A Philadelphia Federal Reserve report in December detailed how the Bureau of Labor Statistics’ (BLS) Current Employment Statistics (CES) survey had overestimated total U.S. employment in a period in 2021 by about a million jobs.

Instead of a growth in a million jobs, the BLS’ Quarterly Census of Employment and Wages (QCEW) survey showed a less than stellar growth in employment—only about 10,000 new jobs.

The discrepancy was noted by critics of Biden, accusing the administration of manipulating statistics to make it seem like their policies were more successful than they were.

The BLS has since stood by their research, stating that the Philadelphia Federal Reserve report was less accurate, that the CES data is the best source of employment data, CES data is regularly benchmarked with the QCEW data, error rates are consistently low (below .05 percent), and if anything their estimate was too low rather than too high.

If anything, that time period may be the most accurate jobs estimate in history.

As it stands the post-pandemic era has shown the smallest gap between the two surveys going back to 2001 when the QCEW was introduced. October of 2021 would have the lowest gap on record: 434,820 jobs.

Survey Differences

The QCEW survey has always underestimated employment compared to the CES. Additionally, the QCEW survey shows month to month seasonal volatility in its estimates that don’t appear in the CES. The CES also rounds its estimates to the nearest thousand—something QCEW does not.

So there is good reason that the two estimates would differ. Their closeness in the 2021-2022 timeframe appears to indicate that other sources of error have been minimized.

Out of concern that more frequent adjustments would lead to more error, the BLS recently abandoned its efforts to update the CES more frequently, to have more accurate data available sooner.

Subscribe now

Currently the total employment, by CES and QCEW data, for June of 2022 is almost exactly back to where it was pre-pandemic in February of 2022 despite COVID deaths topping over 1.1 million.

For CES, its incredibly close—within 601,000 of pre-pandemic values, or less than one-tenth of a standard deviation. Median growth in employment in both surveys is approximately 2 million a year.

Investigative Economics

"Inflated" Employment Estimates Were Actually…