The number of jobs gained in the United States is expected to be revised:

Goldman Sachs Group Inc. and Wells Fargo & Co. economists expect the government’s preliminary benchmark revisions on Wednesday to show payrolls growth in the year through March was at least 600,000 weaker than currently estimated — about 50,000 a month.

While JPMorgan Chase & Co. forecasters see a decline of about 360,000, Goldman Sachs indicates it could be as large as a million.

These are not just numbers; this data has implications for policies and economic conditions. Why are they being revised?

Once a year, the BLS benchmarks the March payrolls level to a more accurate but less timely data source called the Quarterly Census of Employment and Wages, which is based on state unemployment insurance tax records and covers nearly all US jobs. The release of the latest QCEW report in June already hinted at weaker payroll gains last year…

For most of the recent years, monthly payroll data have been stronger than the QCEW figures. Some economists attribute that in part to the so-called birth-death model — an adjustment the BLS makes to the data to account for the net number of businesses opening and closing, but that might be off in the post-pandemic world…

Ronnie Walker at Goldman Sachs says the QCEW figures are likely to overstate the moderation in employment growth because they will strip out up to half a million unauthorized immigrants that were included in the initial estimates.

In other words, this is a measurement issue. The first measure comes from a particular set of data and the revision utilizes a different set of data that takes more time to put together. There might also be discrepancies on what is included in each set of data, not just differences in the sources of data. This mismatch leads to later revisions.

Given our data and analysis abilities today, isn’t there some way to improve the system? Could we get (a) more complete data quicker or (b) better estimates in the first place or (c) even new data sources that could provide better information? To have an initial set of figures that people respond to and then a later set of figures that people respond to seems counterproductive given the stakes involved. .