The Bureau of Labor Statistics is struggling to collect data on inflation and the job market.

It’s remarkable that a basic function of government could still be underfunded in a government that spends $7 trillion per year. Yet the Bureau of Labor Statistics (BLS) is struggling to accurately collect data on inflation and the job market in part because of underfunding.

The Economist reports that lower response rates are making it harder for the BLS to collect data for the consumer price index and the jobs report. The response rate for the employment survey fell from 93 percent to 83 percent between 2000 and 2019. After Covid, the response rate fell to 67 percent.

This long-running problem has been on Congress’s radar for years. I talked with former BLS commissioner Bill Beach, a Trump appointee, about it on my podcast last year, and he said he has raised the concern with members of Congress personally. He said the costs to update the survey methodology and get the accuracy back to where it should be are in the tens of millions of dollars, chump change for the federal government.

Yet BLS funding has fallen, declining by 18 percent in real terms over the past 20 years, The Economist reports. Then, the Trump administration’s hiring freeze and removal of probationary employees hit the BLS especially hard, since many of the workers who collect data in the field were probationary or short-term. The agency is now at roughly 60 percent capacity.

Without that data collection, BLS statisticians have to impute data, i.e., guess about the prices of more categories of goods than before. Some imputation is normal, just because information isn’t available. In the past, about 8 percent of prices were imputed. Now, that rate has soared to 30 percent. The BLS is guessing on nearly one-third of the goods and services included in the consumer price index. It’s very good at guessing, and the data are still of high quality. But they aren’t as good as they could be, and they’re going in the wrong direction.

Again, this is a problem that could be solved with tens of millions of dollars out of $7 trillion in spending. But this is what happens when entitlements and interest payments crowd out everything else.

The bulk of the government’s exorbitant spending goes to just a handful of things: major health care programs, Social Security, and interest payments. The military comes next, then comes everything else that the federal government does. That last category includes a lot of the basic functions of government, such as the courts, law enforcement, tax collection, and diplomacy. It also includes the government’s statistical agencies, such as the BLS.

Collection of official statistics is a vital role of government. There are plenty of private data collection efforts, and government agencies often work with private firms to get their data. But government’s being a focal point for data collection and dissemination is vital to a well-functioning economy.

As Kent Lassman of the Competitive Enterprise Institute — no friend of big government! — has said before, economic data collection is a proper role for government because “the use case for much of this information is in public policy and not routinely commercialized.” That means the market will underprovide it, and policymakers will be in the dark.

It’s very hard to do the work of shrinking government or repealing bad policies without quality information on what the effects of government are. The cost of supporting the data collection efforts that contribute to government statistical reports is minimal compared with the cost of policymakers’ not knowing the relevant information.

The federal statistical system is the gold standard for government data collection around the world. Preserving that should be a priority for Congress. If the federal government is going to spend $7 trillion a year, that should at least buy reliable official statistics.