10.7910/DVN/Y06T2Y Ahlstrom-Vij, Kristoffer Kristoffer Ahlstrom-Vij https://orcid.org/0000-0003-1081-3171 Birkbeck, University of London Harvard Dataverse 2023 Social Sciences Ahlstrom-Vij, Kristoffer Kristoffer Ahlstrom-Vij Birkbeck, University of London Allen, William William Allen Nuffield College, University of Oxford 2023-04-28 2023-06-05 2470 28001 7051 260450 39468 4255 10096 type/x-r-syntax application/gzip type/x-r-syntax text/tab-separated-values application/gzip type/x-r-syntax type/x-r-syntax 1.0 Creative Commons CC0 1.0 Universal Public Domain Dedication. Since the UK’s 2016 referendum on EU membership, politicians on both sides of the debate have invoked the “will of the people” either to claim the matter is settled or to justify further confirmatory action. Building on prior information effects research in political science, we argue that the extent to which voters are informed about politically relevant issues is a key factor in evaluating whether any particular result accurately reflects an electorate’s collective “will.” By applying counterfactual modeling to British Election Study data collected after the referendum (N=2,067), we estimate that support for leaving the EU would have dropped by up to 10 percentage points, had voters been more fully informed. More generally, we suggest that such modeling exercises—accompanied by transparent theoretical assumptions to enable stress-testing under different conditions—provide diagnostic tools for assessing the sensitivity of electoral outcomes to varying levels of information, and therefore offer relevant insight as to whether they truly capture the “will of the people.” In addition to the data files available here, in order to replicate the analysis, one needs to register for British Election Study data access, download the original SAV file from www.britishelectionstudy.com/data-object/2017-face-to-face/, and add this one to the /data folder, along with the other data files.