This Friday is jobs Friday. At 8:30 a.m. ET, the BLS will release its initial estimate of job growth for the month of March.

As a precursor to the biggest indicator of them all, yesterday’s ADP employment report came in at +191K for March, much less than the +215K the market expected. The ADP jobs figure is for private sector jobs only.

The initial reaction from economists was one of credibility, with many economists questioning whether ADP had any.

Bank of Tyoyo-Mitsubishi’s Chris Rupkey sums up this group of economists’ feelings:

A lot of swings and misses for ADP when it comes to forecasting the monthly payroll jobs report.  They thought jobs slowed further to 130K in October and instead jobs went sky-high to 212K.  Actual jobs cooled to 87K in December after two months of big 200K gains yet they told us the number would be the biggest yet, up 238K.  Wrong.  Another swing, another miss.  Why do we continue to take a look at this?

On the other end from the naysayers are those who see the value and predictive ability of the ADP.

Here’s the correlation between the two.

ADP vs. BLS Difference in Monthly Jobs Number

As is no surprise, the correlation is quite high.  The problem is, it’s not perfect.

Here’s the month over month absolute job growth for each of the surveys side-by-side:


The revised January and February 2014 ADP numbers are quite close to the BLS figures, with the January difference at 8K and the February difference at 3K.

Are the economists right that the BLS number is likely to come in above the ADP-implied figure?

Over the course of the two series, the BLS figure has, on average, come in 6K above the ADP number.  Not much when considering the statistical noise already apparent in the BLS figures.

Here’s the difference between two series on a M/M job gains basis (ADP M/M job gains less BLS M/M job gains).

Difference in MM Job Gains, ADP - BLS

Here’s the distribution of the differences.  Given that the benchmark reported BLS figure includes government, it’s no surprise that there’s a slight negative bias to the M/M differences.


Overall, about 52% of the time the BLS figure comes in higher than the ADP figure.

Presuming government employment continues to grow, the economists might be right.  Although, the probability that the BLS figure will come in high probably isn’t as high as what the dead-certain economists would have market observers believe.

In any event, now’s the time to make your bet: ADP or the economists?  The correct response will be rewarded handsomely.