A pay freeze is one of the cleanest instruments in budgeting. It is a single decision, it takes effect on a known date, and its saving can be written down before anyone has argued about it. That is why governments reach for it, and it is why the number that gets quoted is always the one that is easy to calculate.
The alternative pay plan issued in late August holds most federal civilian salaries at 2026 levels through 2027. It does not hold everyone. Federal law enforcement officers are exempted and receive 3.8 percent. Military pay rises between five and seven percent depending on rank.
Those two exceptions are the most informative part of the document.
The exemptions concede the argument
To exempt a category from a pay freeze is to say, in the only language a budget has, that in this case pay determines whether the roles get filled. Nobody carves out an exception for work they believe people will do regardless. The carve-outs identify precisely where the government privately expects a wage cut to cost it staff.
The difficulty is that the same logic applies to the roles that were not carved out, and the evidence for it is already on the record. The Professional Aviation Safety Specialists told reporters the FAA's plan to hire 750 new technicians becomes much more difficult under a freeze. Those technicians are not a rounding error in a safety system; they are the safety system. The National Active and Retired Federal Employees Association described the freeze as effectively a pay cut once inflation and expected health premium increases are counted, which is arithmetic rather than rhetoric.
This publication has been describing the receiving end of this for months. Agencies cannot hire fast enough to spend what Congress appropriated: the money is voted, the posts are authorised, and the shortfall shows up as a vacancy rate rather than as a failed programme. The controller training pipeline is the same story with a longer lead time, where the constraint is not the budget line but the years required to produce a qualified person.
A freeze does not reduce headcount. That is what makes it attractive and what makes it hard to observe. It reduces the government's position in a bidding contest for specific skills — engineers, inspectors, clinicians, technicians — against employers who are not frozen. The result is not a layoff. It is a slower fill, then an unfilled post, then a function that quietly runs at reduced capacity, arriving two and three years after the decision that caused it.
Scale sharpens it. The non-postal federal civilian workforce fell below two million in 2026, the first time since the 1960s, on the Office of Personnel Management's June figures. A smaller workforce administering an unchanged body of law has less slack, and less slack means each unfilled specialist post costs more than the one before it.
There is a version of this decision that would be defensible on its own terms: freeze pay, and simultaneously publish which functions are expected to degrade and by how much. The saving is knowable in advance. The cost is knowable too, and the government has already demonstrated it can identify the exposed roles — it did so when it wrote the exemptions. It has simply declined to extend the exercise to everybody else, which leaves a policy whose price arrives on a timetable nobody has been asked to sign for.
The same gap runs through the fiscal picture more broadly, where a debt path built on assumptions about a workforce is easier to publish than the workforce itself. A budget can be balanced on paper by a decision whose cost lands somewhere that is not a budget line. This is one of those decisions.
The terms of the alternative pay plan, the 3.8 percent law enforcement figure, the five to seven percent military range and the OPM headcount are as reported by Government Executive and Federal News Network on 31 August 2026. Statements are attributed to Everett Kelley of the American Federation of Government Employees, William Shackelford of the National Active and Retired Federal Employees Association and Dave Spero of the Professional Aviation Safety Specialists as quoted in that reporting. The analysis is our own.




