The editorial went on to lay out exactly how officials’ estimates matched up with investigation results in each of these states. (There was consistently a large gap.) The Record-Eagle editorial then bluntly says that:
If there is a problem, it lies within the Secretary of State’s office, where people register to vote. They’re asked then to affirm their citizenship, and if the SOS registers them anyway the problem lies within her office.In a nice bit of cross-state media alliance, the editorial gives a nod (but alas, not a link!) to a solid September 19th Detroit News piece by political reporter Chad Livengood on Johnson’s 4,000 estimate, which turns out to be based on citizenship information for one-fifth of Michigan’s population, and extrapolations from an exhaustive SOS investigation that found 963 non-citizens registered to vote in Michigan—only 54 of whom have ever voted, for a total of 95 votes. On September 23rd, the Detroit Free Press gave space to Johnson to state her case in its “opposing points of view” editorial series. Johnson wrote that Michigan’s voter problem comes from 30 years of a federal requirement that compelled secretary of state employees “to ask all customers, regardless of citizenship, if they wanted to register to vote. Wrote Johnson:
Too many did. There were language barriers. They were trying to do the right thing in their new country. Some of them voted.
Today, our secretary of state clerks use a federal database to check citizenship before offering customers a chance to register. But we must still deal with the problem on our voter rolls, and I believe the check box—until we get the federal government to help us solve a problem it created—is our best tool to do that.
Johnson’s column is paired with one by Jocelyn Benson, director of the Michigan Center for Election Law. (Importantly, the Freep identifies Benson as the 2010 Democratic candidate for secretary of state in her bio, though it should’ve explicitly stated that she was Johnson’s opponent.) Benson’s op-ed notes that, “Johnson has irresponsibly declared that 4,000 noncitizens vote in Michigan’s elections.” She argues that there “is no evidence that requiring voters to re-answer the citizenship question every time they vote will stop noncitizens from voting.” Benson suggests alternative solutions to mis-registered voters would include better training for SOS branch employees, where most voters are registered, and contacting suspected noncitizens directly prior to elections.
But Benson misrepresents Johnson’s words. Nowhere has Johnson “declared” that 4,000 noncitizens actually “vote in Michigan’s elections.” Rather, Johnson consistently indicates that 4,000 noncitizens may be on the rolls, and she always couches the number with qualifiers like “could be” and “may be,” as is the case in a story in The Oakland Press. Considering that the 4,000 number is based on several layers of estimation and incomplete date, Johnson’s habit of bandying it about is certainly open to critique—especially since her office has much more fact-based (and much smaller) numbers of noncitizen voters on hand. But journalists (and Benson) should push back on the 4,000 estimate on its face, and should also fairly acknowledge Johnson’s own caveats to the number.
Reporters should also push voting rights advocates for specific evidence that the citizenship question creates an unreasonable burden on voters. The only specific anecdotes I encountered, often provided by voting rights organizations, are of voters who purposefully refused or challenged the question in order to demonstrate a larger point. (One of the prominent figures to do this was Rich Robinson of the Michigan Campaign Finance Network, who I’ve written about as “the perfect source” for political reporters.) This is not to suggest that inconsistencies in how poll workers respond to voters who do and do not answer the question won’t be a potential problem on Election Day.