How does Moody's Analytics estimate employment for New England MSAs using the NECTA data reported in the BLS Current Employment Survey? How do annual benchmark revisions affect our estimation schedule?
Update Process for New England MSA Employment Data
In its Current Employment Statistics (CES) the U.S. Bureau of Labor Statistics (BLS) publishes employment data for most of the metropolitan areas (MSAs) and metropolitan divisions in the U.S.—but not for New England. Instead, the BLS has retained the New England City and Town Area (NECTA) definitions in Connecticut, Maine, Massachusetts, New Hampshire, Rhode Island and Vermont.
For consistency across the nation, Moody's Analytics estimates employment for the New England MSAs. We combine employment data from two sources: county-level from the Quarterly Census of Employment and Wages (QCEW), plus NECTA-level from the CES.
- Estimate a full county-level employment dataset using available QCEW data.
- Estimate state-level values for New England, using CES and QCEW.
- "Squeeze" the counties to the states, i.e., adjust the values of constituent counties so they sum to the state's value.
- Aggregate constituent counties into New England MSAs.
Effect of Benchmark Revisions on Update Schedule
The CES releases benchmark revisions each year with the January payroll employment data. For non-New England MSAs, we republish the CES results directly, so the annual revisions are visible immediately. In parallel, the CES revisions trigger revisions to our system of county estimates, and only when complete can we aggregate to the New England MSAs.
Hence, you can expect the employment series for New England MSAs (for example, ET.MBOS) to be delayed relative to MSAs everywhere else (e.g., ET.MHOU).
The CES is a sample survey, and its results are estimates. To have validity, estimates need to be calibrated against more authoritative data, ideally a census; this is called benchmarking. Benchmarking can inform future results but can also retroactively amend past results; this is a benchmark revision.
The results of the BLS Current Employment Statistics program are published in phases—national, then state and metro (it publishes no county-level data). Similarly, its annual benchmark revisions are staggered: early February and mid-March, respectively. The revisions are published at the same time as the new January results, and are based on updated population figures.
The QCEW publishes down to the county level and, because it's a census, has no need for benchmark revisions.