EJ Antoni/Heritage Foundation: “some suspect government statisticians are committing lies of omission.”

And could it be that Ancient Aliens helped build the Pyramids? Sure!

EJ Antoni links to article noting Senators Roger Marshall (R-KS),  Ted Budd (R-NC), Rick Scott (R-Fla.), Markwayne Mullin (R-OK) and Marsha Blackburn (R-TN) have jumped in to demand investigation of the BLS regarding the preliminary benchmark revisions:

Sen. Roger Marshall (R-Kan.), who is leading the charge, said the public deserves an investigation into whether the numbers were intentionally fudged to boost the Harris-Biden administration.

“Using taxpayer dollars to mislead the public for political gain is an outrageous betrayal of trust and one of the reasons Americans have lost all faith in this Administration,” Marshall told The Post.

Dr. Antoni should know better. Here’s a time series plot of the benchmark revisions, expressed in % terms (employment is growing over time, after all).

Figure 1: Preliminary benchmark revisions, in % of original employment estimate (blue). Red dashed lines at election years. Source: BLS, and individual year BLS preliminary benchmark revision releases.

The downward revision of 0.5 ppts is less than two standard deviations from the mean of -0.1 ppts revision (std dev is 0.26 ppts). This means the downward revision is not statistically significant at the 5% msl.

Further note that if the BLS were fudging numbers before elections, one would expect systematically positive revisions in election years (indicated by red dashed lines).

So, one could criticize the BLS on methodological grounds (Got a gripe with the birth-death model used? Then suggest something better) but to cast aspersions on the motives of the professional staff of a nonpartisan statistical agency seems particularly reprehensible when coming from a credentialed Ph.D. economist who should know better. (Personally, I think it more likely that BLS will in its final benchmark revise upward employment, because (1) the likely bias in QCEW numbers used to calculate the preliminary benchmark, and (2) the fact that the preliminary benchmark revision to private NFP from the CES  is not reflected in the ADP-Stanford series (the latter does not rely upon QCEW or birth-death model).

Dr. Antoni thus joins the ranks of illustrious purveyors of data paranoia, such as Senator Barraso, Jack Welch, former Rep. Allan West, Zerohedge .

 

 

 

 

 

Source link