We Need a Better Way to Collect Jobs Data

The Labor Department revised previous jobs numbers so dramatically last week that President Trump said they had to have been "rigged." (He even fired the head of the Bureau of Labor Statistics.) Skeptics reject the idea of rigged numbers, but even some in this camp say the controversy should lead to improvements in how the Labor Department collects its data, a problem unrelated to politics. Details:

• How it works: The Bureau of Labor Statistics "surveys 121,000 employers—businesses and government agencies—each month, seeking their total payroll employment during the week in which the 12th day of the month falls," explains Reuters. Typically, they get about 60% of these voluntary responses in time for the initial estimate, but responses are allowed to arrive over two months. Thus, each month's jobs report also has revisions for the two months prior.
• Door to door: BLS also collects data through "old-fashioned door knocking," reports CNN. Agency employees visit homes across the US to collect info on residents' employment status. All the collected data is extrapolated to represent the entire nation, a process that takes some "educated guesswork" involving factors such as seasonal hiring trends, the story notes.

• New system? Usually, the revisions are minor, but large revisions such as last week's are possible, and they undermine confidence in the system when they occur, Bank of America CEO Brian Moynihan said on CBS over the weekend, per Reuters. "They can get this data, I think, other ways, and I think that's where the focus ought to be: how do we get the data to be more resilient and more predictable and more understandable?"
• Ditto: Kevin Hassett, Trump's National Economic Council director, also thinks the system needs modernization, and called for an examination of the process back in 2015, reports the Wall Street Journal. At the time, he said on Fox over the weekend, he was "worried about how the changing economy ... was going to move faster than their ability to measure." Big revisions of late prove the point that "the numbers are not keeping up with the economy."