Now that Pres. Trump has fired the person in charge of the BLS employment report, Axios wrote an article that outlines the general BLS process to the jobs release. Here is the down and dirty without getting too much into the weeds:

  • Release timing: The Bureau of Labor Statistics (BLS), a nonpartisan agency under the Labor Department, releases the monthly jobs report at 8:30 a.m. ET on the first Friday of each month.

  • Initial snapshot: The report includes the monthly change in employment based on surveys from ~629,000 worksites. It's a quick but lower-resolution estimate, but it is not without its quirks and influences that can lead to revisions.

So what are some of the reasons for the BLS to revise the Jobs Data:

  • Delayed responses: Some businesses don’t submit their employment data in time for the initial release, so the BLS updates figures over the next two months as more information comes in.

  • Initial data is a snapshot: The first report offers a quick but lower-resolution view of job trends; later revisions provide a clearer, more complete picture.

  • Impact of large employers: If major companies are late in reporting, it can skew initial figures, leading to significant revisions.

  • Seasonal adjustments: The BLS recalculates seasonal factors (like holidays, weather, or school schedules) to better isolate real employment shifts.

  • Population shifts: Changing labor force dynamics—such as those influenced by immigration and deportation policies—can distort the data and require recalibration.

  • Annual revisions: Each year, the BLS adjusts job numbers using finalized tax records, often resulting in major corrections (e.g., 818,000 fewer jobs were added than originally reported in the year through March 2024).

Revisions are routine, not political: These updates have occurred since 1979 and are part of a methodological, nonpartisan process aimed at improving data accuracy.

That said, the view we had yesterday isn't the same as today. The data was wrong.

The last two months of payroll data were sharply revised—May was cut from 144K to just 19K, and June was slashed from 147K to 14K. Those revisions dragged the three-month average down to just 35K.

On Wednesday, Fed Chair Powell said the Fed was in a good position to wait, with the benefit of seeing two more jobs reports and two more inflation reports. Ironically, the two “more informed” employment reports came not from the future—but from revisions to the past. In effect, they got three jobs reports: May, June, and July.

Put another way: if the next two reports come in at 35K and 35K, the Fed would likely cut rates (or been very tempted to) in September. So, you could argue they would have eased in July had they known what we know now.

Having said that is the changes seen today in the numbers because of manipulation at the BLS? Or is it more of a flaw in the data methodology collection itself?

I tend to trust that the revisions aren’t subjective or manipulated—they’re driven by math and better information which is from the data itself. But it still raises the question: is there a better way to collect and report the data?

Interestingly, the ADP private payrolls which is calculated from actual data from the ADP employment payroll business over the same three-month span told a similar story to the revised BLS story. The ADP employment report showed job gains/losses of:

  • May: +37K

  • June: –33K

  • July: +104

That implies a three-month average of 36K.

Sound familiar?

Despite different methodologies of data collection, both are now painting the same picture and the data, converged to the ADP data, not the BLS data.

My inkling is that if this is not a political witch hunt to manipulate data in Pres. Trumps favor, that if change is at hand, the shift should be in the data collection. That might also require the help of those that have actual payroll data as opposed to surveying companies/people who may or may not respond in a timely way, in an accurate way, or even at all.

Time will tell but 36K is close to 35K and that makes sense.

Source: Forex Live