Employment levels and cumulative change: QCEW stories

The quarterly census of employment and wages was released today. Recalling that some observers argued that the recession occurred in H1 2022 (and/or Q2 2022) as the household survey and QCEW data leveled off. Taking into account the revised data for the first and second quarters and the new data for the third quarter, we have the following picture of the levels of employment in the non-agricultural sector and employment in the private sector.

Picture 1: Nonfarm Employment as of January 2023, CES release including revised benchmarks (blue), ADP (brown), QCEW total workers covered, seasonally adjusted using log-transformed X-13 census (pink), using multiplicative moving average (sky blue), preliminary Philadelphia Fed benchmark data (red squares), all in 000 s.s. Light blue shading denotes a hypothetical 2022H1 recession. Source: BLS (various) and ADP via FRED, BLS QCEW, Philadelphia Fed via FRED and author’s calculations.

Figure 2: Employment in the private non-farm sector as of January 2023, CES release including revised benchmarks (blue), ADP (brown), QCEW of private workers covered, seasonally adjusted using log-transformed X-13 census (pink), using multiplicative moving average (sky blue), Philadelphia Federal Reserve preliminary benchmark minus government employment data (red squares), all in 000 s.c. Blue shading denotes a hypothetical recession in the first half of 2022. Source: BLS (various) and ADP via FRED, BLS QCEW, Philadelphia Fed via FRED and author’s calculations.

Now, since the coverage varies, it is difficult to know if the number of jobs actually increased or not during the 2022H1 period. Hence, I am showing the cumulative change since December 2021 (in 000s).

Figure 3: Change since December 2021 in non-farm payroll employment compared to the January 2023 edition of CES including revised benchmarks (blue), ADP (brown), total number of QCEW workers covered, seasonally adjusted using log-transformed census X-13 (pink) using a multiplicative moving average (sky blue). ), Philadelphia Fed Preliminary Benchmark (red squares), all in 000 s.e. Light gray shading indicates a hypothetical 2022H1 recession. Source: BLS (various) and ADP via FRED, BLS QCEW, Philadelphia Fed via FRED and author’s calculations.

Figure 4: Change since December 2021 in private non-farm payroll employment compared to the January 2023 CES release including revision of benchmarks (blue), ADP (brown), QCEW of private workers covered, seasonally adjusted using log-transformed census X-13 (pink) using multiplicative moving average (sky blue), Philadelphia Fed preliminary benchmark minus government employment data (red squares), all in thousands, s.c. Source: BLS (various) and ADP via FRED, BLS QCEW, Philadelphia Fed via FRED and author’s calculations.

The astute observer will see that every row in fig. 3 higher at the end of the first half of 2022 than at the end of the second half of 2021. Each series, save Philadelphia Fed’s early landmarkhigher in 2022M06 than in 2022M03.

In Figure 4 (private non-farm employment), again each row is higher at the end of the first half of 2022 than at the end of the second half of 2021. Two series—the Philadelphia Fed’s early benchmark minus measured government employment and QCEW, seasonally adjusted using a moving average—show slight declines from March to June. But all other ranks (including the ADP ranks, which are based on actual payroll data and therefore do not rely on CES or CPS) have risen.

In terms of future trends, every single row is higher in 2022M09 than in 2022M03.

So, when evaluating this January comment :

The first concerns the reliability of the sources. What should we believe? CES or HH survey? Menzie competed for CES. In the first half of the year, this seemed to be somewhat problematic as we saw a decline in productivity and a fall in GDP. If we were creating so many full-time jobs, why did productivity fall and GDP fall?

The HH (CPS) poll, in contrast, found that 1) employment has not changed since March and 2) that more than 100% of the job gains came from part-time or multiple jobs, suggesting a surge in low-wage work. which negatively affects productivity. It seemed to me more plausible.

At the same time, I considered it possible that both surveys were in fact correct, but skewed by the effect of suppression recovery, which created a misleading impression because we misinterpreted the data. It still seems possible, although I’ve read that others think CES was manipulated to create a rosier picture ahead of the election. In any case, if one believes that both surveys can be correct in some way, perhaps this discrepancy can be eliminated by several workers. However, as it turns out, a few jobs make up only 314,000 of the 2.7 million jobs at CES in March-November. period, period, I think we discussed. So that assumption turned out to be wrong, as Menzi pointed out, and I admitted it.

We then learned that CES was fundamentally wrong as the Fed cut job gains from 1.1 million to 10,500 from March to June. For that period, this made the whole issue of reconciliation moot, as it claimed that CES was creating phantom jobs. Reconciliation was not required.

Thus, the household survey appears to be the most reliable source, especially during June and probably for much of the rest of the year. And this suggests that the growth in employment has come solely from part-time jobs and a few employees.

We can say that almost everything stated is wrong. good not full time. good KPS vs. CES. When manipulating with CES, there is No evidence (unless you found it on Italian satellites changing the vote count). The fact that lower wages lead to a significant decrease in productivity is so wrong that I can’t even say anything. About the comment: “We then learned that CES was fundamentally wrong as the Fed cut job gains from 1.1 million to 10,500 from March to June.” I will notice that it was Federal Reserve Bank of Philadelphia researchers it reduced the estimated number of jobs, not the Fed as an institution.