The Mismatch Between Good Statistical Data and Governmental Policy Decisions. (May 14, 2012)

Director of the United States Bureau of Census for the past three years, Robert Groves, is to step down to return to academia as provost of Georgetown University in Washington D.C. According to Science's Jeffrey Mervis: "[Dr Groves] was so successful that Democratic and Republican lawmakers alike expressed chagrin when Groves announced last week that he would be leaving the agency in August."

Here we reprint the final two questions and answers of the interview published in Science because they are pertinent to the announcement by the Australian government of the deep cuts to the budget allocations for the Australian Bureau of Statistics.

Science: Why do you want to be a provost?

Robert Groves: One thing I didn't fully understand before I took this job [a director of the Census Bureau] is the gigantic mismatch between the production of really good statistical information about what is going on and how decisions are made. For example, I'm much more aware of that mismatch with regard to decisions that are made affecting the U.S. economy and society.

Science: Was that a disappointment to you?

Robert Groves: Absolutely. I've devoted my life to trying to improve the quality of statistical information, and it is disappointing to see how little of that is used to make decisions. The second observation is how fundamentally data have changed. The Internet has reshuffled the deck of knowledge [to the point where] with a couple of keystrokes, we can become privy to what all disciplines know about any problem.
      So I concluded that the most important thing we have to do is to prepare universities for that new world. That means restructuring curricula and research activities to take advantage of both opportunities, using knowledge for action and utilizing interdisciplinary perspectives. That's the kind of challenge that motivated me.

But what Dr Groves does not address is how policy makers through the remainder of this decade can be persuaded to overcome the gigantic mismatch between the production of really good statistical information about what is going on and how decisions are made.