OIG: Phila. Census workers didn't manipulate jobless data

Workers in the Census Bureau's Philadelphia office didn't falsify data to manipulate the unemployment rate in the months leading up to the 2012 presidential election, a watchdog report has found

The Commerce Department's Office of Inspector General released a report this week saying that its investigation "found no evidence that management in the Philadelphia Regional Office instructed staff to falsify data at any time for any reason."

Last fall, news reports surfaced with allegations that Philadelphia-based supervisors told employees to fake data in order to boost employment numbers as the 2012 election approached. The Labor Department's monthly jobs reports are calculated based on Census surveys, and a former worker told the New York Post that workers had been directed to fabricate the numbers.

"OIG did not find any evidence to support allegations that supervisors in the Philadelphia Regional Office manipulated, or attempted to manipulate, the unemployment rate prior to the 2012 presidential election," the report states. "Nor did our investigation find any evidence to support that such manipulation is likely from a statistical perspective."

The report noted that it would have taken 78 field employees reporting every single unemployed person in their samples as "employed" or "not in labor force" in order to manipulate the jobless figure, "an effort which likely would have been detected." The report added that the unemployment rate continued to drop after the election, making the falsification of data even more unlikely.

However, some falsification has occurred: One Philadelphia worker was fired in August 2011 for faking data. In a Census Bureau analysis of some field employees' work, the bureau found that 78 of the 10,266 interviewers analyzed falsified data between October 2007 and September 2012, including 14 of 889 Philadelphia-area employees.

But the 14 falsifications in that entire time span "would be unable to substantively affect" unemployment numbers and other data, the report said.

The report recommended that the Census Bureau set up an independent system to check for manipulated data. Under its current system, the same supervisors who manage staff -- meaning they could direct subordinates to falsify data -- are the same people responsible for reporting staffers who fake data.

Additionally, the IG said the bureau should no longer let employees suspected of falsifying data continue to work while their surveys are being examined and should update training manuals to include information about detecting and handling instances of falsified data.