U.S. employment data – heavily skewed to low-wage, revolving-door jobs – have blown past consensus expectations for the last three months. The consensus estimate puts the BLS payroll number at 245,000. The ADP private payrolls number came in April 1 at +189,000, well below the consensus +225,000 estimate. If the BLS number is in line with the ADP number, it would come in at 193,000, well below consensus. That’s not the whole story: a shift to hiring by small business might explain the growing discrepancy.
The relationship between the BLS and the ADP payroll measures is sloppy on a month-to-month basis, to be sure, and a great deal of random fluctuation and statistical efforts introduces noise into the equation. There’s information in the noise, however. Asia Unhedged examined the regression equation between ADP and BLS employment data, and found something of interest.
The overall regression fit is about 88% on month-to-month changes in payroll employment. Within this relationship, to be sure, there is a great deal of random fluctuation, shown in the residual line. Notably, residuals are a standard deviation above the mean. In other words, BLS payroll changes have been outperforming ADP payroll changes during the past three months by an exceptionally high margin. There are any number of possible reasons for this. One is that a lot of jobs have been created by smaller firms that do not report to ADP during the past several months. We see this clearly in the National Federation of Independent Business survey of small business job creation plans. If small business hiring continues, the BLS number will continue to outperform the ADP number.
Asia Unhedged has called attention to the bifurcated U.S. economy: services are doing well while manufacturing is fading.
The Institute for Supply Management March survey of manufacturing employment is available, and it shows a sharp downturn. The non-manufacturing employment survey is not yet available.
The services hiring boom might continue and meet the consensus number for BLS payroll employment. But the risks are weighted towards the downside.