Careful New Study Finds at Least Thousands in Two Wisconsin Counties Didn’t Vote Because of Voter ID Requirements, Confusion Over Them

Careful new study led by Ken Mayer:

A survey of registered voters in Dane and Milwaukee Counties who did not vote in the 2016 presidential election found that 11.2% of eligible nonvoting registrants were deterred by the Wisconsin’s voter ID law. This corresponds to 16,801 people in the two counties deterred from voting, and could be as high as 23,252 based on the confidence interval around the 11.2% estimate, which is between 7.8% and 15.5%. The survey further found that 6% of nonvoters were prevented from voting because they lacked ID or cited ID as the main reason they did not vote, which corresponds to 9,001 people, and could be as high as 14,101 based on the confidence interval of between 3.5% and 9.4%.

Roughly 80% of registrants who were deterred from voting by the ID law, and 77% of those prevented from voting, cast ballots in the 2012 election.

Based on these estimates, if all of the affected registrants voted the voter ID requirement reduced turnout in the two counties by 2.24 percentage points under the main measure of effect, and by 1.2 percentage points under a conservative measure. If they voted at 2012 rates, voter ID lowered turnout by 0.9 to 1.8 percentage points.


The burdens of voter ID fell disproportionately on low-income and minority populations.  Among low-income registrants (household income under $25,000), 21.1% were deterred, compared to 7.2% for those over $25,000.  Among high-income registrants (over $100,000 household income), 2.7% were deterred.


8.3% of white registrants were deterred, compared to 27.5% of African Americans.

The study, conducted by Principal Investigator Kenneth R. Mayer, Professor of Political Science and Affiliate Faculty of the Robert M. La Follette School of Public Affairs and UW Madison, with Ph.D. candidate Michael  G. DeCrescenzo, was based on the statewide database of registered voters (WisVote), which records whether a registrant cast a ballot in the November presidential election.  The survey was administered by the UW Survey Center, and funded by the Dane County Clerk’s Office.  The data are based on a sample of 288 nonvoting registrants who were on the rolls on or before election day, November 8, 2016.

The survey asked registrants about their reasons for not voting, the types of ID they possess, interest in the election, confidence in the accuracy of the vote count, and demographics. The survey did not ask voters about who they would have voted for or their party identification.

The survey found considerable confusion about the law. Most of the people who said they did not vote because they lacked ID actually possessed a qualifying form of ID.  This confusion may be the result of a lack of effective efforts educating eligible voters of the requirements of the law, and it is consistent with other studies that show many otherwise eligible voters are confused about ID laws.  There were no significant differences between people who had seen information about the voter ID law and those who had not.

“This study provides better data than previous efforts to measure the effects of ID laws, which have largely been based on aggregate turnout, matching registered voters to state driver’s license and ID databases, or looking at the number of rejected provisional ballots cast by voters without an ID” said PI Mayer.  “By asking nonvoters their reasons for not voting, and about what forms of ID they actually possess, we get a better understanding of how voter ID laws affect individuals, and what types of people are most deterred by the laws.  The data show that poor and minority populations are affected the most.”

“The main conclusion of the study is that thousands, and perhaps tens of thousands, of otherwise eligible people were deterred from voting by the ID law,” said Mayer. “The 11.2% figure is actually a lower bound since it does not include people who don’t even register because they lack an ID.  And while the total number affected in Milwaukee and Dane Counties is smaller than the margin of victory in the 2016 presidential election, that is the wrong measure. An eligible voter who cannot vote because of the ID law is disenfranchised, and that in itself is a serious harm to the integrity to the electoral process.”

See also supporting information and the FAQ.  From the FAQ:

You estimated the number of people in Milwaukee and Dane Counties who were deterred from voting because of Voter ID. Do you know how many people statewide were affected?

No. The sample was drawn from nonvoting registrants in Dane and Milwaukee Counties. The estimate of the effect applies only to the total number of registrants there who were deterred from voting because of Voter ID. The 11.2% figure cannot be directly extrapolated statewide, because we do not know how people outside of Dane or Milwaukee Counties would have answered the questions about their reasons for nonvoting or whether or not they possess a qualifying form of photo ID. The statewide totals outside of Dane and Milwaukee are certain to be greater than zero, but we cannot assume that the effect was the same, 11.2%.

Three brief points:

  1. These effects seem real and this careful study seems much sounder than the earlier Priorities study finding up to 200,000 voters statewide affected by ID (a study Hillary Clinton relies upon in her new biography).
  2. It is interesting how much of the deterrent effects from voter id laws comes from confusion and misinformation. That’s a feature, not a bug, and shows that the details of implementation matter as much as the law itself.
  3. While turnout effects (and electoral outcomes) interest a lot of folks, I continue to believe that this is not the central question about voter id and similar laws. The question goes to the dignity of each voter and asks why the state should be able to make it harder for people to vote for no good reason (and these laws don’t seem to stop any appreciable amount of fraud).
Share this: