Miscounting of stimulus jobs cuts two ways

Spending reports indicate that contract and grant recipients under- and over-counted jobs created and saved.

Confusion over Office of Management and Budget stimulus reporting guidelines has led to under-counting -- in addition to already reported over-counting -- of jobs saved or created by the Recovery Act, according to an analysis by Government Executive.

OMB instructed prime recipients of stimulus funds to provide the public with dozens of data elements on spending, including a brief description of the types of jobs that they and any subcontractors had saved or created. In a separate column, recipients were to enter the total number of jobs saved or created -- a figure that in theory should match the information in the narrative column.

But, a closer look at recipient data released on Oct. 30 on Recovery.gov showed that information in the two columns often did not line up. Repeatedly, recipients of grants and contracts discussed jobs created or saved in the narrative column but failed to include them in the total.

For example, Native American Services Corp., a small business in Idaho, reported that an Army parking lot repair contract helped it retain five jobs. In the next column, however, the company said it had not saved any jobs. Similarly, Eagle Aviation Services and Technology of Reston, Va., said a $70,000 contract with the Army Corps of Engineers allowed it to retain eight workers who otherwise would have been laid off. But, the company put a zero in its jobs saved column.

And, it's not just small businesses that appear to be confused. URS Corp., one of the largest government contractors, reported it created 18 jobs by hiring a small disadvantaged business to assist with a liquid waste project. But URS did not record those workers in the jobs figure column.

Other recipients reported that they were "not sure" about their jobs total. One acknowledged "confusion on submission requirements. Not sure this is correct."

The discrepancies were even more pronounced with grantees. California's Transportation Department filed 684 reports, of which 531 had a zero entered for total jobs created or saved. Nonetheless, the department consistently reported in the narrative column that, "Jobs are created or retained in the construction and construction management industry such as laborers, equipment operators, electricians, project managers, support staff, inspectors, engineers, etc." More than 100 of the projects with no jobs in the totals column are under way or completed. The California Transportation Department did not return a call for comment.

OMB, which updated its recipient guidance repeatedly during the first half of 2009, said some of the discrepancies between the two columns can be attributed to the conservative jobs counting formula the Obama administration provided.

"It may be that funding recipients, after providing the anecdotal information, erred on the side of caution by not including data in the formula column," OMB spokesman Tom Gavin said.

The reporting errors go in both directions. Media reports have found that some recipients overstated the total number of jobs saved or created. Some mistakenly based job retention figures on the number of employees given a cost-of-living increase, rather than the number in danger of getting laid off. Others made typographical errors in reporting their jobs totals.

Because the extent of the under- and over-counting in the 130,000 recipient reports released on Oct. 30 has not been measured, it is unclear how these errors will affect the administration's overall claim the Recovery Act saved or created 640,000 jobs.

"To date any errors we have found do not impact meaningfully on the total of 640,000 direct jobs created -- and indeed, errors we have found of over-counted jobs tend to be balanced by under-counted jobs," Gavin said. Watchdog groups said flaws likely are widespread. "These are not isolated problems," said Phil Mattera, research director for Good Jobs First, a Washington-based nonprofit that is tracking the stimulus. "These are conceptual problems."

Michael Balsam, chief solutions officer for Onvia, a Seattle firm that created the stimulus-tracking site Recovery.org, said he has met with Recovery Act recipients who have significantly different interpretations of the jobs reporting rules.

"You knew that there would be issues, so it's not a big surprise that the reporting was problematic," he said.

Good Jobs First found that 31,000 grant and contract recipients placed a zero in the column for number of jobs created or retained. Nearly 1,200 of those recipients described their projects as more than 50 percent complete and another 1,270 characterized their projects as finished.

There could be legitimate reasons an ongoing or completed project has not resulted in any jobs saved or created, Mattera noted. For instance, a company could decide to use its existing workforce. But the more likely scenario is that recipients misunderstood the reporting requirements, he said.

The Economic Policy Institute, a Washington think tank, argued in a recent report that, "It is clear from the recipient reports that either the jobs-reporting instructions are unclear, that recipients are ignoring the guidance, or both." The group recommended that the administration find a way to automatically screen recipient reports for obvious errors and to flag reports containing mistakes or misinterpretations of jobs instructions.

Ed Pound, a spokesman for the Recovery Accountability and Transparency Board, which runs Recovery.gov, conceded many recipients had difficulty in reporting their job creation numbers. But he said it's not the board's role to fix the data so it reflects the recipient's intent.

Gavin said the administration is examining reports of over- and under-counting and will continue to ask recipients questions and fix errors. "All of this will be completely transparent," he said. "The public will see preliminary data, revised data and updated data, as it is posted."

Making matters more complicated, it also is not clear whether all recipients filed spending reports. "We are doing an analysis to determine how many recipients did not report," Pound said. "We do know that some did not report, but we do not have a precise number at this time."

If the administration does not clarify its reporting instructions, then these problems are only going to be compounded as time goes on," Mattera cautioned.