Charges by John Crudele in the New York Post that the U.S. Census Bureau fudged the monthly unemployment numbers a couple of months before the 2012 election are not persuasive.
First, according to the Census’ statement today, Julius Buckmon, the individual quoted in the New York Post story “left” the Census Bureau payroll (on what terms we don’t know) in 2011; so he hardly was in position to fix the numbers the following year in a September, 2012 monthly report.
Second, there is no corroborating evidence of statistical skullduggery. Reading between the lines, there would seem to be a job performance issue here that is personal, not political.
Third, the ability of one minor functionary to change the numbers is very small. Mr. Buckmon didn’t have that big a role in his regional office. For example, he didn’t consolidate the major numbers reported to the Bureau of Labor Statistics.
Fourth, as AEI’s Pethokoukis points out, the Philadelphia region where Mr. Buckmon supposedly skewed his numbers (with directions from someone else), actually reported an increase in unemployment that month. If Buckmon was finding employed people who didn’t exist and reporting them, why did unemployment go up, rather than down, in the affected region?
Fifth, the unemployment reduction reflected in the September, 2012 household survey is in sync with the Bureau of Labor Statistics’ trend data in the establishment survey that provides an alternative measure of employment.
Sixth, those who say that the White House must have been responsible, since it brought the Census under its direct supervision in 2009, dropped a page out of their contemporary history book. The new Obama Administration tried to take the Census under its wing, but failed after a public outcry by, among others, myself, as a former Director.
Seventh, the Census Bureau, whatever its failings, is justifiably jealous of its reputation for integrity and is ardent about deserving and protecting that reputation. Repeatedly one sees the statistical community rise up to fight any political intrusion. Congress and the President may pass laws describing what kind of data the Bureau collects and how it collects it, but the process of collecting and reporting the resulting data is not easily manipulated. In this case, for example, wrongdoing would be caught in the various checks laid down by other Census personnel who care a lot more about the accuracy of what they do than about the image of politicians. The conspiracy would have to be pretty deep and far reaching–and indifferent to the long standing culture of the Bureau–to succeed.
Eighth, the inspector generals of both the Labor Department and the Commerce Department (where the Census Bureau resides) have these charges before them and will report. You also can be assured that the House Oversight Committee for the Census (Chaired by Rep. Darrell Issa) are looking at this closely. I suspect that they are likely to find that there are far more profitable subjects to investigate. Like, say, Obamacare.