US Private Nonfarm Payroll Employment – What Does the Revamped ADP Series Tell Us?

After a hiatus, ADP released today a new series (compiled in conjunction with the Stanford Digital Economy Lab)  that tries to independently measure private nonfarm payroll employment, rather than predict the BLS series. What does it look like over time, and compared to the old series (and the BLS series)?

First , the various series over the sample we have at least two measures:

Figure 1: Private nonfarm payroll employment from BLS (blue), from ADP (tan), from ADP and Stanford Digital Economy Lab (green), all in 000’s, s.a. NBER defined peak-to-trough recession dates shaded gray. Source: BLS, ADP via FRED, ADP, and NBER. 

Notice that there’s a big difference in level between the new ADP series and the BLS series, reflecting the point that the new ADP series is not aimed at predicting the BLS series (technical description here). In words:

Based on anonymized and aggregated payroll data of over 25 million U.S. employees, this independent measure will detail the current month’s non-farm private employment change and deliver weekly job data from the previous month. Data will be broken out by industry, business establishment size, and U.S. census region. Historical data from the previous 12 years at both monthly and weekly frequencies will be benchmarked and available at launch.

How do the changes in each series look? It’s hard to discern patterns in the month-on-month changes, so I show quarter-on-quarter changes (in logs) in Figure 2.

Figure 2: Three month change (not annualized) in log of private nonfarm payroll employment from BLS (blue), from ADP (tan), from ADP and Stanford Digital Economy Lab (green), s.a. NBER defined peak-to-trough recession dates shaded gray. Source: BLS, ADP via FRED, ADP, NBER, and author’s calculations.

Over the 2010-22M05 period, the adjusted R2 for a regression between the old ADP series and the BLS series was about 0.98, and and new ADP series and the BLS series about 0.78. A lot of this high correlation was due to the 2020 recession. Omitting the 2020M03-2020M12 period, the corresponding R2 was 0.42 and 0.13, respectively. For the 2021M01-22M05 period, the adjusted R2 was 0.01 and -0.03, respectively. In other words, to infer what is happening to the BLS measure from the ADP measure was was unwise even with the previous series.

Hence, in response to the relatively small increase (132K), Goldman Sachs left unchanged their forecast (of +325K).

Here’s a closer look at the recent history of the series, along with the Bloomberg consensus as of today.

Figure 3: Private nonfarm payroll employment from BLS (blue), from ADP (tan), from ADP and Stanford Digital Economy Lab (green), Bloomberg consensus as of 8/30, assuming no revision in July figure (sky blue square), all in 000’s, s.a. NBER defined peak-to-trough recession dates shaded gray. Source: BLS, ADP via FRED, ADP, Bloomberg (accessed 8/30) , NBER, and author’s calculations. 

While the revamped ADP series might not tell us much about what the BLS series on private nonfarm payroll will be, it still tells us that for the companies that ADP covers, growth in employment has slowed. The ADP data can also inform us on what firms are increasing their payroll, and how wages are changing (see the release).

38 thoughts on “US Private Nonfarm Payroll Employment – What Does the Revamped ADP Series Tell Us?

  1. Moses Herzog

    Terrific FYI to the readers. I hadn’t heard about it until this post. Maybe it was in the WSJ and I missed it?? Deems keeping an eye on each month.

    1. AndrewG

      You missed another big NYTimes story about China, ltr. What happened? You’d think corruption and failed dirigiste strategies would interest you. Maybe not if it’s China, I guess?

      Xi Jinping’s Vision for Tech Self-Reliance in China Runs Into Reality
      After heavy national investment in semiconductors to break a dependence on global chips, Mr. Xi seems unhappy with the results.

      By Li Yuan
      Aug. 29, 2022

      Wearing a laboratory coat, China’s top leader, Xi Jinping, inspected a subsidiary of Yangtze Memory Technologies Company, a national semiconductor company based in Wuhan. It was April 2018, shortly after the U.S. government had barred the Chinese telecommunications firm ZTE from doing business with American suppliers.

      The ban was a Sputnik moment for China’s tech industry and its leaders. Despite the country’s success in building smartphones, e-commerce platforms and high-speed railways, they realized that tech boom had been built largely on top of Western technologies, especially chips that power nearly everything. They had to change that — and fast.

      Mr. Xi told the executives of Yangtze Memory, or YMTC, that semiconductors were as important for manufacturing as hearts for humans. “When your heart isn’t strong, no matter how big you are, you’re not really strong,” state media reported him saying. He urged them to hurry and make tech breakthroughs to contribute to the rejuvenation of the Chinese nation.

      Mr. Xi has repeated that message ever since, with growing urgency as the United States tries to restrict China’s access to key semiconductor technologies. But a series of corruption investigations last month into the who’s who of the country’s semiconductor industry suggest that Mr. Xi may not be getting what he expected, or at least not quickly enough.

      Those under investigation include the former chairman of YMTC who showed Mr. Xi around on the 2018 visit, and the head of a giant state fund, known as the Big Fund, which has invested in dozens of China’s biggest chip projects.

      The firms they led are at the heart of the country’s push into semiconductors, the little slices of silicon that act as the brains of computers and other devices. Their downfalls are a public acknowledgment that China is rethinking its gold rush approach of throwing cash at projects in the hope that some work out. And it’s a clear setback for the country’s drive to become technologically self-sufficient.

      Behind the purge lies a tension between Mr. Xi’s vision of government-led tech self-reliance and the very nature of semiconductors.

  2. pgl

    John Eastman is a clown:

    John Eastman, the lawyer who developed former President Donald Trump’s strategy to overturn Joe Biden’s presidential election victory, appeared Wednesday before a Georgia grand jury investigating interference in the 2020 election. Eastman’s lawyers, Charles Burnham and Harvey Silverglate, issued a statement saying they advised him to refuse to answer questions under his Fifth Amendment right against self-incrimination. The lawyers declined to reveal the substance of the questions posed. “By all indications, the District Attorney’s Office has set itself on an unprecedented path of criminalizing controversial or disfavored legal theories, possibly in hopes that the federal government will follow its lead,” the statement said. “Criminalization of unpopular legal theories is against every American tradition and would have ended the careers of John Adams, Ruth Ginsburg, Thurgood Marshall and many other now-celebrated American lawyers.”

    John Eastman is not another Ginsberg or Marshall. He is a criminal and a traitor who is scared to talk to this grand jury.

  3. pgl

    “it still tells us that for the companies that ADP covers, growth in employment has slowed.”

    In a market where people are changing jobs, could it be that people left jobs covered by ADP to working for folks not covered by ADP?

    1. Menzie Chinn Post author

      JohnH: Numerous studies – some cited on this blog – have noted that the household series yields almost no additional information about the business cycle relative to the establishment series.

      1. pgl

        When JohnH is not promoting Putin’s war crimes, his other passion in life is becoming Princeton Steve’s BFF. So we have to excuse his silliness.

      2. JohnH

        Maybe this time is different? Just looking at the graph, the employment series has a pretty good track record of flattening at the onset of a recession.

          1. JohnH

            It so happens that this time lots of things are different than…like the gap between GDP and GDI.

            Krugman even had a column about the puzzling difference between the household and employer surveys.

          2. pgl

            September 1, 2022 at 9:46 am

            Get used to dumb comments like this. Johnny refuses to admit he goofed even when everyone else on the planet knows he did.

        1. pgl

          Gee the employment to population ratio has remained at 60% for several months:

          I cannot imagine any sane person saying this is an indication we were in a recession. But of course JohnH is desperate to be BFF with Princeton Steve so he writes “the employment series has a pretty good track record of flattening at the onset of a recession.”

          Does this lying troll have a shred of evidence to support this absurd claim? NOPE!

        2. pgl

          (1) the employment series has a pretty good track record of flattening at the onset of a recession but
          (2) this time is different.

          Which is it Johnny. Oh wait – (1) is not true. And there is no reason why (2) is true either.
          Johnny’s next comment will tell us: (1) the earth is flat; and (2) the earth is round.

        3. pgl

          BLS has reported. Payroll survey showed an employment increase = 315 thousand.

          Your household survey? Employment increase = 442 thousand.

          Do not go into the forecasting business as you are incompetent at that too.

  4. ltr

    August 31, 2022

    Nobel Prizes given to US 400, given to UK 137, given to Germany 111, given to China 1.

    [ What is so interesting is being completely unaware of the twenty-seven books written by the Cambridge scientist Joseph Needham on “Science and Civilisation in China” (1954-2008). The work is an exceptional triumph of scholarship. Saddening, such a lack of awareness.]

    1. ltr

      June 20, 1971

      Joseph Needham, the Real Thing
      By Richard Boston

      Our vocabulary for describing what is great has been so impoverished by the misuse of Hollywood publicists (Stupendous!!! Colossal!!!) that it is hard to find suitable words to describe the real thing. And Joseph Needham is the real thing: he is one of the great intellects of our time. Merely to call him a polymath gives no idea of his achievement: a fellow scholar at Cambridge University, a man who is not given to making rash judgments and who is well‐versed in the British art of understatement, recently commented to me that you have to go back to Leonardo before you can find anyone with such a grasp of the whole of human knowledge. History, philosophy, religion, mathematics, astronomy, geography, geology, seismology, physics, mechanical and civil engineering, chemistry and chemical technology, biology, medicine, sociology, economics…. just to list the topics covered in his massive “Science and Civilisation in China” would take more than the space of this article (the summary of contents in the publisher’s prospectus covers more than 12 closely printed pages.) And, at the same time as dealing with China, throughout the work Needham compares and contrasts with what was going on all over the rest of the world. In spite of Hollywood, it is stupendous, it is colossal.

      Joseph Needham, Fellow of the Royal Society and holder of the Brilliant Star of China, was born in 1900, the son of a doctor. He took his degree at Cambridge, and in 1924 became a Fellow of Gonville and Caius College, of which he is now the Master. In the same year he married a fellow‐student and fellow‐biochemist, and later Joseph and Dorothy Needham became the first husband and wife both to be made Fellows of the Royal Society.

      When he was 31, Needham published a three‐volume work on “Chemical Embryology.” This led to his “History of Embryology,” and his involvement in the history of science. Then, before the Second World War, an important influence came in the form of a group of Chinese scientists who came to Cambridge to do postgraduate work in biochemistry. Working alongside them Needham developed an interest in Chinese culture and science, and simply for his own interest taught himself the language. He also began to ask himself, first, why it was that during ancient and medieval times China was so far ahead of Europe in science and technology, and then why the advance to modern science took place in the West: these were the basic questions that were later to underlie his great work. One of these Chinese scientists, Lu Gwei‐djen, was a cardinal influence at this time. Not only was she trained in modern science, but also she had access to the old Chinese culture through her apothecary father, to whom the first volume of “Science and Civilisation in China” is dedicated. Later she returned to Cambridge, where she has now been working full‐time with Needham on the project for 12 years.

      When the war came Joseph Needham was, as a Chinese‐speaking scientist, exceptionally qualified to be sent to China as head of the British scientific mission. For four years during the war he directed the SinoBritish Science Cooperation Office, in the course of which he travelled thousands of miles all over China, made contact with a vast number of. Chinese scientists and engineers, and began to build up material about Chinese science.

      Before the war was over he was gathering support for the establishment of an international agency for scientific cooperation, and he met with such success that he has justifiably been called the man who put the S into UNESCO. He was the first Director of UNESCO’s Department of Natural Sciences. When he returned to Cambridge in 1948, he thus had behind him a record of academic and public achievement which was more than enough for a single lifetime. But in 1954 came the publication of the first volume of the great work on which he is still engaged, “Science and Civilisation in China.” *

      * Twenty-seven books (1954-2008) ….

      1. Barkley Rosser

        Yes, ltr, I am fully aware of Needham’s work, and you know that, since I believe you have read one of my books that cites Needham’s work.

        You have been told why I am posting this point and when. It is every time you make YOUR racist point about how other countries have had more than the 4 per million Covid deaths that China has had. You have been repeatedly told that you have way over reported this piece of data, and that every time you post it again, I shall post this completely factual matter about Noble Prizes.

        Got it?

        1. ltr

          I am fully aware of Needham’s work…

          [ I am so pleased, and deeply appreciate all the consideration shown. The kindness and consideration and knowledge are quite wonderful. ]

          1. Barkley Rosser


            I think you also know that I am deeply aware of Chinese culture and history, and I applaud the many genuine achievements China has made and is making now, many of which you publicize here. Many of your messages from official Chinese sources are not only accurate but useful and informative. You get in trouble for repeating some things over and over that maybe are a bit distorted for various reasons, while of course ignoring some unpleasant things.

          2. baffling

            ltr gets in trouble for calling others racist with no evidence. usually done when presented with evidence that contradicts ltr’s pro-communist propaganda. ltr defends vigorously a government which is responsible for the deaths of millions of its own citizens. and the genocide of ethnic minority groups within its own borders. then ltr pretends to have hurt feelings when confronted with these harsh realities.

  5. Macroduck

    This is a huge improvement over the old ADP numbers. ADP collects lots of employment data every month. The old approach was to obscure the information available in that data in order to mimic payroll employment, which we’d always have in a few days, anyway.

    The old method led to forecast revisions which often proved to be a mistake. It was never clear whether divergence between ADP.and private payroll employment represented different information or a poor job of emulating by ADP. The new method removes some of that uncertainty.

  6. ltr

    Really a social-psychological mystery, how we could experience a healthcare calamitous tragedy, as was easily-casually known and as the CDC confirmed just today, but there is an evident need to demean and dismiss the health care efforts of a country that has been exceptionally successful in limiting the spread and effect of the coronavirus. We have a Nobel Prize winning economist, who while American Indians have lost 6 years in life expectancy since 2019 needs to savagely criticize health care experts who have saved countless lives and preserved the health of millions. Evidently the health of American Indians is of no concern, if respecting the efforts of another culture to preserve life could provide reasonable ideas.

    Then there is the curious repeated Nobel Prize comparison, * as though that makes one million seventy thousand American coronavirus deaths just fine. Quite a mystery to me.


    August 31, 2022

    Nobel Prizes given to US 400, given to UK 137, given to Germany 111, given to China 1.

    1. Barkley Rosser


      The mystery is why you feel the need to endlessly repeat a paricular statistic to the point of utter nausea. I am probably going to stop following your ongoing constant repetition of that statistic, which has several problems with it that you never recognize, just because, well, I give up. You are incorrigible and just keep doing it.

      Oh, and super secret agent Moses keeps finding more Nobel Prize winners from the PRC, although the number is still very low. My point was (and remains) that you like to repeat and repeat and repeat a stat that makes PRC look good, but then get all in a huff when somebody else repeats after your repetition another stat that does not make the PRC look so good. But it has clearly become a pointless exercise.

  7. ltr

    Joseph Needham of Cambridge was responsible for twenty-seven books for “Science and Civilisation in China” and Nobel-like prizes were never any concern. The concern was what could be learned by all of us from the science accomplishments:

    “Having written much, whether well or ill, I know not,
    But with devout intention for the healing of the nations.”

  8. ltr

    You’d think corruption and failed dirigiste strategies…

    [ I notice such a comment and immediately stop reading and move on. I am interested in dirigibles but nonsense is rarely even amusing. Dirigibles? ]

  9. pgl

    Trump hires the dumbest lawyers:

    The Justice Department’s argument was “deeply flawed,” Trump’s team said, “There is no question…that the matters before this Court center around the possession, by a President, of his own Presidential records.”

    Deeply flawed? What was the flaw again? That our national security secrets belong to an individual? Oh that’s right – their client is King Donald I.

  10. pgl

    Is Alina Habba the girlfriend of Rick Perry Mason Stryker?

    “They say themselves in these papers that they filed… that this is under the Presidential Records Act,” Habba told Kirk. “So what they did was try and criminalize Donald Trump as they always do. They found these three mundane statutes: espionage and the two others ― obstruction. And they are trying to claim there is some sort of criminal activity.”

    This is precisely the dumb spin Rick has been spinning. Espionage is mundane? Bimbo alert!

    This incompetent lawyer also admitted Trump had those classified documents neatly stored in his boxes but the evil FBI threw them on the floor. Oh how untidy! I guess this bimbo never heard of how law enforcement often wants to take pictures as evidence. BTW Trump tweeted how evil it was for the FBI to make his office so untidy. MAGA.

  11. James

    Thanks for posting – interesting to me is that some of the sectors reporting the greatest decreases in employment are also among the sectors reporting the highest percentage of remote work – IT, financial services, professional services – see page 14:
    Although I would expect with an increase in interest rates – we see a cut in jobs in the FIRE sectors – . finance, insurance, and real estate.
    Menzie – as a Wisconsin resident and worker – I take special interest in economic policies of Governor candidates – I find Michels’ to be a poorly stated rehash of GOP meaningless talking points that GOP has pushed for 40 years – cut taxes/regulations for job creators! and then all the workers can scratch out an existence in his libertarian paradise
    One interesting feature – Michels’ wants to “Increase American energy production and distribution” – no details on how – but my guess would be that Michels will pump WI taxpayers’ $$$ into his family pipeline business. In other words don’t run Wisconsin like a business – but run Wisconsin as his business.

  12. ltr

    June 1, 2022

    By Ryan K. Masters, Laudan Y. Aron and Steven H. Woolf


    BACKGROUND  Prior studies reported large decreases in US life expectancy during 2020 as a result of the COVID-19 pandemic, disproportionately affecting Hispanic and Black populations and vastly exceeding the average change in life expectancy in other high-income countries. Life expectancy estimates for 2021 have not been reported. * This study estimated changes in life expectancy during 2019-2021 in the US population, in US racial/ethnic groups, and in 21 peer countries. The study compared outcomes across five US racial/ethnic groups and is the first to estimate changes in life expectancy during the pandemic in non-Hispanic American Indian/Alaska Native and Asian populations.

    METHODS  US and peer country death data for 2019-2021 were obtained from the National Center for Health Statistics, the Human Mortality Database, and overseas statistical agencies. The 21 peer countries included Australia, Austria, Belgium, Canada, Denmark, England and Wales, Finland, France, Germany, Israel, Italy, Netherlands, New Zealand, Northern Ireland, Norway, Portugal, Scotland, South Korea, Spain, Sweden, and Switzerland. Life expectancy was calculated for 2019 and 2020 and estimated for 2021 using a previously validated modeling method.

    RESULTS  US life expectancy decreased from 78.85 years in 2019 to 76.98 years in 2020 and 76.44 years in 2021, a net loss of 2.41 years. In contrast, peer countries averaged a smaller decrease in life expectancy between 2019 and 2020 (0.55 years) and a 0.26-year increase between 2020 and 2021, widening the gap in life expectancy between the United States and peer countries to more than five years. The decrease in US life expectancy was highly racialized: whereas the largest decreases in 2020 occurred among non-Hispanic (NH) American Indian/Alaska Native, Hispanic, NH Black, and NH Asian populations, in 2021 the largest decreases occurred in the NH White population.


    1. ltr

      August 31, 2022

      U.S. Life Expectancy Falls Again in ‘Historic’ Setback
      The decline during the pandemic is the sharpest in nearly 100 years, hitting American Indian and Native Alaskan communities particularly hard.
      By Roni Caryn Rabin

      The average life expectancy of Americans fell precipitously in 2020 and 2021, the sharpest two-year decline in nearly 100 years and a stark reminder of the toll exacted on the nation by the continuing coronavirus pandemic.

      In 2021, the average American could expect to live until the age of 76, federal health researchers reported * on Wednesday. The figure represents a loss of almost three years since 2019, when Americans could expect to live, on average, nearly 79 years….


Comments are closed.