Why The Business Dynamics data is the most authoritative source for job stats

Updated on

In the State of the Union address President Trump asserted, “Since the election, we have created 2.4 million new jobs….” This is incorrect, and by a wide margin.

In Q2 2017 alone there were 7.6 million new jobs. For a full year, it is more like 30 million.

Get The Timeless Reading eBook in PDF

Get the entire 10-part series on Timeless Reading in PDF. Save it to your desktop, read it on your tablet, or email to your colleagues.

The President (or his speechwriters) made the error through a common shortcut – discussing net job gains, a rather small part of the overall picture. Many media sources sloppily do the same. The table below shows the data for Q2 2017, the most recent available.


The Business Dynamics data is the most authoritative. It is drawn from actual state employment filings. No “fake jobs” there, unless businesses voluntarily pay taxes on phantom employees! The gross job gains are mostly offset by the 7.1 million in gross job losses. The result is an accurate count of net gains, reported quarterly and about seven months after the quarter ends. It is an important number, but almost universally ignored by the media and economic observers.

Essentially, it provides a good scorecard for Q2. At the time of the original BLS releases, the net gain in private jobs was reported as 428K. After the first revisions, which merely include more responses to the survey, the number was 554K. ADP, which does not try to match the first BLS estimates, choosing instead to align with the revised numbers, estimated net job growth of 588K.

Both were a bit high.

Why should we care?

Understanding the gross changes provides a much better picture of the dynamics of the labor market.

  1. A mere 2% increase in jobs created would add 50K per month to the net, an important increment.
  2. The net changes are small, the froth at the top of the wave, compared to the overall changes.
  3. The size of employment “misses” from expectations are almost trivial when compared to the overall size of these changes. The monthly discussion and market reactions are overdone.
  4. Even the “revised” numbers include sampling error of +/- 100K or more. Some incorrectly believe that this is eliminated with the revisions. Not so. The employment dynamics numbers are needed for that purpose.
  5. The gross job losses show that 25 million jobs are lost each year, even in fairly good times. No wonder so many are concerned about their employment.
  6. The growth in new businesses should put to rest those criticizing the BLS, birth/death adjustments, phantom jobs, fake data, and the like.

Sadly, the President is not alone in misinterpreting employment data. We should understand the value of older, but more accurate data, even if used only to give grades to the early estimates.

Leave a Comment