Photo: psdA couple years ago, we wrote a column called “Why Vote?” It didn’t advocate for people to not vote; it just argued that, because of the way the world works, there’s very little value in a single person’s voting.
But according to Donald Redelmeier, a professor of medicine at the University of Toronto, voting might actually be dangerous to your health.
How so?
In a research letter just published in the Journal of the American Medical Association (gated), Redelmeier describes the research conducted by himself and a co-author, Robert J. Tibshirani:
We hypothesized that mobilizing approximately 50 percent to 55 percent of the population, along with U.S. reliance on motor-vehicle travel, might result in an increased number of fatal motor-vehicle crashes during U.S. presidential elections.
Here are some numbers:
A total of 3,417 individuals were involved in fatal crashes during the hours of polling on the 8 election Tuesdays and 16 comparison Tuesdays. The modal person was a young adult driving in a southern state (demographic characteristics generally stable over time). The 8 election days accounted for 1,265 individuals, equivalent to 158 per day, or 13 per hour. The 16 control days accounted for 2,152 individuals, equivalent to 134 per day or 11 per hour. This yielded a relative risk of 1.18 on election days (95 percent CI, 1.10-1.26; P< .001), equivalent to an absolute increase of 189 individuals over the study interval (95 percent CI, 104-280). The net increase in risk was about 24 individuals per election and was fairly stable across decades of time.
In a nutshell: A lot of people are out doing a lot of extra driving on Election Day — which, according to R. and T.’s research, produces extra motor-vehicle deaths. They offer some specific mechanisms:
… distraction (driver inattention), rerouting (unfamiliar pathways), enforcement (decreased police presence), and demographics (mobilizing unfit drivers).
Here’s one writeup of the study; Redelmeier was also featured on NPR.
Redelmeier wrote to tell us that our voting column helped inspire his research on Election Day fatalities. Furthermore, as he put it: “Behold! Chance of dying exceeds chance of casting a pivotal vote!”
If it’s really true that extra people die on Election Day because they’re in a hurry to get back to work from the polling place or whatnot, this is perhaps the most compelling reason yet to make Election Day a national holiday; or at least to switch to mail-in ballots.

It depends how far the polling place is from one person’s office or home and they are open till 8pm so there is no need to rush. Near our house we have a school, a library and a fire station that serve as polling places and all 3 are less than a mile away. My husband and I usually walk there to drop off our “absentee” ballots. This is for Calif since other states have their rules.
Considering early voting, mail in ballots etc, is it really that dangerous to vote? How will this hypothesis apply to Oregon?
Shouldn’t the solution be to use absentee ballots instead of in-person voting rather than proclaiming “voting” in general to be “dangerous”? And if you were to cite increased costs of paper and mailing resources of absentee voting, how about an option of online voting with heightened security measures? If people are comfortable banking and trading online, surely they could be comfortable voting online (especially considering how unlikely it is to be a swing voter).
I am all with Bobby G that online voting should be explored as a viable alternative to in person or absentee voting. The old arguments that online access disadvantages certain demographics is no longer valid with the overwhelming availability of internet access points in modern society.
I am still registered in the neighborhood I grew up in where I usually drive to my old elementary school (some 20 miles south of where I live) to cast my vote. While the nostalgia of the trip is fulfilling, it certainly makes no sense to be making the drive there. I’ll readily concede that it is selfish of me (and possibly illegal?) to keep my registration such that I can affect the vote in my old school district. It is actually gas prices that have led me to switch to now becoming an absentee voter. I found this site convenient for that end: BeAbsentee.org
I already registered to vote by mail! My only concern now is papercuts.
how do they know the people that died voted?- if not, the study is flawed and the conclusion is that not voting on election day increases chance of death
Mail-in ballots? Internet, please! The only way to significantly increase US voter turnout without making it mandatory is to eliminate the “turnout” part and let voters stay home and do it.
Although that raises the question: do we really want the 50% of the population that isn’t making the effort to vote now to be able to make a few simple, uninformed clicks from the comfort of their padded black desk chairs?
Very funny.
All activities have a cost. It’s kind of neat to see that quantified.
I am all for
1) Moving election day to a weekend (perhaps spanning two days) or
2) Making election day a national holiday, or
3) Somehow coming up with a national civic/service day or
4) Combining labor day and election day.
The way things are right now, voter turnout would always be suppressed by logistics, among other things. Elections like this one in November are too important to avoid just because one’s chances of getting in an accident is slightly higher on election day.