The Statistics Blackout: What Six Missing Weeks Did to the Data
Federal statistical agencies restarted after the shutdown with gaps that cannot be filled retroactively, leaving policymakers and investors to make decisions on a patchwork of private data.
The federal government's statistical agencies came back to work this month facing a problem that no amount of overtime can solve: some data simply cannot be collected after the fact.
The 43-day shutdown halted the surveys that produce the monthly jobs report, the consumer price index and much of the input to quarterly GDP. Household surveys ask about a specific reference week. Price collectors record what an item cost on a particular day. When those windows close, they close.
What was lost and what was merely delayed
Establishment data — payroll counts drawn from employer records — can largely be reconstructed, because the underlying records still exist. That is why the payroll figures came back relatively intact, if late.
Household survey data is different. The unemployment rate comes from interviews with about 60,000 households, and an interview conducted in December cannot tell you who was looking for work in October. Price data sits somewhere in between: some prices can be recovered from receipts and online archives, many cannot.
The consequences ripple forward
Seasonal adjustment is the quiet casualty. The models that convert raw counts into the smoothed figures markets react to are estimated on long, unbroken histories. A hole in the series distorts those adjustments for months in both directions, which means several 2026 releases will carry larger-than-usual revisions.
Filling the gap
Private data providers had their most consequential six weeks ever. Payroll processors, card networks, job-listing aggregators and price-scraping firms became the primary evidence for what the economy was doing. Their coverage is real but unrepresentative in known ways — skewed toward large employers, toward card users, toward online prices.
The Federal Reserve met in December with this patchwork in hand and cut rates a quarter point over three dissents. Officials were explicit that they were operating with less information than usual.
The institutional question
The episode revived a long-running argument about whether the statistical agencies should be insulated from appropriations lapses the way some other functions are. Data collection is inexpensive relative to almost anything else in the federal budget, and the cost of losing it is borne by every business and household that relies on markets to price risk correctly.
