reform package. The United Kingdom provides a clearer view of a stand-alone polling-place requirement, but the government’s evaluation found that it lacked evidence to assess the effect on fraud through personation, noting that even if such fraud was found to decline over time, it would be very difficult to know if it was due to the new voter ID law or not.²⁹

Likely benefits in the United States. The SAVE Act addresses noncitizen registration and voting, rather than the fraud targeted by most polling-place ID requirements. Available evidence suggests that noncitizen voting is rare, although its precise prevalence is difficult to measure. In a study of 42 jurisdictions that counted 23.5 million votes in 2016, election officials referred approximately 30 suspected incidents for investigation.³⁰ This was not a count of confirmed votes or a nationally representative estimate. It does, however, indicate that the proposal addresses conduct that was rarely detected in the jurisdictions studied. The extent to which documentary proof would prevent additional cases or increase confidence remains unknown. Assessing the proposal therefore requires weighing these potential benefits against the burdens it would impose on eligible citizens, and the administrative demands and public resources required to implement it.

Broader effects. Identification requirements could also affect turnout and public confidence. Documentation burdens could reduce participation, while greater confidence or mobilization in response to the law could increase it. Existing U.S. research is mixed: a Government Accountability Office review of ten studies of state voter ID requirements found four reporting decreases in turnout, five finding no statistically significant effect, and one reporting an increase.³¹ Studies of state-level voter ID also found no impact on public trust in elections.³²

Reducing the Cost of Voter ID in the U.S. The international cases make it clear that how the requirements are designed and implemented can reduce the chance that eligible citizens are unable to vote.  For example, accepting multiple forms of identification, ensuring that the ID types are those that citizens already use, issuing national ID cards that all citizens uniformly possess, and automatically registering citizens to vote can all help reduce the burden on citizens, thereby ensuring that any possible reduction in cases of fraud does not come at the cost of disenfranchising eligible voters.

Authors

Rodrigo Balvanera

Laura Paler

Julie Anne Weaver

Sources

1. H.R.22, 119th Congress. “Safeguard American Voter Eligibility Act.” Congress.gov, www.congress.gov/bill/119th-congress/house-bill/22/text; “Safeguard American Voter Eligibility Act (SAVE America Act)
   and Federal Voter Registration Policy and Law Congressional Research Service, Feb. 24, 2026, https://www.congress.gov/crs_external_products/IF/PDF/IF12902/IF12902.5.pdf
   [↩]
2. “Voter Identification.” MIT Election Data Science Lab, electionlab.mit.edu/research/voter-identification. [↩]
3. Cantoni, Enrico, and Vincent Pons. “Strict ID Laws Don’t Stop Voters: Evidence from a U.S. Nationwide Panel, 2008–2018.” Quarterly Journal of Economics, vol. 136, no. 4, 2021, pp. 2615–2660, academic.oup.com/qje/article-abstract/136/4/2615/6281042. [↩]
4. Stewart, Charles, III, Stephen Ansolabehere, and Nathaniel Persily. “Revisiting Public Opinion on Voter Identification and Voter Fraud in an Era of Increasing Partisan Polarization.” Stanford Law Review, volume 68, June2016, review.law.stanford.edu/wp-content/uploads/sites/3/2016/08/6_-_Persily_-_Stan._L._Rev.pdf; Cantoni and Pons, “Strict ID Laws Don’t Stop Voters.” [↩]
5. “Six Things to Know About the SAVE America Act.” Bipartisan Policy Center, 17 July 2026, bipartisanpolicy.org/article/five-things-to-know-about-the-save-act/. [↩]
6. Cantoni and Pons, “Strict ID Laws Don’t Stop Voters”; Hajnal, Zoltan, et al. “Voter Identification Laws and the Suppression of Minority Votes.” Journal of Politics, vol. 79, no. 2, 2017; U.S. Government Accountability Office. “Elections: Issues Related to State Voter Identification Laws.” GAO-14-634, Sept. 2014, www.gao.gov/products/gao-14-634. [↩]
7. Barton, Tom. “Understanding a Key Electoral Tool: A New Dataset on the Global Distribution of Voter Identification Laws.” Representation 60.2 (2024): 253–277. Data: Harvard Dataverse, doi:10.7910/DVN/RU2F9A. [↩]
8. Goel, Sharad, Marc Meredith, Michael Morse, David Rothschild, and Houshmand Shirani-Mehr. “One Person, One Vote: Estimating the Prevalence of Double Voting in US Presidential Elections.” American Political Science Review 114, no. 2 (2020): 456–469; U.S. Government Accountability Office, “Elections: Issues Related to State Voter Identification Laws”; Christensen, Ray, and Thomas J. Schultz. “Identifying Election Fraud Using Orphan and Low Propensity Voters.” American Politics Research 42, no. 2 (2014): 311–337. [↩]
9. “Explainer: How France’s Old-School Voting System Works.” Associated Press, The Hill, 24 Apr. 2022, thehill.com/homenews/ap/ap-business/explainer-how-frances-old-school-voting-system-works/; “French Voters Should Consider Election Integrity Issues as They Vote in 2022.” German Marshall Fund, www.gmfus.org/news/french-voters-should-consider-election-integrity-issues-they-vote-2022. [↩]
10. Service-public.gouv.fr. “Elections: Identity Papers to Be Presented to Vote.” French Government, www.service-public.gouv.fr/particuliers/vosdroits/F1361?lang=en. [↩]
11. “Identity Card (France).” Wikipedia, en.wikipedia.org/wiki/National_identity_card_(France); Service-public.gouv.fr, “Elections: Identity Papers to Be Presented to Vote”; Service-public.gouv.fr. “Lists of Electors: In Case of First Registration.” French Government, www.service-public.gouv.fr/particuliers/vosdroits/F1367?lang=en. [↩]
12. https://passeport.ants.gouv.fr/home/online-services/applying-for-a-national-identity-card-for-the-first-time [↩]
13. “Mexico’s National Electoral Institute — Explainer.” Wilson Center, www.wilsoncenter.org/article/mexicos-national-electoral-institute-explainer; Klesner, Joseph L. “Electoral Reform in Mexico’s Hegemonic Party System: Perpetuation of Privilege or Democratic Advance?” Kenyon College, www2.kenyon.edu/Depts/PSci/Fac/klesner/Electoral_Reform_in_Mexico.htm. [↩]
14. “Electoral Registry.” Instituto Nacional Electoral, ine.mx/electoral-registry/. [↩]
15. “Mexico’s National Electoral Institute: Ensuring Fair Elections at the Local Level.” Baker Institute, Rice University, 30 June 2014, www.bakerinstitute.org/research/mexico-ensuring-fair-elections-local-level. [↩]
16. “Voter ID.” House of Commons Library, Research Briefing CBP-9187, 5 October. 2026, commonslibrary.parliament.uk/research-briefings/cbp-9187/; “Briefing on Voter ID.” Electoral Reform Society, 2026, https://electoral-reform.org.uk/latest-news-and-research/parliamentary-briefings/briefing-on-voter-id-2026/)
17. What voters must do: The Elections Act 2022 introduced a photo ID requirement for polling-place voting, beginning with local elections in England in May 2023. ((House of Commons Library, “Voter ID.” [↩]
18. Fieldhouse, Edward, Ralph Scott, Christopher Prosser, Jack Bailey, and Jonathan Mellon. “Who Lacks Voter Identification? The Electoral Implications of the Elections Act 2022.” Parliamentary Affairs, vol. 78, no. 4, Oct. 2025, pp. 695–715, academic.oup.com/pa/article/78/4/695/7917523; Electoral Reform Society, “Briefing on Voter ID.” [↩]
19. IFF Research (2025) Electoral Integrity Programme Evaluation: Year 2. London: Ministry of Housing, Communities and Local Government. Available at: https://assets.publishing.service.gov.uk/media/68245b4deef73e50f3644af9/Electoral_Integrity_Programme_Evaluation_Year_2_Report.pdf (Accessed: 2 October 2026). [↩]
20. Electoral Reform Society, “Briefing on Voter ID. [↩]
21. Scott, Ralph, and Edward Fieldhouse. “How Have Voter ID Requirements Affected British Elections?” The Constitution Unit Blog, UCL, 21 Feb. 2025, constitution-unit.com/2025/02/21/how-have-voter-id-requirements-affected-british-elections/. [↩]
22. Electoral Reform Society, “Briefing on Voter ID.” [↩]
23. OSCE/ODIHR. “Republic of Albania, Parliamentary Elections, 28 June 2009: Election Observation Mission Final Report.” 14 Sept. 2009, www.osce.org/sites/default/files/f/documents/9/8/38597.pdf; “General Elections 2009 Albania.” Robert Schuman Foundation, www.robert-schuman.eu/en/monitor/893-general-elections-in-albania-28th-june-2009. [↩]
24. OSCE/ODIHR, “Republic of Albania, Parliamentary Elections, 28 June 2009.” [↩] [↩]
25. OSCE/ODIHR, “Republic of Albania, Parliamentary Elections, 28 June 2009”; Robert Schuman Foundation, “General Elections 2009 Albania.” [↩]
26. Barton, “Understanding a Key Electoral Tool.” [↩]
27. H.R.22, “Safeguard American Voter Eligibility Act.” [↩]
28. “What You Need to Know About the SAVE Act.” Campaign Legal Center, campaignlegal.org/update/what-you-need-know-about-save-act; Center for American Progress, “The SAVE Act: Overview and Facts.” [↩]
29. “Millions of Americans Don’t Have Documents Proving Their Citizenship Readily Available Brennan Center for Justice, 2024, https://www.brennancenter.org/our-work/analysis-opinion/millions-americans-dont-have-documents-proving-their-citizenship-readily. [↩]
30. IFF Research, Electoral Integrity Programme Evaluation: Year 2. [↩]
31. “Noncitizen Voting: The Missing Millions”, Brennan Center for Justice, 5 May 2017,https://www.brennancenter.org/our-work/research-reports/noncitizen-voting-missing-millions [↩]
32. U.S. Government Accountability Office, “Elections: Issues Related to State Voter Identification Laws.” [↩]
33. Stewart, Ansolabehere, and Persily, “Revisiting Public Opinion on Voter Identification and Voter Fraud”; MIT Election Data Science Lab, “Voter Identification.” [↩]