Senate to Census Bureau: Survey This!

As reported in the Washington Post a couple of weeks ago, the US Senate is threatening to enact severe cuts in the Census Bureau's budget for next year. The Post's editorial page reports today that the nascent American Community Survey would likely have to be abandoned if these cuts go through. Compared to the "long form" survey used in the 2000 decennial Census, the ACS is a shorter, monthly survey designed to provide the same information in a less painful way. The Post article explains pretty succintly why the ACS is a good thing:

The centerpiece of the reengineered 2010 census is the ongoing monthly American Community Survey, on which the bureau has spent $700 million. It asks the same questions as the long form, but by spreading questionnaires through the year, census officials say they can better handle complaints that some questions are an invasion of privacy and take too long to answer... For localities, the survey's selling point is that it produces updated data each year, not once a decade as the census does.

So the ACS gives a more continuous flow of information over time, and is (at least in theory) less of a drag for people to respond to than the traditional "long form" survey, for reasons which should be obvious from the name. To hear folks at the Census tell it, the demise of the ACS would be doubly bad for US taxpayers because (despite the short-run savings) in the long run, eliminating the ACS and returning to the old "long form" decennial Census would actually end up costing more, even though it provides less useful information.

Losing Census data probably doesn't sound like a big deal for most people. When the Census announced, after years of delay, that it was going to stop publishing its State and Local Government Finances data annually (going to an every-two-years schedule), you couldn't fill a room with the US taxpayers who ever noticed. But the idea of losing the ACS--to say nothing of the apparently "floated" idea of canceling even the old long-form Census in 2010 to save money-- should be a big deal for good-government advocates working on all sorts of issues. As D'Vera Cohn put it in the Post, the threatened cuts

would mean no long-form data about commuting times, housing costs, immigration patterns, marital trends, income inequality or other topics that help shape public policy.

Anyone who's seen the way tax policy is made knows the importance of an informed decision-making process. Without good data on social, economic and demographic trends, lawmakers can't even diagnose emerging policy failures-- let alone diagnose correct public policy remedies. The threat posed by these cuts is real enough that the Brookings Institution held a panel discussion on the implications for public policy research. (A transcript is available here.)

It may seem almost flippant to say this when Congress is considering cuts in the Food Stamp program, but the potential information loss from eviscerating Census funding is huge. Congress needs to know that Americans value an informed policy-making process-- and that we know the Census makes this process possible.