There is a new study that reports evidence that traffic accidents increase on election days. This is an unexpected phenomenon, and I'm not sure why it might be, but I have two main questions about the methodology. First, where are the accidents happening? Is there an uptick near the polling place or are the accidents more or less randomly distributed across town? It wouldn't surprise me that accident increase near a driving to a polling place, where a lot of people are headed who rarely go there. Polling places are sometimes hard to find. They also attract a lot of people walking to and from, so you'd expect an overall increase in pedestrian and vehicle traffic. Perhaps the volume of crashes is consistent with the amount of activity, it's just that there is more of it on election day. I suspect this explains quite a bit of the increase.
The second question is about who is getting into crashes. Elderly people vote in greater numbers than younger people. They also get in more crashes. Maybe this helps explain the problem.
Maybe New York (and a few other states) intuitively know about the dangers of driving on election day which is why they've made it a state holiday. Do the study results hold across states?