The three major federal statistical agencies — the Bureau of Labor Statistics (BLS), the Bureau of Economic Analysis (BEA), and the Census Bureau — have access to information about US businesses. But complicated rules govern how the Internal Revenue Service (IRS) can share the information it gathers as part of its mandate to collect taxes with those three statistical agencies. And information the IRS has would greatly assist the statistical agencies in understanding what’s happening with US businesses, and the economy overall.
What’s needed is data synchronization — a limited form of data sharing. More specifically, there is a growing need for the statistical agencies to synchronize data that would allow them to accurately assign businesses to their proper industries (e.g., accurately classifying Target as a business in the retail industry, or Filomena as a business in the restaurant industry). Census and BLS each have their own information about businesses, but the information produced by one agency often contradicts the other — according to a 2006 study, one-third of businesses are in industry X according to Census, but in industry Y according to BLS. (In other words, the same business is assigned to different NAICS industry codes by each agency.) These discrepancies could be fixed if Census and BLS could take a look at each other’s industry information and harmonize industry assignment data across the two agencies. But Census gets its business information in part from IRS data, which BLS is not legally allowed to access — therefore, Census and BLS can’t compare notes about which firms belong to which industries.
Search This Blog
Thursday, August 4, 2016
Problems with Federal Data
At AEI, Michael R. Strain describes a little-understood problem with federal data: