University of York Study Finds High Error Rates in Polling Clerks’ Photo ID Checks at 2024 Election

The results are consistent with previous research which shows high error rates in ability to check ID.

(IN BRIEF) New research from the University of York published in Royal Society Open Science suggests that the UK’s introduction of mandatory photo ID at the 2024 general election may not have improved electoral security as intended, after polling clerks made errors in 36% of cases when checking real ID photographs and still made mistakes 16% of the time on simpler image comparisons, reinforcing longstanding scientific evidence that unfamiliar face matching is inherently unreliable, with similar error rates observed among university students and a notable decline in accuracy among older clerks, leading lead researcher Professor Mike Burton to argue that current training is unlikely to resolve the issue and that manual identity checks at polling stations remain a fundamentally flawed method of preventing voter fraud.

(PRESS RELEASE) YORK, 4-Feb-2026 — /EuropaWire/ — University of York research suggests that the UK’s introduction of mandatory photo identification at the 2024 general election may not have delivered the security benefits that were originally promised by the government.

A new study indicates that polling clerks responsible for checking voter identity made frequent errors when assessing real identification documents, raising questions about the reliability of human face-matching as a safeguard against electoral fraud. The findings, published in Royal Society Open Science, build on a wider body of evidence demonstrating that even trained professionals struggle to accurately match unfamiliar faces.

During the UK’s first election requiring photo ID at polling stations, researchers evaluated clerks using established face-matching tests. When presented with the Kent Face Matching Test, which uses authentic identification photographs, clerks incorrectly accepted mismatched IDs or rejected valid ones in 36% of cases. Even when the task was simplified — comparing high-quality images taken just seconds apart — errors still occurred 16% of the time.

The research challenges the assumption that in-person identity checks are a dependable barrier to voter fraud. While much of the political debate around voter ID focused on potential impacts on turnout, the effectiveness of the checks themselves received comparatively little scrutiny before the policy was implemented.

Professor Mike Burton from the University of York’s Department of Psychology, who led the study, said the results aligned with decades of scientific research into face recognition. He noted that similar high error rates have been observed among passport officers, police officers, and other professionals tasked with verifying identity.

The study also compared polling clerks with a control group of university students who were not involved in the election. Despite differences in age and experience, both groups made a similar number of mistakes, suggesting that the difficulty lies with the task itself rather than the individuals performing it. The researchers also found that older clerks tended to perform less accurately than their younger counterparts.

Professor Burton explained that the public often underestimates these limitations because people are generally very good at recognising familiar faces, such as friends and family. However, matching unfamiliar faces — as required in a polling station — is a fundamentally different and far more challenging cognitive task.

While some clerks demonstrated excellent performance, achieving perfect accuracy, others performed little better than chance. The researchers concluded that existing training approaches are unlikely to significantly improve outcomes, as previous studies show that face-matching training has minimal impact on accuracy.

Overall, the study suggests that even when carried out diligently and in good faith, reliance on manual face-matching at polling stations is an inherently unreliable method of verifying voter identity.

Media Contact:

Shelley Hughes
Deputy Head of Media Relations (job share)
UK
Tel: work+44 (0)1904 323918
shelley.hughes@york.ac.uk

SOURCE: University of York