Early Benchmark suggests a downturn, as does QCEW, Powell conjecture:
Figure 1: Official nonfarm payroll (NFP) employment (blue) ,implied preliminary benchmark (tan), Powell conjecture applied to implied preliminary benchmark (sky blue), early benchmark (green), CPS adjusted to NFP concept (red), and QCEW total covered seasonally adjusted by author using X-13/log transform (purple), in 000’s. Source: BLS via FRED, BLS, BLS, BLS, Philadelphia Fedand author’s calculations.
Implied benchmark calculated by wedging in the preliminary benchmark between April 2024 and March 2025, and using reported changes thereafter. The early benchmark is calculated by applying the ratio of sum of early benchmark state employment to sum of CES state employment to official BLS NFP employment, and reported changes thereafter. The Powell conjecture applies the 60K overcount estimate cited by Jay Powell to April-November NFP changes.
QCEW total covered series is seasonally adjusted in log terms using Census X-13, with X-11 ARIMA default.
