Treasury draft proposes auto-opening 63.36 million child Trump Accounts by Social Security number

Draft Treasury rules put a number on the childhood enrollment machine the agency wants to build: 63.36 million children who have Social Security numbers could have Trump Accounts opened automatically, without a parental election, and the agency projects more than 60 million accounts added this year. What the regulation does not say is how the government will know which children those 63.36 million are.

The draft replaces a March proposal with a simpler administrative posture: the Treasury Secretary may open the account on a child's behalf, and the legal requirement stops there. The enrollment problem becomes less a matter of getting a parent to sign a form or an employer to verify a payroll file and more a matter of matching a Social Security number against records that would prove a child belongs in the program—records the rules do not name.

An enrollment event in the tens of millions is not a pilot, and the Treasury's own projection of more than 60 million new accounts this year runs ahead of any manual repair the agency could reasonably perform. The rule text describes the authority to open the account but not the data field, agency file, or eligibility screen that would identify the child; the headline promise is automatic enrollment, and the drafting leaves the identification step as an implementation detail.

That missing detail is where the Saver's Match coverage problem re-enters, because the program's existing sign-up path has shown the familiar low-income participation gap and the draft's answer is to remove friction by opening accounts automatically. But removing friction does not remove the need to know who should receive one: if the identification step begins and ends with a Social Security number, the system inherits whatever gaps exist in the underlying records, and the regulation gives no indication of how it would handle a child whose name, date of birth, or guardian relationship is inconsistent across datasets.

The prior design placed the burden on employers to bring savers into the system, a bottleneck that plan-design data from Vanguard makes concrete: 26% of small plans use automatic enrollment, against 61% of large plans, and even at small plans that do auto-enroll, participation reaches 79%, still short of the 94% at large plans. The draft flips that burden from the employer to the Treasury, but it does not supply the birth-record infrastructure the flip implies, and the temporary regulations leave open what happens when the auto-generated account meets a child who has never appeared in an employer system at all.

There is another way to read the draft's silence: Treasury may already have a federal data source in mind and simply did not put it in the regulatory text. But the temporary regulation that will govern this year's projected adds is exactly the place a matching standard would live, because a rule that opens an account automatically has to define the population it can open them for. A Social Security number alone is a payment identifier, not the eligibility test the program's target population implies.

What the regulation does specify is that the Secretary may act without a parental election, which marks the substantive difference from the March proposal and puts Treasury in a position plan sponsors never occupied: the agency would originate the account and hold the enrollment record, and it would be responsible for knowing the account belongs to a real, eligible child. If the identity data is not described, the responsibility has no test attached.

That 15-point gap in Vanguard's plan-design data is the employer-side version of the coverage problem the Treasury is trying to bypass, and if the new draft flips the mechanism without naming the matching data, it may simply move the same gap from the plan sponsor to the master file. A 63.36 million account projection is too large for that gap to be hidden, because the first-year cohort is the program's entire automatic-enrollment experiment and there is no prior-year file to check against.

The 63.36 million figure deserves to be held against the program's purpose: if the goal is to reach the workers the subsidy targets, an auto-opened account for a child who already has a Social Security number is only as good as the underlying record's ability to find that child in the low-income population the program means to cover. The rule does not say whether the matching universe will be the Social Security Administration's enumeration files, state birth records, tax return dependents, or something else; it says only that the Secretary may open the account, with the child's Social Security number as the identifier the draft names. The middle step—the part that determines whether the subsidy lands where it was intended—is not in the text.

The March replacement matters because it shows the agency chose scale over consent: a parent no longer has to elect the account, and an employer no longer has to verify anything, but the government now has to know who is eligible. That is a data matching problem, not a distribution problem, and it is the one part of the draft left blank. The rule can auto-create accounts by the tens of millions; whether those accounts will sit in the names of the right children is the question the regulation does not answer.

What the first-year projection makes concrete is the failure mode: more than 60 million accounts added this year, or roughly 63.36 million children under the draft's stated authority, and if even a modest share of those accounts are opened for the wrong child, merged with the wrong Social Security record, or attached to an address that never receives the statement, the error is population-scale before the first annual statement goes out. The notice does not say how Treasury plans to avoid that, and the temporary rules will govern the rollout before the matching machinery is described.