Why December Private Payrolls Aren’t a Great Predictor of the Jobs Report

December ADP data are especially unreliable for forecasting payrolls, says Jim O’Sullivan, chief U.S. economist at High Frequency Economics. 

WILFREDO LEE/ASSOCIATED PRESS

Be wary of ADP data as a predictor of monthly nonfarm payrolls numbers, some economists warn—especially in December.

ADP, with its monthly private payrolls report, doesn’t aim to replicate the government’s nonfarm payrolls survey. The samples aren’t the same and ADP includes different inputs as well as lagged government data. But many traders and analysts value the ADP report as a timely labor market indicator and one last clue ahead of the key government data that comes two days later—despite an average miss of 61,000 by ADP in predicting the same month’s private nonfarm figure.

The Bureau of Labor Statistics counts a person as employed when he or she gets a paycheck. That contrasts with ADP, which relies on companies to report when an employee leaves, said Moody’s chief economist Mark Zandi. Moody’s partners with ADP to compile the monthly report.

Because of that, December ADP data are especially unreliable for forecasting payrolls, says Jim O’Sullivan, chief U.S. economist at High Frequency Economics. Big December misses likely reflect difficulties adjusting for December purges, when many private employers, prompted by the need to file year-end tax documents, finally drop departed workers from their books.

The December ADP figure is susceptible to missing in either direction, Mr. O’Sullivan says, because there is a reversal of the cumulative effect of the other 11 months, raising the scope for an error in the last month of the year and “adding another layer of complication for all the things ADP has to adjust for.” A notable exception was 2014, when ADP was off by just 1,000. In December 2013, the miss was a hefty 151,000 and in 2012, it was 47,000. Since 2008, the average December miss has been 97,000 versus an average miss, covering all months, of 61,000. The survey’s methodology changed in October 2012 when Moody’s replaced ADP’s previous partner. But ADP has still tended to overshoot private nonfarm payrolls in the last month of the year.

To be sure, the ADP methodology adjusts monthly for the expected purge factor—just like it adjusts for factors such as weather—and some companies do provide more timely information on departed employees. But Mr. Zandi from Moody’s says the margin of error is bigger in December because that’s when purging actually happens for a fair share of business and ADP gets updated records. He said the purge factor is more of a problem when the economy is accelerating or decelerating more rapidly and employers are adding or shedding jobs at a quick pace.

But with the economy on steady ground and “moving in a straight line,” the “bias from the purge effect should be relatively small this year,” he said.

Related reading:

5 Things to Watch in the December Jobs Report

Private Payrolls Rise a Better-than-Expected 257,000 in December

from Real Time Economics http://ift.tt/1S6i7Rx