by Calculated Risk on 1/05/2012 01:29:00 PM

Tomorrow (Friday) the BLS will release the December Employment Situation Summary at 8:30 AM ET. Bloomberg is showing the consensus is for an increase of 150,000 payroll jobs in December, and for the unemployment rate to increase slightly to 8.7%. The consensus is probably moving up based on recent reports.

Here is a summary of recent data:

• The ADP employment report showed an increase of 325,000 private sector payroll jobs in December. Although ADP seems to track the BLS over time, the ADP report hasn't been very useful in predicting the BLS report. Also note that government payrolls have been shrinking by about 24,000 on average per month this year, so this suggests around 325,000 private nonfarm payroll jobs added, minus 24,000 government workers - or around 301,000 total jobs added in December.

However we need to use caution with the ADP report in December. From Jan Norman at the O.C. Register:

... Joel Prakken, chairman of Macroeconomic Advisers, cautioned that the number may be inflated by an annual accounting correction that ADP does with its customers' data every December. That procedure tends to overstate seasonal adjustment of job gains in the aftermath of a recession.

And an explanation from Goldman's Andrew Tilton last year:

Because the ADP report counts the number of people on payrolls, regardless of how many hours they worked (or whether they worked at all), its accuracy depends on company payrolls being up to date. In reality, some companies do not immediately delete departing workers from their payroll records. Those that do not often wait until the end of the year. As a result, December in particular typically sees a meaningful decline in payrolls in the raw ADP data. The reported data are adjusted in an attempt to account for this behavior, but insofar as “purging” occurs to a greater or lesser extent than usual, it could affect the reported numbers. In particular, there was probably less purging in 2010 than in recent years, since data from the Labor Department make plain that the level of “separations” (layoffs or quits) declined this year. If this was less-than-fully accounted for by seasonal adjustment, the reported figure could show a large gain. (Note that the Labor Department’s survey does not suffer from the “purging” problem, since in that survey a person has to have actually reported hours in the survey period to be counted as employed.)

So ADP is probably overstating employment gains again this year.

• The ISM manufacturing employment index increased to 55.1% from 51.8% in November. Based on a historical correlation between the ISM index and the BLS employment report for manufacturing, this reading suggests the gain of ten thousand or so private sector payroll jobs for manufacturing in December.

The ISM service employment index increased to 49.4% from 48.9% in November. Based on a historical correlation between the ISM non-manufacturing employment index and the BLS employment report for service, this reading suggests the gain of around 45,000 private payroll jobs for services in December.

Overall the ISM surveys do not suggest a strong employment report.

• Initial weekly unemployment claims averaged about 375,000 in December, down from 396,000 per week in November, and down from 405,000 per week in October.

For the BLS reference week (includes the 12th of the month), initial claims were at the lowest level since May 2008. This is a very positive sign.

• The final December Reuters / University of Michigan consumer sentiment index increased to 69.9, up from the November reading of 64.1. This is frequently coincident with changes in the labor market, but also strongly related to gasoline prices and other factors. In general this low level would suggest a weak labor market - but slightly better than in the July through November period (the BLS reported an average of 132,000 per month for those five months).

• And on the unemployment rate from Gallup: Gallup Finds U.S. Unemployment Holding at 8.5% in December

Gallup finds U.S. unemployment, not seasonally adjusted, at 8.5% in December -- the same as at the end of November, but down from 9.6% a year ago. Gallup's unemployment measure suggests the government is likely to report essentially no change for December 2011 in its seasonally adjusted unemployment rate, though this December is especially hard to predict.

NOTE: The Gallup poll results are Not Seasonally Adjusted (NSA), so use with caution. This does suggest little change in the headline seasonally adjusted unemployment rate.

There always seems to be some randomness to the employment report, but the overall situation has improved (lower initial weekly unemployment claims, more job openings). However the ADP report is probably overstating December job growth, and the ISM surveys still suggest sluggish job growth - I'll take the over (above 150,000), but I don't expect a strong report as suggested by ADP.