ADP has put together a fairly miserable track record in forecasting BLS job data. This month, they have, once again, revamped their methodology in an attempt to more closely track BLS data, which they claim after the BLS revisions one year later their data will match up.

Why? Why is that goal is even remotely desirable?

ADP has their own proprietary data sources; they track employees (new and existing), they can track payroll dollars (total gains and losses, and per employee changes), off of the actual payroll checks going to employees. Why try to imitate the BLS output each month? ADP can create a very specific set of reports that ARE PURELY DRIVEN BY THEIR OWN PAYROLL RECEIPTS, that stands on its own.

Why even bother messing around with trying to imitate or forecast BLS data?

The BLS Non-Farm Payroll numbers are somewhat flawed, subject to massive revisions, and fatally flawed due to how the Birth Death adjustment has been applied.

If ADP wants to contribute something valuable, they should stop trying to forecast BLS, and instead generate their own, proprietary, data driven monthly NFP numbers.

Consider how NFP lags the end of recessions, continuing to be negative long after the recovery begins. Employers often expand output without new hiring: They offer full time employment to part–time workers, and offer overtime to full timers. ADP can track that, and capture a real time improvement in the economy long before it shows up in the BLS data.

Inquiring minds want to know: Why bother aping a data series that is not particularly helpful in real time?


via Barron’s Econoday


ADP Reports 693,000 Private-Sector Jobs Lost in December
January 7, 2009, 8:28 am

U.S. Economy: Companies Cut Payrolls at Faster Pace in December
Bob Willis and Courtney Schlisserman
Bloomberg, Jan. 7 2008

Category: Data Analysis, Employment

Please use the comments to demonstrate your own ignorance, unfamiliarity with empirical data and lack of respect for scientific knowledge. Be sure to create straw men and argue against things I have neither said nor implied. If you could repeat previously discredited memes or steer the conversation into irrelevant, off topic discussions, it would be appreciated. Lastly, kindly forgo all civility in your discourse . . . you are, after all, anonymous.

19 Responses to “ADP Goes Back to the Drawing Board”

  1. VennData says:

    ADP, the supply side economists of employment forecasting.

  2. Bruce N Tennessee says:

    Well, yes. But there was another interesting fact in the ADP numbers….

    He added that the U.S. economy probably contracted at a 5.5 percent annualized rate in the fourth quarter and would shrink 3.5 percent in the first quarter of this year.

    “After that economic growth is going to depend on the size and timing of the fiscal package that is being discussed in Washington right now,” Prakken said.

    OK>>>>> If the general accepted definition of a depression is >10% decrease in GDP..IT APPEARS TO ME THAT 3.5 +5.5=9.0….isn’t that in the ballpark??


    BR: GDP is reported in annualized numbers. A -4.0% Q GDP is actually a 1% drop X 4 quarters . . .

  3. km4 says:

    Americans, by nature are an optimistic bunch. Even in tough times, there is something to be optimistic about. …That is probably the only reasonable explanation for the findings of this survey conducted by Glassdoor, a Sausalito, Calif.-based start-up that ranks employers by taking anonymous feedback from the employees.

    Despite the dismal global economy, and wide spread layoffs and rising unemployment, 61 percent of surveyed employees are not be willing to take a pay cut if they discovered their job was in jeopardy. A whopping 40 percent expect a pay raise in the next 12 months, despite job cuts at their employer. Of those eligible for an annual bonus, 57 percent expect a bonus and 40 percent do not expect a bonus

    Most amusing part — four out of five employed adults say they are not concerned about being laid off from their job in the next six months. One in five employees, are concerned they will be laid off during the same period. Perhaps the pessimistic 20 percent are reading the news.

    To the optimistic 80% – enjoy what you’re smoking for now ;

  4. Bruce N Tennessee says:

    Got ya. Thanks. Once again, trying to work at the salt mine, and blog, didn’t see that for what it was.

    Panic over.

    I think I’ll go buy my Boxster now…

  5. AmenRa says:

    I’ve always liked the ADP numbers because they came from actual payrolls. Usually the BLS is revised in the future and comes closer to the ADP number at that time.

    BTW Paulson is again begging for more money. I think he realizes that Congress planned on not giving him the other 350 billion until the next administration (and probably with conditions).

  6. AmenRa says:

    I almost forgot. TPTB always come out to try and talk up the market when it looks like it will be a big down day.


    BR: WTF is TPTB ?

  7. The reason ADP tries to ape the BLS is that they understand the deep and abiding belief that people of all strata (from fed economists to J6P blogger) have for data (i.e., bullshit) that comes out of the government. For some strange reason, people mostly believe government data is unbiased, whereas data from a private source is necessarily biased by the need to make money. In my opinion, this is a lazy and not-strongly-held belief, and fails quite quickly once it is proved that the government is lying (as it nearly always is).

    If ADP wants to quit aping the BLS, yet still have the same currency, it will need to prove the BLS is bullshit. Shouldn’t be too hard.

  8. mark mchugh says:


    You know what’s hard to find right now? Companies run by dunces, in contracting sectors, that are only off their 52 week high by about 12%.

  9. BG says:

    Curmudgeon said:

    “For some strange reason, people mostly believe government data is unbiased, whereas data from a private source is necessarily biased by the need to make money. ”

    REALLY? I guess we are polar opposites.

    My first thought was that everyone knows the government reports are damn-near useless and in general are a total waste of MORE taxpayer money. I was hoping that the credibility of a private report might actually tell the rest of us what is really going on.

    You know things are so fucked up right now, it would be oh so nice if someone would do the RIGHT thing for a change and give us the straight numbers without all of the usual massaging.

    Will we ever learn anything? IMO, it is time to cut the bullshit and get down to the facts of what is really going on. It is time to stop the little government-spin games and give it to us stratight. And like I said earlier, it was my hope that a private report might actually provide us with that. Any report coming out of DC has been massaged, pampered and pilfered to the point its final conclusion is 180 degrees out from where the original data was pointing. I have long sense felt government reports are a means unto an end.

    If the spin doctors can talk the prescribed message long enough then maybe some people will actually give credence to these government reports when they come out. I’m sorry; but, I’m not one of them.

  10. Bill McDonald says:

    Publishing ADP’s payroll information is worth millions in free advertising and name recogniton for ADP, and tracking the BLS numbers better will insure that wonderful free coverage continues.

    ADP data is from a subset of US business, they tend not to cover very small companies. ADP’s market share is not uniform across all business sectors, as an example, Bank of America has a large market share for many high tech companies. So the ADP data is not a perfectly representative sample across all sizes and sectors.

    While ADP data is a large enough sample of the whole to have value, the BLS numbers are still more popular, and get more converage from the financial news media.

  11. leftback says:

    Wonder what would happen if BLS reported -1,000,000 on Friday?

  12. wunsacon says:

    I bet ADP was asked or decided on their own not to present their own data for political and/or business reasons.

  13. wunsacon says:

    Er, I meant really just “political” reasons… I don’t see how this is better for their business (notwithstanding Bill’s reasonable attempt — because I think anyone who comes out with their own, more accurate numbers will gain more of a following).

    Maybe they don’t want to rock the boat and end up in a situation where they need TARP money but can’t get it.

  14. Concerned American says:


    The market would go up. Less business expense equals more profit. Crazy, but that seems to be the pattern.

  15. MatB says:

    I never understood why they tried to mimic BLS numbers either. ADP pays 1/6th of all non-government employees in the U.S., that seems like a good enough set of data to stand on it’s own, completely apart from the BLS numbers.

  16. jasonch says:


    I completely agree with your point about the ADP data being pure, and grounded in some objective reality, based on all the transaction volumes that they process.

    The question it raises, however, is one of BLS methodology. Wasn’t the BLS methodology created way back when, before there were computers involved in payroll? Perhaps it’s time that the government updated that methodology. But they shouldn’t need ADP.

    The IRS receives withholding on every single paycheck transfer in the US from employer (with Tax ID numbers) to employees (with social security numbers). The data should be easily joined with employees address to make it geographically relevant, and SIC codes for the employer’s industry to identify sector breakdowns.

    I just don’t see why the employment data is even a medium difficulty problem to solve.

  17. Everyone seems to be ignoring the fact that for the last 6 weeks unemployment claims have been decreasing. For two weeks straight now the 4-week moving average is *negative*.

    We don’t need ADP or anyone else to tell us that job losses are *slowing*.

    The data is there if you choose to use:

    Last week everyone ignored the fact that the claims *drop* was the largest in 16 years.
    The basis is negative right now, and no matter who reports the data (ADP, BLS, DOL, etc, etc), if it is good, it will be dismissed.


  18. mark mchugh says:

    Just cycling back through old posts.

    Barry, TPTB is conspiracy theorist for “The Powers That Be”.

  19. [...] the history of the ADP report’s failure to correlate very well with its bogey (BLS’ Non-Farm Payroll), the reaction is revealing of [...]