New paper w/ @justingrimmer: we link individuals’ ID-holding status to their turnout in North Carolina to estimate the effect of strict voter ID laws. We find NC’s law differentially deterred those w/o ID, and the effect persisted after the law's repeal.
dropbox.com/s/ggl2ry6dyqnp…
dropbox.com/s/ggl2ry6dyqnp…
We show the mechanical effect of the law, where a voter shows up to the polls, casts a provisional ballot because they lack ID, and has the ballot rejected, totaled 1,169 votes in the 2016 primary. These voters were more likely to be Dems & non-white compared to all registrants
To estimate the deterrent effect of the law, we use exact matching and a difference-in-differences design, showing that the photo ID law caused a 1 percentage point turnout decrease for voters without ID relative to those with ID in the 2016 primary election
After the law was suspended by a court ruling, this differential deterrent effect persisted: those without an ID were 2.6 percentage points less likely to turnout in the general election relative to those with ID
Finally, we show because minority voters and Dems are less likely to have ID, the law disproportionately affects them, and because very few voters lack ID, the effect of the law on *overall* turnout is small (3,095 votes), consistent w/ aggregate-level null effects in prior work
