Prior research predicts that election administration changes that increase voting costs should decrease participation, but it fails to consider that some interpret those changes as attacking their franchise. Drawing on psychological reactance theory, this study tests whether such perceived attacks might instead activate those citizens. It leverages the State of Florida’s multi-stage effort in 2012 to purge suspected non-citizens from its voter rolls, comparing the voting rates of suspected non-citizens whose registration was and was not formally challenged by the state. Within-registrant difference-in-difference and matching analyses estimate a positive, significant participatory effect of being challenged, particularly for Hispanics (the vast majority of the sample). Placebo tests show that those challenged were no more likely than those not challenged to vote in previous elections.
To date, roughly 72% of all ballots cast early in-person and by mail in Florida as of Sunday have been by voters who cast ballots four years ago in the 2014 midterm elections.
In one of the better signs for Democrats, Republicans have cannibalized more of their Super-voters thus far: 76.6% of votes cast by Republicans have been by those who voted in 2014. Slightly fewer, 71.5%, of votes cast by Democrats have been by those who voted in 2014 (so, by definition, no newly registered 18 & 19 year olds). The most interesting development thus far is that only 60.7% of votes cast by No Party Affiliates have been by those who skipped (or were not yet registered) in the 2014 midterms. Of course, NPAs have cast fewer than half of votes cast by either Republicans or Democrats.
The daily patterns are remarkably similar across the three groups, as the figures below reveal. Regular voters were much more likely to vote by mail, earlier in the cycle; more infrequent voters started to tick up during the first week of early in-person voting.
I plan to provide similar figures in the coming days, focusing on age and race, broken down by 2014 voters and nonvoters.