The Obama administration recently gave what it says is a more accurate account of how many jobs were created and saved by the federal stimulus bill. The new guess is that an estimated 640,000 jobs were created or saved by the federal spending.
Milwaukee Journal Sentinel reporter Ben Poston took a look for himself to see if the government’s numbers would hold up. When it comes to the state of Wisconsin, the government’s report was full of problems.
Poston discovered that some jobs were counted twice because federal and state money was involved. He found that some jobs were listed as “saved” just because an employee got a less than 2 percent cost-of-living raise. In other cases, the data was just flat wrong.
I interviewed Poston about his story to find out what journalists need to know to tackle this story in their states. You can read his edited responses below.
Al Tompkins: Walk us through the process you used. How did you assemble and examine the data?
Ben Poston: This was the easy part. I visited the Recovery.gov Web site and downloaded national summary data on contracts, grants and loans from the site’s download center and saved them as Microsoft Excel files. I sorted out all records related to Wisconsin recipients and began analyzing the data to look for projects generating the most jobs.
As I reviewed the data, I began finding stimulus-funded projects that had received no funding, but showed dozens of jobs had been created. I knew I had the makings of a decent story. The data work was straightforward, but the reporting was crystallized by making calls to recipients and agency heads to nail down the inconsistencies and get them to go on the record.
How did the federal government count some data twice?
Poston: I found a $7.3 million sewer replacement project in rural Douglas County that was showing 50 jobs created or saved in the loan data. But the sewer district treasurer in the town of Parkside admitted that he meant to enter “5″ into the Department of Agriculture’s online reporting system, but added an extra zero. He tried to correct the mistake but was told it was too late.
Those 50 jobs showed up twice — once in the loan records and again in the grant data — because the project was a combined loan and grant. So the five jobs saved by the project (which won’t begin until May) were reported as 100. I found three other cases where jobs were counted twice.
So a job can be counted as “saved” if you just award a 1.8 percent cost of living raise?
Poston: Apparently so. There were problems with Head Start grants in most states. Recipients of the grants in Wisconsin reported nearly 160 jobs saved by giving those employees 1.8 percent cost-of-living raises. I think there was general confusion among the recipients about the guidelines on reporting full-time equivalents. A spokesman for Health and Human Services told me the agency was working to correct the numbers.
Is it even possible to accurately count the indirect jobs that get created?
Poston: That’s a good question. When direct job recipients spends money they would not have earned without the stimulus, they create new economic activity that generates more employment, hence the indirect job. The White House says there is a multiplier effect of one indirect job for every direct job created.
The White House reported that 640,329 direct jobs were created or saved by the stimulus through September, but that figure balloons to 1.3 million if you include indirect and induced jobs. Those jobs include workers who make the steel that contractors use to complete their project, or the waiter who serves construction workers at a nearby restaurant.
In the end, you wind up arguing with a formula, since the government does not try to actually measure the jobs.
So, after all you have done, do you have any idea how many jobs were created and preserved in Wisconsin?
Poston: Not really. All I can say for sure is that the total is less than 10,073, the figure first reported. From what I found, it’s at least overstated by hundreds of jobs. Depending on who you talk to, some people would say it’s exaggerated by the thousands because more than half of the jobs in Wisconsin were comprised of teachers, law enforcement or local government employees. It’s unlikely those school districts, municipalities and counties would have imposed all the necessary cuts solely through layoffs.
What did you learn in covering this story that would help other journalists cover the story in their state/community?
Poston: Interview the data like you would interview a source. Look carefully at the grants, contracts and loan data for any discrepancies. If the information seems irregular or improbable, it probably is.
Make calls to the agencies who administer the program or to the recipient who reported the data for explanations. In this first round of stimulus reporting, the data was riddled with errors because this is new territory for private contractors and all levels of government. Most recipients I spoke with were open to discussing the reporting process and any issues they had with it.