King's College London

A study focussed on a local election in Bromley, London, which piloted voter ID in 2018, estimates that turnout was reduced by between four and five per cent, with as many as 10,000 people eligible to vote put off from doing so.

The results of the study will be of interest to policymakers as, since October 2023, a new law means all elections in the UK now require voters to show ID.

The findings were revealed in a study published in Political Science and Research Methods, authored by Dr Tom Barton, a postdoctoral researcher with the Department of Political Economy at King’s.

Dr Barton’s research focussed on an election held in the London Borough of Bromley in 2018, which was among a handful of local authorities in the UK that volunteered to pilot voter ID requirements and the only council in London.

Under the requirements, voters needed to show one piece of photo ID or two pieces of non-photo ID in order to receive their ballot paper. Under normal conditions, voters were only required to provide their name and address.

Dr Barton’s study focussed on ward-level voting data from 2002-2018, comparing the average voter turnout over time in Bromley with turnout data from wards across London in the same period.

Dr Barton said: “I found that the pilot scheme did significantly reduce turnout in Bromley, compared to a synthetic counterfactual. The effect estimated implies that of the 240,249 people registered to vote in 2018 in Bromley, 10,571 did not vote who would have otherwise done.

“The number of people turned away and not then returning estimated by the Electoral Commission in Bromley was 154. This large difference between the two figures implies that voter ID laws reduce turnout by deterring people who lack ID rather than catching them unaware.”