Break Up the Data Nerds
What can be done about the perfectly predictable political capture of the Bureau of Labor Statistics
Last week, President Trump used that power to fire the BLS chief after an unfavorable jobs revision, replacing him with a loyalist. The problem isn’t just Trump; it’s the design. Today, Hamilton is paradoxically (but not so paradoxically) valorized by centrist technocrats as laying the blueprint for breaking through supposed sclerosis. Yet the Trump presidency reveals not just the shortsightedness of Hamilton but the prescience of the pseudonymous Anti-Federalists. At ground-level, this episode raises the question as to why the president has this authority, and what can be done to decentralize economic data-gathering.
Well, if we want to really dig into the roots, one can blame Alexander Hamilton, a Founding Father today revered by centrist technocrats for having set forth a vision for a “unified executive” — a premise not exactly aging like wine.
But more to the point, here’s an irony: BLS isn’t even a “unified executive” project. Most of its core numbers—employment counts, unemployment rates, wage data—already come from state workforce agencies, collected through long-standing cooperative agreements. The household survey that determines the official unemployment rate is run by the Census Bureau, not BLS field staff. Inflation numbers are gathered directly from businesses in cities nationwide. In other words, the “national” numbers already rest on a web of state-level inputs.
The vulnerability comes not from how the data is gathered but from how it’s compiled and released. Right now, a single presidential appointee sits atop the process, and that’s the point of leverage. If the commissioner is captured—whether through loyalty to the president or pressure from the administration—the release schedule can be slowed, the messaging reframed, or inconvenient results buried under technical noise.
The fix isn’t to blow up BLS. It’s to harden the system so no president or single appointee can choke off the flow of information. That means re-architecting the federal–state framework to make it more resilient and less centralized at the decision point.
First, keep BLS as the standard-setter and aggregator, but give the commissioner removal-for-cause protections and a fixed term that does not align neatly with the presidential election cycle. The Federal Reserve Board and certain regulatory commissions already operate on this principle.
Second, formalize the role of state agencies as independent data suppliers. Much of what becomes “BLS data” already originates in state workforce agencies—unemployment insurance claims, employer payroll reports, and price samples collected in cities and towns. Some of that information, like weekly unemployment claims, is already released directly by states before it appears in any BLS report. Other core measures, especially monthly jobs numbers from the Current Employment Statistics survey, are held under strict embargo until BLS’s coordinated national release—even though the states have the data in hand.
A reformed system would simply undermine the president’s ability to dictate the narrative around jobs. It would collect and disperse the data the way the U.S. is structured and represented, by states. A reform system would take a dual-track approach toe releasing data. States could publish their raw, anonymized data sets, including preliminary jobs numbers, directly to the public in real time while still sending them to BLS for aggregation and standardization. That could let analysts, journalists, businesses, and local policymakers work from the same underlying facts without waiting for a single federal agency’s blessing. It could also make it far harder for any BLS head—or the president—to suppress, delay, or spin politically inconvenient numbers, because the source material would already be in the open. Over time, such transparency could encourage methodological competition among states and reveal which approaches yield the clearest, most useful economic signals. This hybrid preserves the benefits of BLS’s national coordination—shared definitions, historical continuity, a single portal for aggregated data—while aiming to eliminate the single point of failure. In other words, much of the solution to combating Trump’s clearly autocratic impulses has less to do with what he is doing and more to do with explicitly disempowering the presidency going forward.
Critics will argue this adds complexity. It does. But complexity is a feature when it comes to safeguarding trust. A president shouldn’t simply install a loyalist and expect to control data from fifty state agencies, the Census Bureau, the Department of Labor, the Bureau of Economic Analysis, and the Occupational Safety and Health Administration. That’s just plain nuts.
Trump’s firing of the BLS commissioner is being treated as a partisan skirmish. In truth, it’s a stress test of an outdated design. The question isn’t whether the next president will be better behaved; it’s whether we’ll actually learn from the clear risks of concentrating more authority over information in a single office that can be reshaped at will. The problem is not the president. The problem is the power.
Please consider donating to Austin Disaster Relief Network to help Central Texans recover from recent and ongoing flooding. Donate here.
If you want to decentralize power, start by decapitating the democrat and republican parties. Those two parties have an obscene, and unconstitutional level of power.
No 'solution' to the concentration of power is going to come from within either party.