As this blog noted, Cleta Mitchell was invited to give a “special legal presentation” to a recent RNC donor retreat — either because of or despite her role in proceedings attempting to overturn the 2020 election. 

There has already been widespread reporting on Mitchell’s comments on student voting.  There’s a lot more there there in another round of audio posted on Friday, including claims about litigation that she voluntarily dismissed before it could draw sanctions, and calls for tax-deductible contributions to her nonpartisan 501(c)(3) nonprofit less than 60 seconds after criticizing Alaska’s ranked-choice voting system because of its perceived partisan electoral impact.

In this (longish: apologies) post, though, I want to briefly address two of her claims concerning government entities that aren’t in a position to fight back.  One claim is almost certainly false and the other is false on its face.  To be abundantly clear, I don’t speak for either government entity.  But in an era of attack on institutions, it’s (still) important to have actual facts out there.

First: the Census.  Mitchell claims that the Census Bureau “literally manipulated” the apportionment system to give seats to blue states and take seats from red states, and admitted their “mistake” last May.  What she seems to be talking about is post-enumeration analysis that Census has run in some fashion every decade since 1950, attempting to assess what the decennial Census got right and where there’s room for improvement.  These aren’t corrections to the decennial Census, which is, by law, not correctible for apportionment purposes using statistical samples; instead, it’s an after-action review to help prepare for surveys to come and for the next enumeration.  In May of 2022, Census released its post-enumeration analysis for the 2020 Census; the assessments of error are themselves estimates and subject to error, but Census’s best calculation was that there were likely undercounts in six states (AR, FL, IL, MS, TN, and TX) and overcounts in eight (DE, HI, MA, MN, NY, OH, RI, UT). 

It is true that Minnesota very narrowly gained a seat in apportionment (by just 26 people), and that Florida and Texas results did not measure up to their expected count.  It is not true that Minnesota’s gain is straightforwardly Florida or Texas’s loss: that’s not how the formula for apportioning congressional seats works.  (Indeed, if anything, that next seat would likely have gone to even-bluer New York, which dropped a seat it would have kept with 89 more individuals.)  

The casual charge of partisan manipulation doesn’t fit the facts.  If the over- and undercounts revealed by the post-enumeration analysis were the result of partisan manipulation, that conspiracy would have to account for at least five factors: 1) the fact that the data for the 2020 Census were collected and collated during the Trump Administration by the same career officials who then reported them the next year, 2) the fact that the alleged conspirators transparently published and voluntarily released their self-reflective analysis (in line with practice for the last 70 years), violating Bell’s First Law, 3) the fact that deep-blue Illinois was undercounted and deep red Ohio and Utah were overcounted, 4) the fact that Latino, Black, and Native American populations were significantly undercounted and non-Hispanic whites were significantly overcounted, and 5) the fact that Montana was also on the knife’s edge to stay pat with one seat in Congress but actually gained another by a hair, while New York lost a seat it would have kept with just 89 more people.  None of this makes sense in the world of the conspiracy.

Fortunately, there are simpler explanations for the under- and overcounts.  First and foremost, the pandemic threw the biggest wrench in a century into Census operations, particularly in states like New York hammered by the front wave of the virus on Census Day; some significant degree of error is to be expected from that fact alone, and the fact that 14 states had under- or overcounts is far less surprising than the fact that 36 states did not.  Second, in the middle of a prominent public campaign around immigration enforcement, the Trump Administration’s yearlong pursuit of a citizenship question on the decennial Census form, though ultimately overturned by the courts, may still have succeeded to some degree in dissuading the participation of a vulnerable population; it was entirely predictable that states comparatively less responsive to significant immigrant communities would see an undercount. Third, some states chose to invest heavily, in partnerships or funding or both, in ensuring a complete count — and some states, like Texas and Florida, did not … or invested only at the very last minute.  It’s not hard to imagine that investment (or its absence) had consequences, particularly for an undercount.

It’s not possible to know which of these factors — or others not dependent on allegations of malfeasance — actually contributed to the under- and over-counts in what degree.  But they’re all more facially plausible sources of deviation than the unsubstantiated claims of rigging the system.  If Mitchell has evidence that the Census Bureau manipulated the count, I’d love to see that evidence.

Second: the Department of Education.  Mitchell claims that the Biden Administration’s executive order on promoting access to voting spurred ED to issue guidance that every college and university receiving federal funds has to “have a voter registration as part of the student enrollment package.  They have to have a voter registration form as part of registering for classes.  If you apply for student financial aid, you have to fill out a voter registration form.”

Look, I think Mitchell’s idea would make great policy.  Giving eligible Americans more opportunities where they can conveniently choose to register to vote or update their registration is in my view a good thing; providing these opportunities when Americans are otherwise filling out paperwork so they can do two things at one time more efficiently is in my view a good thing; increasing the routinization of registration at institutions so that local officials get reliable and updated information more effectively, to help keep the rolls cleaner, is in my view a good thing.  At one point in her presentation, Mitchell decried hordes of students lining up for same-day registration in Wisconsin, and the single best way to avoid those lines is to get the eligible students registered up front, when they’re enrolling or registering for classes or applying for aid.

But while I think it’s sound policy, and applaud steps in this direction — like in Maine, Ohio, and Tennessee, where state law requires public high schools to facilitate voter registration for eligible students — it’s not currently a federal requirement.  Which may be why, despite Mitchell’s claim, ED didn’t actually discuss any such mandate in its guidance.

Here’s the guidance document, from April 21, 2022.  There’s no requirement in the document that supports Mitchell’s assertion.  Indeed, the only mandate in the letter is ED’s quoted reminder of a federal statute: postsecondary institutions in states subject to the NVRA are required to “make a good faith effort to distribute a mail voter registration form, requested and received from the State, to each student enrolled in a degree or certificate program and physically in attendance at the institution, and to make such forms widely available to students at the institution.”  That’s verbatim from 20 USC 1094(a)(23)(A), and it’s been federal law for 25 years now.  The guidance also notes that the institution can send an electronic registration form or a link to an electronic registration form instead of a paper form, which is just 20 USC 1094(a)(23)(D), with a 15-year history.

That’s it.  There’s no broader boogeyman in the mandate ED posted than a recitation of words in the U.S. Code for decades.  Just so we’re all working from the same set of facts.