By Paul Vigna

For months now, the data on the U.S. economy has displayed an odd schism. Most of the data has been pointing to a weak economy – except for the jobs numbers, which have consistently pointed toward a strong economy. Most observers, from Street denizens to the punditry to the practitioners of the dismal sciences, have written off downbeat data as somehow misleading, and expect those figures to eventually jibe with the stronger jobs numbers.

Thursday’s jobless claims provided the latest instance of a strong labor market. The four-week moving average of weekly jobless claims – a measure that smooths out some of the volatility of the weekly numbers and provides a clearer picture – fell to 274,000, a 15-year low. Claims have not been this low since April 2000. In terms of the other major sounding – the monthly jobs report – the unemployment rate is so quickly approaching what usually constitutes “full employment” that some are suggesting the threshold of “full employment” should be lowered.

But what if it’s the jobs numbers that are misleading?

“Falling claims have been sending a false economic signal,” SouthBay Economics’ Andrew Zatlin wrote in a report. “Jobless claims are falling for non-economic reasons.” One big reason he points to is that many states tightened the rules and time frames for unemployment benefits, so people who previously could or were claiming benefits were dropped off the roles. This has had a marked impact on continuing claims.

Beyond that, many states simply aren’t reporting the full data, Mesirow Financial economist Diane Swonk said this morning on the MoneyBeat show. Ms. Swonk said she was in Washington last week talking to the stats people who put together the claims data, and was told that the states aren’t submitting complete reports, for various reasons. “So it’s not clear how good that data is anymore, and that’s worrisome as well.”

Read More Why the Good Jobs Numbers Are Misleading – MoneyBeat – WSJ.