Jobs Truthers Are Wrong: Unemployment Numbers Can’t Be Cooked
The unambiguously positive jobs report, which showed 114,000 new jobs created in September and the unemployment rate dropping to 7.8 percent from last month’s 8.1 percent, caused some predictable reactions. Elation on the left, sober assessment from the center. And conspiracies from the right. Jack Welch, Congressman Allen West, Joe Scarborough, a collection of Fox News hosts, and a motley crew on Twitter all seemed to conclude there was something fishy about the report: the fix is in.
Jack Welch, the former General Electric CEO, was perhaps first out the gate, even speculating about a possible conspiracy theory last night, when he tweeted “Tomorrow employment numbers for Sept. with all the assumptions Labor Dept. can make..wonder about participation assumption??” And when the news came in, it was certain: “Unbelievable jobs numbers..these Chicago guys will do anything.. can’t debate so change numbers.” That’s pretty rich, considering Welch’s GE was known for producing quarterly earnings results that were just right.
Is there any reason to believe Welch and his fellow brave truth tellers? No.
Betsey Stevenson, the former chief economist in the Labor Department, wrote this morning, “Anyone who thinks that political folks can manipulate the unemployment data are completely ignorant about how BLS works and how data are compiled.” Alan Krueger, the chairman of the White House Council of Economic Advisers and one of the nation’s preeminent labor economists, told Bloomberg today, “No serious person would question the integrity of the Bureau of Labor Statistics. These numbers are put together by career employees. They use the same process every month. So I think comments like that are irresponsible.”
In fact, the Bureau of Labor Statistics, the division of the Labor Department responsible for the monthly job report, is incredibly scrupulous in how it compiles and presents the unemployment data. The bureau’s acting head, Jack Galvin, has been at the BLS since 1979 and is a career nerd who ran its employment data collection and analysis from 1998 to 2011. The data, which come from surveys of households and employers, arrives at the BLS on the Wednesday of the week before the report is released.
During the eight days they assemble the data, the BLS team uses encrypted computers and locks the information in a safe whenever they leave the conference room where the entire team is working. Some senior members of the president’s economic team get a peek at the data, but only on the afternoon before it is released Friday morning.
The conspiracy theorists might sound less loony if the data in this report were particularly surprising. But they’re not. While the decline in unemployment is a solid 0.3 percent and takes the rate below the psychological landmark of 8 percent, it is not necessarily inconsistent with past data showing consistent, but not particularly high, job growth. The drop in the rate isn’t unprecedented either; in fact, there’s something about September, maybe seasonal hiring, that brings up the jobs numbers.
For the numbers to be manipulated, career officials at the BLS—who are not Obama appointees and are part of a department with an overwhelming ethos of being nonpolitical—would have to manipulate two surveys. They’d have to manipulate the “establishment survey,” which comes from employers and provides the payroll-jobs figure of 114,000 posts added. Then they’d have to manipulate the survey data from households, which showed a 873,000 rise in total employment. The last time the BLS was accused of fudging the numbers in a political context was when Richard Nixon thought a “Jewish cabal” of Labor Department civil servants were out to make him look bad. Really.
Most importantly, even if the administration were somehow manipulating the data, then they’re doing a terrible job at it. Only with this month’s data has the unemployment rate dropped below the level it was at when Obama was inaugurated. And it wouldn’t really be all that surprising to anyone if the rate went back up to 8 percent next month.
Finally, look at the revisions. Just last week, the BLS released a preliminary revision to a year’s worth of jobs data showing that the employment level was actually 386,000 higher than they previously estimated. I imagine the campaign team in Chicago would rather have had that 386,000 spread out over 12 months of jobs reports.
The BLS also always goes back and revises the monthly reports. The best example is August 2011. The initial report said the economy had added zero jobs. That number was revised up 57,000, and then with a second revision, the final number was 104,000 jobs created. But did that change the initial headlines about job growth foundering or the mistake the widespread impression that the economic recovery had skidded out? The answer to that question is the same as the answer to whether the Labor Department is cooking the books.