Accurate statistics are part of the foundation of any modern society.
This may sound like an exaggeration, but it's absolutely true. Policymakers and businesses alike have come to rely heavily on accurate statistics, as they allow leaders to make informed decisions.
That's why this news, via Jared Bernstein, is so infuriating. Faced with $22 million in budget cuts, the Bureau of Labor Statistics is scaling back a couple of programs to keep its core functions operating. Here's what's getting cut:
Curtail the Quarterly Census of Employment and Wages (QCEW). The QCEW program provides national, state, metropolitan statistical area, and county data on monthly employment and quarterly total wages and the number of establishments...The BLS will achieve savings largely by reducing the scope and frequency of collection for select units in the QCEW survey...
Curtail the International Price Program (IPP). IPP Export Price Indexes measure the price change of goods and services sold to foreign buyers. The BLS will discontinue production and publication of its Export Price Indexes. [BLS]
Bernstein, formerly the chief economic adviser to Vice President Joe Biden, comments:
The QCEW provides critical benchmarking data to the monthly payroll survey so an even slightly "degraded" sampling frame will potentially reduce the accuracy of the payroll survey as well. And in a world were inputs are increasingly globalized/outsourced, we need more, not less information about import pricing...
$22 million is 0.0006 percent of current outlays, so…um…it ain't exactly the problem. The whole damn agency budget, as per above, amounts to just 0.02 percent of 2013 outlays ($3.45 trillion). [On the Economy]
First of all, budget cutting right now is utterly pointless — in fact, it is counterproductive. A time of labor market slack is not the time for austerity, and there are strong arguments that budget-cutting will actually increase the deficit in the long run. We should be increasing spending and cutting taxes. And, as Bernstein points out, the amounts here are just pitifully small. Our national debt is roughly $17.5 trillion. Is another $20 million going to tip us into crisis? Would we even notice if someone hadn't pointed it out? Of course not.
Beyond that, the BLS is just about the worst place to be looking to save a few measly pennies. Having high-quality information about what is happening in the nation is absolutely critical for policymakers to make informed choices. How would the Federal Reserve set monetary policy, for example, if they weren't sure what is happening to unemployment and prices?
And it's not just policymakers. Private companies depend on the same thing. Some time ago I attended a meeting in the basement of the Capitol about efforts to preserve the BLS and other government statistics efforts, like the Census' American Community Survey. Paul Emrath, a lobbyist from the National Association of Home Builders, said without government statistics the entire business model of the construction industry would be threatened. Without good population statistics, putting up a new apartment building suddenly becomes much more risky.
Free government statistics are the dictionary definition of a public good. They are highly valuable to society as a whole, but very hard to make a profit from. Absent government action, they'll be systematically under-produced. A private version of the ACS would be proprietary, very expensive, and not nearly as comprehensive as the federal databases, Emrath said.
Sadly, as we've learned over the past few years, having a government that provides public goods competently — indeed, the very idea of public goods, or impartial statistics — is sharply politically contested. Luckily, these particular cuts are quite small, but the trend is ominous: Ultraconservative Republicans in the House have repeatedly tried in the past to cripple the Census Bureau on ideological grounds. At some point, Republicans will probably win power in Washington again, and unless something changes, American businesses had better be ready to start counting.