> It's the "perceived" part that I am suspicious of (and I assume the GP too). Poll citizens and ask how much voter fraud they think happened in the last elect

> It's the "perceived" part that I am suspicious of (and I assume the GP too). Poll citizens and ask how much voter fraud they think happened in the last election, in their state, before a voter ID law is passed. Wait for a voter ID requirement to be passed, and another election to happen. Ask again. Compare. Repeat a few different places if you want to be thorough. That's one method, anyway. ("Polled citizens' perception of a thing about the world is wildly wrong, and changes in that perception don't follow changes in reality the way you'd hope they would" is a so-common-it's-boring result in political science)