University of Wisconsin-Madison political science professor Ken Mayer released results of a study this fall that found changes to the state’s voter ID law led to enormous confusion and deterred thousands from voting.
He’s received some criticism of the study — including questions on his methodology by state Elections Commission member Dean Knudson — but on a Sunday appearance on the talk show “Capital City Sunday" on WKOW/TV, he stood by his findings.
“I think the reason Commissioner Knudson was critical of the study is because the results conflict with what he thinks is true,” Mayer said. “It’s easy to sit back and criticize a study because you don’t like the results. But the issue is that the study we did here is not an outlier.”
Mayer released the study results in September that estimated the Wisconsin voter ID law deterred 16,800 registered voters in Dane and Milwaukee counties from voting in the 2016 presidential election. Mayer said Sunday that some people didn’t vote because they didn’t realize they had valid IDs.
The survey listed 12 forms of ID, asking voters if they believed each was valid for voting. On average, respondents correctly answered about 5 of these questions. Not everyone knew that a Wisconsin driver’s license qualified as a valid ID, Mayer said.
“We found there is an enormous amount of confusion,” he said. “There is almost no one who completely understands the elements of the law in terms of what forms of ID qualify and which forms do not.”
He pointed to one of his colleagues in the UW-Madison political science department, who has a PhD from Harvard. She, too, was confused about what ID was acceptable, he said.
But the survey also found that the laws are less likely to affect Harvard PhDs, as “the burdens fall disproportionately on vulnerable populations,” Mayer said.
Respondents who cited a lack of ID as the reason they didn’t vote only got about 2.5 of the 12 ID questions correct.
“The more likely you are to be affected by this law, the less likely you are to know about it,” Mayer said.
When Mayer reported the findings to the state Elections Commission last week, Knudson, a former Republican state representative, challenged the methodology and validity of the findings.
Knudson was “essentially claiming the survey results were invalid,” Mayer said, and was “insistent that the survey was done incorrectly,” and that it should have been done by phone rather than by mail.
A phone survey was not an option because the survey used WisVote voter files, two-thirds of which did not list a phone number, Mayer said, meaning “essentially the only way you can do a survey like this is through the mail.”
There’s also been objections to what critics say is a small sample size. The project sent out 2,400 surveys to people who were registered to vote in Milwaukee and Dane counties, received 293 back and was able to use 285 that had complete information, Mayer said. The margin of error is calculated to reflect that, but even so, it’s “entirely enough to draw inferences,” he said.
“Two-hundred and eighty-five, to a lay person, that’s not a lot,” show host John Beard said.
“It is unintuitive that you can estimate the properties of a population by sampling one tenth of one percent,” Mayer said. “But the math works.”
He pointed to national surveys which use 1,200 respondents to draw inferences from a population of 300 million.
Mayer said his findings are consistent with, and even more conservative than, similar studies all around the country.
“I’ve been doing surveys for 30 years. There’s no reason to believe that our numbers are incorrect,” he said. “We have been poking at them, looking for evidence that something might be wrong. There’s no evidence of it.”
Dane County Clerk Scott McDonell paid for the $55,000 survey, but Mayer said McDonell didn’t have control or input in how the study was conducted and did not have any advance notice of the effects.
Asked what can be done to better educate voters, Mayer said that went “beyond what we actually studied,” but was skeptical that further information education efforts would fix the problem. Thirty to 35 percent of respondents said they had not heard anything about the voter ID law from anyone, he said.
“In the state’s defense, this was the first time, 2016, that (the voter ID law has) really been put into effect,” Beard said.
Even so, voter lists are constantly changing, Mayer said, as voters move away from or into Wisconsin, lose their eligibility to vote or reach voting age. The 2016 eligible voting population will not be the same eligible voting population for the next race.
“I have some doubts about whether even a large-scale educational campaign will resolve all of this,” Mayer said.