Illustration: Rebecca Zisser/Axios
Can we trust this morning's surprisingly good employment report?
- The short answer: Yes.
The longer answer: No statistics are entirely reliable, and the more of an outlier you're trying to measure, the less reliable your statistics become.
The big picture: No one at the Bureau of Labor Statistics is trying to commit fraud. The agency was doing its very best to be accurate in October 2012, when employment figures made Obama look good, and it's doing its very best to be accurate in June 2020, when the same report makes Trump look better than many expected.
- What they're saying: It's possible that "the models used to produce these numbers — they aren't really raw data — have gone haywire in a time of pandemic," says Nobel economics laureate Paul Krugman.
By the numbers: The official unemployment rate includes millions of Americans recorded as employed but absent from work due to "other reasons." Most economists agree those people should have been included as being unemployed on a temporary layoff, but for consistency's sake they were not.
- The household survey, which measures unemployment, only managed to reach 67% of the people it tried to poll. That's down 15 percentage points from normal. Still, its results are in line with the results of the establishment survey, which measures employment, and whose response rate held up much better.
The bottom line: All macroeconomic statistics should be taken with a large grain of salt right now. That said, this morning's report is undeniably good news.