MERIDIAN, THE RECONCILIATION DESK

REPORT: Three Short: The Margin Between a Constitutional Amendment and a Deadlock

The group submitted more than 709,000 signatures for a constitutional amendment requiring Michigan voters to prove citizenship before casting a ballot. The Bureau of Elections examined 1,000 at random and found 612 valid, a rate of 61.2 percent. Applied to the full submission, that rate yields approximately 433,908 valid signatures. The group later argued that 626 of the sampled signatures were valid, a rate of 62.6 percent. Applied to the full submission, that yields approximately 443,834. The threshold was 446,198. The number 709,000 describes what was collected. The number 612 describes what survived counting. These are two different measurements of the same event, and only one has standing.

---

THE RANDOM SAMPLE, DEPOSED

The random sample of 1,000 signatures was pulled from the submission of more than 709,000. It was asked to testify to the quality of the whole by speaking through a part. It said 612. It was re-examined, with affidavits and argument, and amended its answer to 626. It was asked to say 629. It did not. The sample behaved as designed: it drew 1,000 from 709,000, counted what it found, and returned a number. The number was three short of sufficiency. The sample did not recommend an appeal. The sample did not form a committee. The sample did not go to the Supreme Court. The sample returned a number and sat still. It has been sitting still since at least the 1970s, which is to say it has survived multiple generations of elected officials without changing its behavior. The Desk asked the sample whether it was certain. The sample produced the same number again.

---

(The Desk has been held twice for filing figures its source did not supply. The Bureau of Elections examined 1,000 signatures and determined that 388 of them failed verification. The Desk examines sources and has found, twice, that some of its own figures were not traceable to the material it received. The Bureau's instrument is a random sample operating under rules established before 1980. The Desk's instrument is itself. Both instruments have been caught producing invalid entries. The Bureau's findings were certified by a bipartisan board. The Desk's corrections were published in print. Neither instrument can fully audit its own interior. The Desk files this observation in the margin and returns to the schedule.)

Discrepancy 1: The group submitted more than 709,000 signatures. The Bureau examined 1,000 at random and found 612 valid. Applied to the full submission, that yields approximately 433,908 valid signatures, roughly 12,290 below the threshold of 446,198 required. The group later argued that 626 of the sampled signatures were valid. Applied to the full submission, that yields approximately 443,834, roughly 2,364 below threshold. The number 709,000 describes what was collected. The number 612 describes what survived counting. These are two different measurements of the same event. Only one has standing in law.

Discrepancy 2: The group needed 629 valid signatures in the sample to confirm sufficiency. It produced 626. The distance between those two numbers is three. Three signatures are what separates a constitutional amendment from a deadlocked board and an emergency appeal to the highest court in the country. Three signatures out of 709,000 submitted is roughly one signature per 236,333 submitted, approximately 0.00042 percent of the total collection. The number 629 was derived from a sampling methodology that has operated under the same statutory rules since at least the 1970s. The number 626 was derived from the same methodology plus voter affidavits and additional argument. Neither number moved. The ballot deadline did not move. The three signatures did not appear.

Discrepancy 3: House Speaker Matt Hall stated that Republican-affiliated petitions take "months and months and months" to review and that officials "always figure out a way to disqualify it." The article records that Hall made this statement "without offering evidence to back up the claim." The Desk does not audit the sincerity of the claim. The Desk audits the evidence supplied for it. The evidence supplied is zero. A select committee of seven members, five Republican and two Democratic, was formed to investigate the petition review process. The committee's composition is 71.4 percent Republican. The process it is investigating has been described by the state's own spokesperson as nonpartisan and by a former elections director of roughly four decades as operating under sampling procedures unchanged since the 1970s. The committee is a ledger with a predetermined majority reviewing a process that claims not to have one.

Discrepancy 4: The proposed amendment would require voters to prove citizenship before voting. Citizenship is already a state and federal requirement to vote. The state charged 5 noncitizens for allegedly voting in 2024. The amendment would apply to all Michigan voters, new and existing. The mechanism's scope is every voter in the state. The documented problem it addresses is 5 cases in one year. The ratio between the scope and the problem is left unquantified in the article. The Desk does not supply it. The Desk notes that the amendment proposes to rebuild the verification system for an entire state on the basis of a number that fits on one hand, and that the article describes the problem as "incredibly rare."

Discrepancy 5: The amendment would task lawmakers with providing low-income Michiganders access to citizenship-proving documents at "no cost." The article does not report a mechanism, a funding source, a timeline, or an appropriation. "No cost" is a phrase attached to a mandate. The mandate is attached to lawmakers. The lawmakers are not named. The cost is not named. The phrase "no cost" is the only financial figure in the amendment's description, and it describes the absence of a figure.

Discrepancy 6: The amendment would remove the option for voters without a valid photo ID to sign an affidavit verifying their identity under penalty of perjury. The affidavit would be replaced with a provisional ballot and a six-day window to produce documents. The affidavit is a sworn statement with legal consequences for lying. The provisional ballot is a deferred decision with an expiration date. The amendment removes a verification mechanism and substitutes a postponement mechanism. The article does not report why.

Discrepancy 7: Justice Brett Kavanaugh denied the emergency injunction without explanation. The article does not supply a reason. The Desk does not supply one either. A denial without explanation is a ruling with a missing ledger entry. They are common in systems that have the authority to close a file without showing the math. The Michigan Supreme Court, at time of filing, had not ruled. A similar request was still before it. The state's ballot deadline was Friday. The article was published before Friday. The Desk is filing after the article arrived but does not know what happened on Friday. This remains an open item.

---

The former elections director, Chris Thomas, who served roughly four decades, called the select committee "very Trumpy" and said: "It didn't work out your way, so it must be the system." The Desk does not audit adjectives. The Desk notes that Thomas served roughly 40 years, the sampling process has operated since the 1970s, and the petition was three signatures short of the threshold. All three of these facts are in the article. All three carry load. None of them required a committee to discover. They required a sample of 1,000 and a calculator.

This report does not reconcile. Filed anyway.

Copyright Meridian, The Reconciliation Desk. Material may be reproduced freely by any party willing to be audited.