DHS voter investigations should account for the cost of a mistaken knock.

A questionable record in a government database is an administrative problem. A knock at someone’s door can become that person’s problem. The distance between those two moments is where public accountability belongs.

The October 21 order in California Nurses Association v. Mullin limits specified DHS election field operations involving likely U.S. citizens in nine states. It is a preliminary ruling, not a final decision on every allegation. Its start is delayed until 5 p.m. on October 23, with conditions and the possibility of further court action. The government retains authority to investigate unlawful voting. None of those qualifications makes the underlying question small.

My argument is straightforward: the agency initiating an encounter should carry the burden of explaining why that encounter was justified. A resident should not become an unpaid quality control department for a government dataset.

There is a reasonable counterargument. Investigators cannot wait for certainty before asking questions. Some investigations exist precisely because the available records disagree. A rule that forbids every inquiry until a case is already proved would make the inquiry pointless. That objection deserves a hearing rather than a slogan.

But an investigative lead and a decision to use an intrusive method are different judgments. Before officials turn uncertainty into a field visit, they should be able to explain what less intrusive verification was possible, who approved the escalation, and how an error will be corrected. Those are standards for competent administration. They do not require pretending that an investigation can never be legitimate.

The person on the receiving end also has costs that do not appear on a department’s activity report. Imagine having to interrupt work to locate records, explain an encounter to family, or untangle an inaccurate entry. These are possible burdens, not accounts of people interviewed for this column. Counting an agency’s completed investigations while ignoring the work it transfers to residents produces a flattering and incomplete picture of efficiency.

The remedy should extend beyond stopping a particular visit. In my view, a serious accountability process would ask whether the underlying record needs correction, whether that correction reaches other systems, and whether the affected person can learn what happened without navigating an expensive procedural maze. A quiet database repair may protect a citizen more effectively than an impressive press conference.

The measure of a successful correction should be whether it reaches the person affected. A department can close a file while leaving someone else to live with its consequences. That is why process needs an owner, a clear explanation, and a way to verify that a mistake is actually resolved.

A preliminary injunction cannot supply the entire administrative reform program. It can force attention onto a basic allocation of responsibility. If the state chooses the data, designs the process, and sends the officials, the state should have to explain its mistakes. Public confidence should be measured partly by how well lawful participation survives the investigation, not merely by how forcefully an investigation is announced.