Supreme Court Revives Controversial Citizenship Data System
The Instrument Without a Logbook
This policy is a hypothesis. The evidence so far suggests that it has been adopted with more confidence than its test warrants. When the Supreme Court of the United States revived the federal data system for checking the citizenship of registered voters, it did not settle whether noncitizens are voting in numbers that matter. That question remains disputed. What the Court did was allow an instrument built on that disputed premise to begin operating across the United States. A genuine experimentalist does not ask first whether the ruling was correct. He asks what we will be able to learn from it, and whether anyone has arranged to learn anything at all.
The theoretical problem is the one the lawyers argued: may a state compare its voter rolls against federal citizenship records? Put that way, the answer seems obvious. Only citizens may vote. The government holds records of who is a citizen. Checking one against the other looks like diligence. But the problem people actually experience is different. It is not whether the state may check. It is whether the check can tell a citizen from a noncitizen reliably enough that the cure does less harm than the disease it was prescribed for, a disease whose existence is itself in dispute.
The strongest case for the system deserves to be stated plainly, because it is not foolish. Registration has long rested largely on the applicant’s own word, sworn under penalty. A skeptic may fairly say that an oath is a weak filter, that matching against official records replaces trust with verification, and that a Court declining to block a program is exercising restraint, not endorsing a policy. Let the states try it, the argument runs, and let experience judge.
I accept that argument entirely, and that is exactly where the difficulty begins. Experience judges only when it is recorded, compared, and made public. A citizenship database is a photograph of a person’s legal status taken at some moment in the past. Naturalization changes status, and records follow the change at the pace of bureaucracy, not of life. So the system’s characteristic error is predictable in kind, even where no one yet knows its size: it will flag people who were once noncitizens and are now citizens. Picture a woman who raised her hand in a federal courtroom, took the oath, and posed for a photograph with a small flag and her grandchildren. Some months later, a letter arrives from the county asking her to prove what that courtroom already established. She takes a morning off work, finds the certificate in the drawer where she keeps her mother’s rings, and stands in line. She is cleared. Nothing about her morning enters any account of whether the program works.
This is the defect that matters. The program’s errors are experienced one household at a time, at a clerk’s counter, in a letter, in a lost wage, while its successes are announced in aggregate at press conferences. An experiment that counts its confirmations publicly and lets its refutations dissolve privately is not an experiment. It is a slot machine that reports only its jackpots. Every flag that turns out to be a citizen is data. Every flagged registrant who never answers the letter and quietly drops off the roll is also data, perhaps the most important data, and it is precisely the kind no one is obliged to collect.
A community that treats its policies as hypotheses would build the logbook alongside the instrument. It would require states using the system to publish how many registrants were flagged, how many were confirmed as noncitizens, how many proved citizenship, and how many were removed without resolution. It would ask the federal agencies holding the records how quickly naturalizations reach the database. None of this presumes the program is bad. It presumes only that we intend to find out, which is the minimum a democracy owes to any measure that touches the franchise, the one instrument by which the public corrects its other mistakes.
The Court has shown it will let the experiment proceed. It has not required, and perhaps cannot require, that anyone watch the results. That task falls to legislatures, election officials, and the public, and it will not perform itself. Until someone does it, the argument over noncitizen voting will be conducted exactly as before, by assertion, and the only people gathering real evidence will be citizens like that woman, standing in line with a certificate in her hand, running the experiment on themselves while no one takes notes.