“Not a Scientific Survey…”

by Andrew Rudalevige on May 21, 2012 · 8 comments

in General Politics,Legislative Politics,Methodology

The House vote to de-fund the American Community Survey, the 3 million-person strong supplementary sample to the US census, has been noted on this site before but perhaps deserves to be highlighted. A useful NY Times primer from the weekend is here (with a h/t to Rob Mickey). The data are used for many purposes, not least for driving funding formulas for the distribution of federal money to states and congressional districts.

In the story, Rep. Daniel Webster (R-FL) complains that the survey “intrudes on people’s lives, just like the Environmental Protection Agency or the bank regulators.”  Leaving aside his casual equation of bankers to people (sorry! rimshot here), Monkey Cage readers might be most disturbed by Webster’s assertion that the ACS is an especial waste of money because “in the end this is not a scientific survey. It’s a random survey.”

The congressman has at least escaped the danger of being confused with his 19th century senatorial namesake. That Daniel Webster observed in 1825 that “Mind is the great lever of all things; human thought is the process by which human ends are ultimately answered.”

 

{ 8 comments }

Adam Thompson May 21, 2012 at 12:34 pm

Hmm, intrusion into everyday people’s lives? Attacking a wide ranging and important survey? Why, it’s a page right out of the Stephen Harper playbook when they got rid of the mandatory census! So, can we say that we Canadian took the lead for once in doing something stupid? I’ll answer my own question in the affirmative.

Boris Shor May 21, 2012 at 1:07 pm

Of course, but it must be said this is aggregious overclaiming from the NYT piece:

Maurine Haver, the chief executive and founder of Haver Analytics, a data analysis company. “The reason the Great Recession did not become another Great Depression is because of the more current economic data we have today that we didn’t have in the 1930s.”

Why, article author, did you need to include that?

Andrew Gelman May 21, 2012 at 2:02 pm

If only we had two American Community Surveys, we wouldn’t have had a recession at all. With three such surveys, the economy would be humming along just fine With four, it would be dangerously overheated.

Peter May 21, 2012 at 1:58 pm

As Nelson Polsby once said, members of Congress do not spend time reading studies, instead they just “throw them at each other like dead cats.” In other words, don’t expect to win them over with reasoned debate.

The Republican opposition to statistical sampling in the Census dates back over a decade and is based in the fact that the “actual enumeration” of the American people tends to undercount hard-to-reach Democratic voters, while statistical sampling will more accurately assess true numbers. Rather than trying to shame the proponents by pointing out they say silly things, the strategy here for opponents should be to make sure the Senate drops the language and to make sure this appears in the administration’s SAP against the bill. I suspect a quick analysis would show which states will lose money from relying only on enumerated census results to distribute funds and this will rally opposition to the provision very quickly.

Andrew Rudalevige May 21, 2012 at 2:32 pm

You’re right of course about the usual utility of studies in substantive legislative debate – though I would note that Jordan Tama’s recent book is a bit more hopeful, and that the ACS is not “that kind” of study, exactly. It is a data gathering exercise (going far beyond the enumerated census) without an advocacy argument behind it. Which brings us to your also-correct point that Republicans have been objecting to sampling for some years now. That said my point was not to shame Rep. Webster into changing his mind but to despair of the disconnect between his comment and pretty basic numeracy. And, I have to say I agree with Andy Gelman, below…

Andrew Gelman May 21, 2012 at 2:00 pm

There should be a law that politicians named Daniel Webster aren’t allowed to be dumb.

eric May 21, 2012 at 3:28 pm

I can’t help but think that this post and the previous post on this blog are related.

Debbie October 14, 2012 at 10:00 pm

Make it part of the regular every 10 year census and I will answer all the questions. Anytime between those ten years is a waste of taxpayer dollars and my time!!!

Comments on this entry are closed.

Previous post:

Next post: