Political Questions About the Jobs Report

With the economy dominating the campaign for the White House last fall, did the government cook the books to make the unemployment rate move lower and help President Obama’s re-election chances? Was manipulation responsible for lowering the unemployment rate to 7.8 percent in September 2012 from 8.1 percent in August, a big deal when the numbers came out in October 2012, just weeks before the election?

That’s the contention of an article Tuesday in The New York Post by John Crudele headlined “Census ‘faked’ 2012 election jobs report.” And it seems to confirm the suspicions of Jack Welch, the former G.E. chief executive who said on Twitter at the time, “these Chicago guys will do anything … can’t debate so change numbers.”

Mr. Crudele’s article begins with an employee of the Census Bureau, which collects the data for the unemployment rate report, who says he was told by higher-ups to “fabricate” results of the surveys he was supposed to be doing.

The Labor Department had targets for how many households the Census Bureau needed to poll each month, and when it fell short in the New York and Philadelphia regions, “Philadelphia filled the gap with fake interviews,” the Post article states. To make matters worse, “a knowledgeable source says the deception went beyond that one employee â€" that it escalated at the time President Obama was seeking re-election in 2012 and continues today.”

Those are the allegations reported by Mr. Crudele â€" the only problem is that a lot of them don’t add up.

Before the heavy lifting of analyzing how the data is prepared and crunching the numbers, consider the perspective of Keith Hall.

Mr. Hall was commissioner of the Bureau of Labor Statistics from 2008 to 2012 under both President George W. Bush and President Obama, and was appointed to the job by President Bush. Before that, he was the chief economist for the White House Council of Economic Advisers, again under President Bush, from 2005 to 2008. He is now a senior research fellow at the free-market-oriented Mercatus Center of George Mason University.

In other words, he doesn’t have a dog in the hunt politically. “I’m skeptical,” he said. “It sounds like a workplace performance issue. This is somebody being lazy, or a supervisor really cutting corners. It’s certainly not evidence of an attempt to move the numbers.”

So let’s look at the facts in question. For starters, the former Census Bureau employee named by Mr. Crudele in the article Tuesday hadn’t worked for the agency since August 2011, more than a year before the election, according to the Census Bureau.

But what if it continued after he left, with pressure from supervisors in the Philadelphia office? Well, as James Pethokoukis of the American Enterprise Institute notes, the jobless rate for Pennsylvania actually rose in September 2012.

In fact, as Mr. Pethokoukis points out, the big drops in unemployment that month were in states nowhere near the supposedly tainted Northeast - instead happening in California, Utah, Louisiana, Florida and South Carolina.

The methodology of the Census Bureau’s survey also makes the manipulation argument seem far-fetched at best. For starters, the household survey is based on 54,000 household interviews per month. The typical field representative handles about 37 cases a month, although that varies by office.

So if a few employees in one office fudged the numbers one month, it would certainly be troubling and cause for a top-to-bottom investigation, but it would not be enough to alter the nationwide figure by much. In addition, in each 15-month period, the Census Bureau goes back two or three times and re-interviews a handful of the original subjects to make sure the results were gathered correctly.

“Making up entire caseloads would be caught,” said a veteran Census Bureau field manager on condition of anonymity, because she wasn’t authorized to speak publicly. In terms of the former employee, she said, “No matter how hard he tried, he couldn’t move the number.”

In addition, while the manipulation is alleged to have taken place in the household survey that produces the unemployment rate, subsequent revisions of the establishment survey showed parallel increases in job creation by employers.

When the numbers were revised on Feb. 1, 2013, the B.L.S. concluded that the first estimate of job creation in September undercounted activity by 6,000 positions, while the figures for the next three months were revised upward by 150,000 jobs from the initial count.

Indeed, while three-tenths of a percentage point is a big drop for one month in the rate, it was followed by steady downward movement in unemployment levels, or what you’d expect if the numbers were right in the first place, as Joe Weisenthal of Business Insider writes.

“It’s not completely implausible that someone would make up a report and the Census Bureau tries to prevent this from happening,” said Katharine G. Abraham, who was commissioner of the B.L.S. in the Clinton administration, and was a member of the Council of Economic Advisers under President Obama. “But it’s completely implausible it could move the unemployment rate by .3 percentage point.”

Mr. Crudele did not respond to requests for comment Tuesday, nor did a representative.

CNBC, meanwhile, put the story into heavy rotation, and Jeff Cox, a finance editor at CNBC, concluded that “Jack Welch may not have been so off the mark after all.”

“Should the controversy go full-blown,” Mr. Cox warned ominously, “it could shake the government’s data-gathering operations at their foundation.” Rush Limbaugh said he is waiting for an apology from the “Drive-By Media” for mocking him when he echoed Mr. Welch more than a year ago.

Before the shaking and apologizing begin, remember there have been even more outlandish accusations of B.L.S. tampering, which turned out to be baseless. More than four decades ago, President Richard M. Nixon ordered aides to count the number of Jews at the agency, because he suspected them of liberal tendencies and manipulating the figures to hurt him politically.

For now, the Census Bureau’s parent, the Commerce Department, has referred the matter to its internal inspector general. “We have no reason to believe that there was a systematic manipulation of the data described in media reports,” the bureau said in a statement. In addition, Bloomberg reports that the Labor Department’s inspector general is reviewing the information, and the House Oversight and Government Reform committee is launching an inquiry, too.

Whatever those inquiries find to have actually happened, it is unlikely to have moved the needle, said Mr. Hall, the former B.L.S. commissioner.

“They ought to look at it,” he said. “But it’s pretty unlikely it would affect the overall number.”