Willing to Vote, but Disenfranchised by a Costly Registration Process: Evidence from a Randomized Experiment in France

Abstract

We study the impact of door-to-door canvassing interventions implemented in 2011 on the registration rate and participation of newly registered citizens at the 2012 French Presidential and General elections. 4,118 buildings hosting 38,000 citizens were randomly allocated to a control group or one of six treatment groups varying by the timing and number of visits as well as their content: simple door-to-door canvassing (providing encouragement and information), or home registration. In addition to varying the intensity of the treatment, this experimental design enables us to disentangle the selection impact from the treatment impact of the more intensive home registration visits on the propensity to vote of newly registered citizens. Our interventions increased the number of new registrations by 30% on average, and citizens registered thanks to our visits were almost equally as likely to cast a ballot at the subsequent elections as newly registered citizens in the control group: arguably, higher registration costs play a more important role than lower interest in politics in explaining their failure to register, absent any intervention. Home registration selected additional citizens characterized by a lower propensity to vote than those also selected by door-to-door canvassing, a difference which results from a slightly lower underlying interested in the elections, rather than a negative treatment impact of getting registered at home on the beneficiaries' motivation and involvement. Finally, the interventions increased the share of registered citizens who were born abroad and selected younger people, compared to previously registered citizens. They selected citizens living in addresses in which the average propensity to vote was lower than for the newly registered citizens in the control group: although identical for all, the registration process imposes a higher cost to some subgroups of the population, it reinforces preexisting exclusions and undermines the representativity of the electoral outcomes.