MSNBC’s Dubious Insinuation of Job Data Manipulation

A paragraph in an article written by Mike Stuckey on MSNBC.com insinuates that the White House manipulated the Recovery.gov job count total [1] to match its previous claims of job growth numbers. I can't tell if Stuckey simply has his facts wrong, is intending to mislead to create controversy, or has been misled by an unscrupulous source, but, wow, this paragraph is wrong:

When the government site [Recovery.gov] was saying last month that 30,000 jobs had been created, the White House was insisting the figure was much higher, although its estimate ranged from 600,000 to 1.5 million. The government site was subsequently updated to show 640,329 jobs created.

What this paragraph doesn't make very clear is that these three estimates are from different sources, and are trying to measure vastly different things.  But by obscuring the source and basis for each estimate, Stuckey makes it seem like the White House is purposely fixing the numbers.  Here's a fact-check for each of the three numbers.

• 30,000: This jobs figure is reported on the government site -- Recovery.gov [2] -- and is the number of jobs created directly by Recovery Act contracts (for instance, when the Department of Energy contracts directly with a company).  The number was based on some 9,000 recipient reports covering $16.6 billion in Recover Act contracts, which came out on Oct. 15.  The number is so low because it does not cover the other $194 billion reported in the first reporting cycle, which is in the form of grants and loans.
• 600,000 to 1.5 million: This estimate is likely an expansion of the number the White House was touting, which comes from a Council of Economic Advisors report [3] that counts jobs based on economic models.  The CEA report is different from Recovery.gov in a couple ways.  First, it takes into account the jobs created by the enhanced buying power of Recovery Act recipients, or so-called "ripple effects."  Second, included in the CEA job count estimates are jobs not just created directly by Recovery contracts, but by the $151.4 billion estimated to have been injected into the economy by the act -- an amount that includes some $63 billion in tax cuts and more than $30 billion in aid to individuals (like unemployment insurance and one-time $250 Social Security payments). Job estimates on tax cuts and aid to individuals are not reported on Recovery.gov.  Thanks largely to this greater pool of money, the CEA report estimated the Recovery Act had created or saved between "600,000 and 1.1 million" jobs this year. Finally, the CEA report also came out on Sept. 10, about six weeks before the initial release of recipient reports on Recovery.gov on Oct. 15.
• 640,329: Today, Recovery.gov indicates that 640,329 jobs have been created by the Act, because on Oct. 30, tens of thousands of additional recipient reports on another $194 billion in Recovery Act funds were published.  This simple fact explains the difference from the earlier 30,000 figure, and in that sense the 640,329 figure is basically an update to the earlier number.  These new reports are from recipients of grants and loans, and along with contract data, provide information on about $214 billion in Recovery Act awards. And again, these job counts are jobs created directly by contracts, grants, and loans. Unlike the CEA report, this count excludes jobs created by tax cuts and direct payments to individuals and jobs created indirectly from contract, grant, and loan expenditures.

Readers of this blog, I'm sure, are familiar with my criticisms of Recovery.gov (here [4], here [5], and here [6]), but one thing that I commend the Board on is its adherence to the principles of transparency.  The fiercely independent Recovery Accountability and Transparency Board, which runs Recovery.gov, is going to great lengths to ensure that the data reported by recipients is exactly what gets posted on the website.  There are perfectly good reasons for the different job estimates cited in the MSNBC article, and they are clearly not the result of any political influence from the White House.  It's too bad Stuckey didn't read the Vice President's Oct. 30 report [7] on Recovery Act jobs, which had a very helpful section entitled "How Do The Recipient Reports Compare With Other Job Estimates?" that compares the various job estimates, and explains where they all come from.  Reading this five-page document would have answered many questions for him, and might have stopped this article from being written.  The Recovery Act can be confusing,  but that didn't contribute to this train-wreck of an article.  It's simply a victim of poor reporting.