Business Employment Dynamics data for Q2 2022 came out today. This data is based on the QCEW. The Philly fed spoiled the headline that there was negative net job change in the quarter on a SA basis. The BED reported a net job change of -287k vs +1 million www.bls.gov/news.release/cewbd.nr0.htm

for the same quarter in the CES. There are caveats in interpreting 1 quarter divergences between QCEW/BED and CES, which I discuss in the linked thread. The BED does give some context on what is driving the divergence. The BED links establishments

over time in the QCEW, which means it can distinguish between job changes at existing establishment and changes at entering/exiting establishments. The -287k net change can be divided between a +70k at existing and -357k net change from entering/exiting establishments. This is

noteworthy because the CES only directly measures changes at existing establishments. Changes from entering/exiting are estimated. Net birth/death estimate adds +617 of the +2702 NSA change in CES employment. Assuming a similar % contribution to the 1 million SA change would be

about 200k. Replacing that with net change due to entering/exiting firms from the BED could explain about 500-600k of the 1.3 million divergence. However, it still leaves more than half the divergence unexplained with the CES still showing relatively strong growth in existing

establishments compared to the BED. In terms of the bigger picture, I would expect estimated birth/death vs actual birth/death to diverge at turning points in the labor market. Due to the lumpiness of the QCEW, I would want to see another quarter to see if this is a trend or

a one off. It's also interesting that job growth from entering firms doesn't seem to be the problem in Q2 and we'll want to see if entry holds up. Finally, note that the Q2 QCEW data won't show up in the CES benchmark revision that will be reflected in the next release.