Automated Data Processing
reported this morning that the private sector added 135,000 seasonally adjusted new jobs in May. It was a modest performance and well below the expert consensus of 171,000. But given how large ADP's revisions of these numbers will likely be in the next couple of months, the value of the report as a guide to any decision-making—or even as an omen for what the government's Friday job report will say—is dubious.
Since the expert consensus is that the government's Bureau of Labor Statistics' report of both public and private jobs will show a gain of 167,000 on Friday, the obvious conclusion ought to be that the results ADP announced today means the federal figures will be lower than expected, too.
The supposedly obvious, however, may very well not mesh with the reality. Ever since Moody's Analytics was commissioned by ADP last October to help it measure the economy's job performance by implementing a new methodology, the results have been off the mark by as much as 50 percent. Its best showing was a 16 percent difference. In other words, off as much or more than with the old methodology.
Last month, for example, the ADP reported 119,000 new private jobs had been added in April. The BLS reported 176,000 new private jobs. That's a gap of 57,000 between them, or 48 percent. In March, the gap was large, too, but the numbers went the opposite direction. ADP reported 158,000 private jobs added, while the BLS reported only 95,000, a gap of 63,000, or 40 percent,.
The average difference between the ADP and BLS reports during the previous seven months: 44,000. Thus, anybody betting the mortgage on what Friday's BLS report will look like based on the results of today's ADP report better have some rent money stashed somewhere.
Below the fold, you can read more about the disjunction between the two jobs reports.
Over the next couple of months, the ADP and BLS will make what could be huge revisions in their reports for May.
For example, in its report for November 2012 the BLS first reported a gain of 147,000 new private jobs. By the time it made its final revision two months later, that figure had risen to 247,000, a 68 percent difference. For that same month, ADP first reported a gain of 118,000 new jobs. Its final revision? A gain of 276,000, or a 134 percent difference.
The gathering of job statistics is an ever-on-the-move affair, fraught with all kinds of obstacles, with formulaic tweaks worked out going forward to account for an increasingly slippery labor market whose past behavior is no long any guarantee of future performance. As better data become available for any particular month, ADP and the BLS make their adjustments. Not conspiratorially but under parameters agreed upon. This is not a diss of the statisticians.
However, the very nature of the parameters skew the big picture.
Skip for today the issue of McJobs and poorly compensated part-time work and lower benefits that plague the economy. And forget for now that a job market recovery that has been consistently labeled weak is actually weaker still because of the vast numbers of Americans who have dropped out of the workforce. The drop-outs have cut the labor participation rate to its lowest level in 30 years which, in turn, has lowered the unemployment rate. If those drop-outs had stayed in the wokrforce and kept looking for jobs they couldn't find, the national unemployment rate would now be close to 10 percent. (Contrary to claims that the drop in the participation rate is all about the baby boomers hanging up their spurs, studies show that only about one-quarter of the ["missing workers"] have retired.)
Even ignoring the impact of those skews, however, it's frustrating knowing that each month the stock market is going to be affected and media anchors and not a few analysts are going to treat those statistics as if they're some exact measure of what's going on despite the fact that the measurement will be revised, perhaps heavily, in the couple of months to come.
••• ••• •••
The ADP National Employment Report is calculated from the payroll accounts of some 400,000 U.S. business clients and about 24 million U.S. employees working throughout the private sectors of the economy.