ADP vs. BLS

The ADP survey says the United States lost 84,000 private sector jobs in December.

Nonfarm private employment decreased 84,000 from November to December 2009 on a seasonally adjusted basis, according to the ADP National Employment Report®. The estimated change of employment from October to November was revised by 24,000, from a decline of 169,000 to a decline of 145,000.

The decline in December was the smallest since March of 2008. Employment losses are now rapidly diminishing and, if recent trends continue, private employment will begin rising within the next few months.

December’s ADP Report estimates nonfarm private employment in the service-providing sector increased by 12,000, the first increase since March of 2008. However, this employment growth
was not enough to offset continued losses in the goods-producing sector. Employment in the goods-producing sector declined 96,000, with employment in the manufacturing sector dropping 43,000.

If you recall the BLS November unemployment report claimed the United States only lost 11,000 jobs while ADP reported a 169,000 loss of private sector jobs (now revised to 145,000).

Now someone, somewhere is wrong. Add to these discrepancies, this Zero Hedge post, which found a 32% unemployment actual payout vs. the numbers of those collecting unemployment insurance.

One of the biggest factors is ADP does not count government jobs. According to the WSJ, ADP has tracked the BLS data consistently being 62,000 job losses below the BLS data, but this year and according to ADP, the BLS is less accurate. What has happened in this last year to cause such large variances from the ADP and BLS is an exercise for another post. We do know seasonal adjustments are not published as well as a birth/death model which is revised every February.

So, we hope to lost past the headlines and examine revisions as well as make sure the seasonal adjustments do negate themselves on the year to year.

Subject Meta: 

Forum Categories: 

This may be the reason

What has happened in this last year to cause such large variances from the ADP and BLS is an exercise for another post.

I don't have a link, but I recall reading a biz article around a year ago that an increasing number of companies were outsourcing their payroll bookkeeping to ADP, which should work to increase the accuracy of their payroll inventory tracking.

I strongly suspect this might be the answer.