The auditors in charge of overseeing the $787 billion economic stimulus package acknowledge there are problems with the first batch of spending data reported last month — but they say those problems were expected, and that the next batch of reports in January will be more accurate.

Recipients of stimulus funds — state and local governments, private contractors and nonprofits — reported on their spending last month. The Recovery Accountability and Transparency (RAT) board posted that data on Recovery.gov, the stimulus oversight Web site; it showed that roughly $150 billion in stimulus spending to date has created or saved 640,329 jobs.

But the data has been widely criticized since its release. It shows multimillion-dollar contracts that created zero jobs; stimulus spending performed in congressional districts that don't exist; and companies that reported "saving" more than half of their workforces. As many as one in 10 recipients didn't report anything at all.

Earl Devaney, chairman of the RAT board, acknowledges that the data was full of errors.

"What we are all seeing, at least following this first reporting period, is not particularly pretty," Devaney said at Thursday's hearing of the House Oversight and Government Reform Committee. "Everybody engaged in this is involved in a lessons-learned exercise. ... We have learned a lot from this first reporting period."

A Government Accountability Office report released last week showed that nearly 4,000 recipients reported creating or saving jobs — despite having received no stimulus money. More than 9,200 recipients reported the opposite: They said they received stimulus money, but didn't create any jobs.

"There was some erroneous or questionable information in the database that merits additional scrutiny," said Gene Dodaro, acting comptroller general. "There are also questions about the quality of review that was done by federal departments and agencies."

The board is working to develop better automated tools to check the data as it's reported, Devaney said. For instance, one tool might check to make sure reports are identified with valid congressional districts.

Agencies also are planning to update their policies for reporting data. The Transportation Department, for example, found that several states were improperly coding their stimulus spending, or submitting the wrong paperwork; the department is working to clarify its guidance. And the Education Department is planning another nationwide outreach program to tell recipients about problems with their first round of reporting.

"We'll develop a ‘lessons learned' document and begin another round of outreach in advance of the next reporting period," said Tony Miller, the department's deputy secretary.

Many agencies already have systems in place to confirm data. Education, for example, developed an automated program that analyzes recipient reports and looks for obvious outliers. The results are forwarded to the appropriate program offices, which then manually review them.

One-third of the recipient reports submitted to Education last quarter — 742 of 2,229 — were corrected through this process.

Dodaro said the Office of Management and Budget should also come up with a standard definition of "full-time equivalent." That's how recipients are required to report the number of jobs created or saved. But the definition of a full-time equivalent can vary state by state, which makes it impossible to compare data across the country.

Legislators pushed Devaney to punish the estimated 10 percent of recipients who received stimulus funds but failed to report data. Devaney said he'd also like to see penalties applied to recipients that don't report any spending data.

Asked whether he would impose penalties, though, he said that was up to Congress.

"I believe in penalties. … My law enforcement background leads me to believe that penalties have a deterrent effect," said the former Secret Service agent. "But Congress didn't put any penalties in."