ADP reports 162,000 new private jobs were created in September

Automated Data Processing reported Wednesday morning a seasonally adjusted 162,000 new private jobs were created in September. A consensus of economic and business experts surveyed by Bloomberg earlier this week had estimated that ADP would report 140,000 new jobs. The number of jobs for August was revised to 189,000. ADP only reports gains or losses in private-sector jobs.

The question today is whether the firm has again set everyone up for disappointment when the government's Bureau of Labor Statistics report comes out on Friday. That's been the way of things for the past six months.

In August, ADP reported that the economy had produced a seasonally adjusted total of 201,000 new private jobs (now revised to 189,000, as noted). But the BLS reported that only 103,000 private jobs had been created, with lay-offs of 7,000 in the public sector, for a net gain of only 96,000. That is not quite enough to keep up with monthly additions to the working-age population.

ADP bases its report on the payrolls it processes for some 430,000 businesses covering about 24 million employees. [Methodology here.] BLS bases its report on its Current Employment Statistics survey of 410,000 worksites. [Methodology here.]

Over the past six months ending in August, the BLS has reported 629,000 new seasonally adjusted private-sector jobs have been created, with 49,000 lay-offs in the public sector. The ADP, on the other hand, has reported the creation of 978,000 new private-sector jobs, also seasonally adjusted. A huge disparity.

However, if you take into account the past 18 months, it's another story. For the eight months from March-October 2011, the BLS reported more private-sector jobs than ADP. In fact, 388,000 more. This was followed by a four-month "transition" in which ADP reported more jobs one month and the BLS reported more the next. Since March of this year, ADP has reported more.

For the entire 18-month period, ADP has reported a gain of 2.887 million new private jobs. BLS has reported 2.917 million new jobs. Remarkably close. Especially when compared with often large differences between them each month.

But there's a lump in the pudding. Last week, the BLS provided a preliminary reading on its annual "benchmark revisions." That is, just as it does every year at this time, it gave people a heads up about its reanalysis of data it had gathered in the 12-month period ending in March 2012.

In the reanalysis the bureau found 453,000 additional private-sector jobs had been created between April 2011 and March 2012 that it didn't catch the first time. It also found 67,000 more government lay-offs, for a net gain of 386,000 jobs not counted previously. That is, as noted, a preliminary figure; the final won't be added until February. But the final is usually very close to the pre-announcement.

So, when those 453,000 private-secgtor jobs are added to the BLS's total over the 18-month period from March 2011-August 2012, it comes out to 3.37 million new private jobs. That's closing on half a million jobs that BLS says were created in the period that ADP doesn't.

When you subtract the 282,000 lost government jobs, the difference between the BLS and ADP numbers for the past year and a half is 201,000 jobs.

You might be tempted to ask, if the BLS undercounted for the 12 months before March, then why shouldn't we assume it is still doing so? That's possible. But, remember, this benchmark revision is an annual affair. Last year, the bureau reported that 366,000 fewer jobs had been created from April 2010 through March 2011.

All these numbers and the differences between the two sources may make your eyes water a bit. But while they are an (obviously) imperfect tool, they do serve an important purpose in showing us what direction we're headed and the general speed we're moving. They don't, however, answer some questions that really matter, among them being, how good are these jobs and what can (and will) be done politically to generate more of them faster starting Nov. 7?