Data Paranoia Watch: “I’ve read that others think the CES was manipulated to provide a more rosy picture heading into the election”

That’s a quote from Mr. Steven Kopits, on why the CES survey showed such a rosy picture on NFP growth. This statement joins a long pile of such allegations, e.g.,  Senator BarrasoJack Welchformer Rep. Allan WestZerohedge, Mick Mulvaney, among others. All I can say is that (1) if there was a conspiracy, they didn’t do a very good job, or (2) it’s a conspiracy so vast that it encompasses not only the other offices of the BLS, but also to ADP.

Figure 1: Private nonfarm payroll employment, change from January 2022, from BLS October  release  (blue), from ADP October release (tan), and from private covered employment in Quarterly Census of Employment and Wages, seasonally adjusted by author using X-13 (green). Source: BLS, ADP via FRED, BLS (QCEW), and author’s calculations. 

The BLS and ADP series in the October releases would’ve been known before the election date on November 8th. The QCEW series would’ve only been known up to 2022M03 at that time.

In both these cases, the cumulative change in employment according to the BLS CES would’ve been below the QCEW (if anybody was looking at that) and certainly below the ADP.

But, perhaps “they” got to ADP too…



32 thoughts on “Data Paranoia Watch: “I’ve read that others think the CES was manipulated to provide a more rosy picture heading into the election”

  1. Not Trampis

    I do find it interesting that people who mostly do not have much expertise about statistics can wax lyrical about statistical conspiracies’.

    1. Moses Herzog

      BlueStatesResidentKopits has assured us, in another past life he was a TA for a Statistics prof. I mean, he’s confused about what confidence intervals are but…… I see no reason why that sounds like a peculiar claim. Can you see any reason that would sound funny to people?? After all, he knows the stats terms “holistic”, “suppression”, “multiple job holders”. I nominated Kopits for this year’s Deming Lecturer Award. Oh, excuse me, I got a pain in my parietal lobe, I gotta go take my B-12 now.

  2. Macroduck

    “And you almost think — I’m not saying that, and I’m not a conspiracy person. . . . Half the people in this room are saying it.”

    — Donald Trump

    “I don’t know. Maybe? A lot of people think so.”

    — Glenn Beck

    “I’ve read that others think the CES was manipulated to provide a more rosy picture heading into the election.”

    — Stevie Kopits

    “Hillary Clinton will be arrested between 7:45 AM – 8:30 AM EST on Monday – the morning on Oct 30, 2017.”

    — Q Clearance Patriot

    “There will be a very large meteor shower that lasts for two weeks, it will be seen in the Northern hemisphere, containing the Nozic Message.

    “One of the meteors will seem different than all the others, that is because it is a ship landing on Earth, starting preparations for the first Nozic War.”

    — @aesthetictimewarper

    One way of presenting a notion is as valid as any other, I suppose. Right?

    1. Moses Herzog

      Could it be that the CES survey started out in a lab in Wuhan?? [ Can you hear the X-Files theme music playing??]
      Barkley Rosser April 13, 2020 at 2:06 pm
      “I do not know about this remark, but the current leading theory is that this did not start from somebody eating a bat (or pangolin) from a Wuhan wet market. Rather now the best evidence suggests that it came from a lab where they were trying to develop a vaccine against bat-induced illnesses, but were careless with health security.”

      One of the many bend-over-laughing-’til-your-stomach-aches hilarious parts of Barkley “FoxMulder” Rosser’s arguments about the virus coming from a Wuhan lab was that it was difficult for Barkley to imagine that a bat with the virus would have crapped on other animals, and apparently the only way Barkley could imagine bat crap getting on food before/during its transport to Wuhan was if the bat had “cavorted” with seafood.

      Barkley Rosser April 14, 2020 at 1:10 pm
      “…….. It is probably not from a Wuhan seafood market, long the most popular theory. Problem is we are pretty sure it is from bats (possibly pangolins), but bats were not sold at this “wet market.” If it is from there, then somehow a bat infected some animal that was there, but, as its label says, it was a seafood market, not the usual animals bats interact with…….. “

      1. Menzie Chinn Post author

        Moses Herzog: The post was about labor market developments, but if you are to quote, you should try to quote without omitting important caveats that the commenter has made. The full comment [link] is:

        Wow, Moes, you have gone full-bore hysterical here. No, I am not going to read any of your liinks. Obviously there is debate about this, and it is unresolved. I suspect some of these might be a few weeks old and about the clearly discredited theory pushed by Sen. Tom Cotton and some others that it was consciously cooked up in a bioweapons lab in Wuhan. No, that is not what this is about.

        I first heard it on NPR and then read about it in a WaPo column on April 2 by David Ignatius, who is about as well-informed and smart as anybody at WaPo. Last time I saw him he was with Madeleine Albright. He is super well connected and not remotely a goofball. It seems that parts of that column are holding up, although the bottom line remains we do not know for sure.

        A more recent link is at Daily Signal by Fred Lucas from two days ago, . Here are the four things, according to him (who takes the Ignatius column seriously, while also saying much of it is unconfirmed):

        1. Definitely not a bioweapon.
        2. It is probably not from a Wuhan seafood market, long the most popular theory. Problem is we are pretty sure it is from bats (possibly pangolins), but bats were not sold at this “wet market.” If it is from there, then somehow a bat infected some animal that was there, but, as its label says, it was a seafood market, not the usual animals bats interact with. As it is, the Chinese have shut it down and thoroughly swabbed the place, so we shall probably not get to know for sure one way or another.
        3. There is indeed a lab 300 yards from the suspect but unlikely seafood market that is a branch of the Chinese CDC and Prevention, and indeed they have been studying bat and other originated diseases over the last several years, as I said and you so hysterically dismissed. In this case, it would not be a plot or a conspiracy, but, as I said, a careless mistake (and Chinese sources note that this lab had been criticized for carelss safety practices). Lucas views this theory as “complicated,” but it is certainly plausible, and, if anything, gaining ground now given how unlikely its main competitor, the seafood market, seems to be.
        4. China has been poor with information on all this, which has clearly fed all sorts of theories.

        So, Moses, there is a source, hardly the only one. If you want to dump on this story, try and upend the column by David Ignatius. Do any of your links specifically address his column? If so, tell us what is in the link. As you know by now, you have zero credibility with me regarding your links, almost always totally worthless junk.

        As the entire comment makes clear, the seafood market is still considered a possibility. Never is it stated that it is impossible.

        I would also refer you to the DNI’s declassified report on the sources of the infection, as linked to in this post. One of the IC elements argues for the lab accident hypothesis.

        I would finally note that back in April of 2020, one could be reasonably uncertain as to how the virus escaped.

        1. Moses Herzog

          Menzie, I respect you for defending a colleague and possibly a friend. You know I really can’t attack you for wanting to defend someone, that’s a positive attribute (possibly personally annoying to me in this specific instance, but nonetheless a positive attribute and something to be admired that you defend a colleague).

          @ Menzie
          Now, I’m gonna try to explain to you my problem with Professor Rosser. He has played, what I consider to be a type of “trick” or sham as it were, multiple times. And this is not as “personal” for me as you might imagine, although I’ll concede maybe personal related to other issues. But here is the crux of my problem with Professor Rosser, he states these things rather emphatically. Even with the entire comment, if you’re being honest that even in the context of the entire comment, he states (and stated in other comments even more emphatically, for the love of God are you going to make me hunt them down??) that it most likely came from the lab. Now….. here is what he does over and over and over, He waits (on the about 1 in 20 times he’s correct and says “See, I was right, my theory was right”. He wants it both ways. He wants credit for strongly leaning to one side of the argument when it happens to go his way, then goes “but I didn’t say that” the 85% of the time he is wrong. It’s disingenuous Menzie, any way you care to slice it It’s a form of LYING PERIOD.

          And most scientists, most of the better ones at that time, were laughing their heinies off on the lab theory. It wasn’t holding any water with respected scientists. But you know the scientists the lab theory got attention from?? The same type personalities as Barkley Rosser~~those grabbing on to the most oddball scenario in a sad grab for personal attention.

          And I’ll tell you, I’m not sure I would find Rosser refusing to admit he’s wrong so annoying/ AS annoying, if it wasn’t so patently obvious what a scream ploy it is for attention. Most men outgrow this around what??~~age 25?? He’s doing it in his 70+ years. Watching him grasp out for these bad theories (BECAUSE it is the one-in-fifty bad theories that end up being right he figures will buy him attention, and then refusing afterwards to admit he was wrong, is like watching Milton Berle and Bob Hope at the end of their careers get up and tell the joke to an audience they know won’t get laughs, all on account they get to be on stage,

  3. Steven Kopits

    If I believe the Philly Fed, the CES was not only off, it was way, way off. By two orders of magnitude for an error that clearly began in the month of March and accumulated month after month until at least December.

    Now, that’s either deliberate or some sort of systems failure, and a major one at that. How did that happen? Unlucky? Assumptions changed? Wild incompetence?

    I don’t know, but the error was large, persistent and cumulative. It’s now nine months later. Has the error been addressed? Is some sort of working group or committee working to fix it? What is the status of remedial work? Has it been finished? When will it be completed? What was the cause of the problem?

    What is a member of the public to think?

    1. Macroduck

      The structure of Stevie’s argument:

      From “If you believe (an author at) the Philly Fed…” to the next paragraph which leaves behind doubt and takes as give that the Current Employment Statistics survey is wrong.

      In that second paragraph, there is no “if”, but rather “that’s either deliberate or some sort of systems failure…” How didwe get from “if” to “deliberate”? Through bad faith argumentation.

      In the next paragraph, we get pure Glenn Beck-style blathering: “I don’t know, but…” and then a shower of concern about how to fix a problem which two paragraphs earlier may not have existed, depending on “if you believe”.

      This is terrible stuff. High school students would be told to rewrite a paper which relied on such poorly executed logical sleight of hand.

    2. pgl

      “If I believe the Philly Fed, the CES was not only off, it was way, way off. By two orders of magnitude for an error that clearly began in the month of March and accumulated month after month until at least December.”

      Why lie to us? You do not believe a word you utter. Here’s the real deal. You fish for whatever statistics you can find to argue for your RECESSION nonsense and then just trash anything else that contradicts your utter BS.


    3. Menzie Chinn Post author

      Steven Kopits: It might benefit you to study up on when BLS benchmark updates its CES estimates. It’s done once a year. I used the preliminary benchmark for March 2022 in my construction of the implied benchmark. In March of 2023, BLS will provide the final benchmark revision. You should also study up on how they do that updating. Notes are here.

  4. Macroduck

    Oh, my! There it is again!

    “If we were adding so many full time jobs, why was both productivity collapsing and GDP declining?”

    — Stevie Kopits

    When work hours increase while output is falling, productivity falls. I’ve explained this to you before. Output/hours = productivity. In fact, if you look at the BLS website, you will see the terms “labor productivity” and “output per hour” used interchangeably. The answer to this question you keep asking is in the question, itself. And it has been pointed out to you repeatedly.

    Couching your question in terms of jobs, in fact full time jobs, rather than hours is a further sign that you don’t understand how productivity is calculated, that you don’t understand the concepts being discussed. If you don’t have a textbook handy, there are plenty of resources on the internet which could help you. Take your time.

    By the way, why this renewed insistence on half-baked ideas and nonsense expressions?

    1. Macroduck

      Here’s the picture:

      Notice that periods of relatively high employment growth are roughly simultaneous with periods of low productivity growth, amd vice versa. Note also that upward spikes in productivity growth tend to correspond roughly with downward spikes in employment growth, and vice versa.

      Asking “If we were adding so many full time jobs, why was both productivity collapsing and GDP declining?” indicates a lack of familiarity with the data, aswell as the calculation of the labor productivity series.

      1. baffling

        nice graph. thanks for posting. it certainly makes sense that as you hire a bunch of new employees, productivity drops, at least at first. and it appears that productivity then begins to increase once those workers begin to gain more experience. spikes in hiring seem to drop productvity. slow or steady state hiring seems to improve productivity. anybody who has managed workers understands this dynamic.

  5. pgl

    15th round of voting for the next Speaker:

    McCarthy 216
    Jeffries 212
    Present 6

    I demand a recount. The Republicans clearly manipulated the data.

  6. pgl

    Stevie Kopits has never argued in good faith on any issue. He is making this issue as convoluted as possible but it is actually quite easy. Stevie picks a data where the Payroll survey kept showing strong growth whereas the Household survey did not. Of course we have seen episodes of this before but of course Stevie is not interested in either longer term view or even a holistic view. No Stevie is making out like Lawrence Kudlow back in the Bush43 years.

    As this year went on – one could trust the strong growth news from the Payroll survey or the less reliable news from the Household survey. But as any GOP hack might do – Stevie touted the latter since reported growth was lower. He even made up some BS about multiple job holders which of course he cherry picked and misrepresented.

    But now that the latest news from the Household survey confirms that employment growth was indeed strong this year, in true Kudlow style Stevie says this survey is not reliable either. We all know Kudlow is a professional liar. Which begs the question – who is paying Stevie to spread one lie after another?

  7. Jwil1

    Steven’s error was large and persistent. Is his error deliberate? Is he incompetent? Unlucky? Bad assumptions?
    Why hasn’t he addressed his error? Formed a working group? What is the status of his efforts to correct his large error? When will he finish his efforts to correct his initial error?
    What is the public to think when a guy on the internet can be this wrong and make literally no effort to correct his error or become more informed? What is the public to think of such a failure?

    Given the frequency of his errors, one might think……

    1. pgl

      Steven’s error was large and persistent. Is his error deliberate? Is he incompetent? Unlucky? Bad assumptions?
      Why hasn’t he addressed his error?

      All good questions. Of course Stevie’s “performance” on other issues are similar. As are JohnH’s performance on a variety of issue with the same persistence, incompetence, and failure to address his errors. It is a common theme here – alas.

  8. Macroduck

    Menzie wrote:

    “The results they (researchers aobtainedt the Philly Fed) were not definitive; in fact if one reads the notes, one finds that their results depend (substantively I would guess) on the method of seasonal adjustment used.”

    Stevie responded:

    “So you’re saying the Philly Fed screwed up its analysis and we should ignore its work? That’s your view?”

    Putting aside Stevie’s attempt to put words in Menzie’s mouth, this response suggests that Stevie suffers from a profound misunderstanding of how economic research – in fact, scientific research in general – works. A single study, using a particular data set and particular methods, is not the end of the quest – not definitive. Others review the study, attempt to replicate the results, suggest alternative approaches and potential weaknesses. If the single study seems to have turned up new information, that information become part of our corpus of knowledge, but may still not be definitive.

    The authors of any piece of research, if not burdened with a consultant mindset, are likely to acknowledge that their work is not “the answer”, but is instead part of a process of intellectual endeavor.

    What we have in Stevie’s efforts at rebuttal is evidence that he:

    – Is unaware of how productivity is calculated.

    – Is unfamiliar with the behavior of productivity and employment data relative to each other (as he was formerly unfamiliar with the number of multiple jobholders relative to total employment, etc.).

    – Is unfamiliar with how individual empirical efforts fit into the wider process of research and intellectual progress.

    Stevie isn’t alone, but Groucho had a point. Is that a club you’d want to join?

  9. Macroduck


    Seeing a bunch of “504 Gateway Timeout” errors here. Might want to give your service provider a heads-up.

  10. Macroduck

    About a year ago, the NY Fed (not “the Fed”, but rather the New York Region Federal Reserve Bank) launched a new measure of supply chain performance, the Global Supply Chain Pressure Index. That index had shown an easing of pressure through mid-year, but then more recently had begun to show renewed signs of strain. This article from the NY Fed looks into that renewed stress on the supply chain and finds that China’s Covid policies are the root of the problem:

    Supply-chain problems tend to reduce output and increase inflation. This is a problem for Fed policy in that Fed tools cannot address inadequate supply directly. Fed tools can only reduce demand in order to reduce supply/demand imbalances. This, in fact, tends to compound supply-chain problems by reducing incentives to invest in supply-chain improvement. It sucks.

    The good news is that global sourcing has been moving away from China to countries with better public health policies. Even so, China’s role in supply chains isso large that the shift away from sourcing from China has till now had limited effect in reducing global supply chain vulnerability to China.

    China’s long list of economic and policy problems does suggest, however, that sourcing will continue to move away away from China. ltr’s recent focus on propaganda featuring direct foreign investment into China suggests that ltr’s master’s are aware of the migration of production out of China..

  11. Macroduck

    BlackRock, in its 2023 outlook, has this to say about inflation and central banks:

    “Central bank policy rates are not the tool to resolve production constraints; they can only influence demand in their economies. That leaves them with a brutal trade-off.

    “Either get inflation back to 2% targets by crushing demand down to what the economy can comfortably produce now, or live with more inflation. For now, they’re all in on the first option. So recession is foretold. Signs of a slowdown are emerging. But as the damage becomes real, we believe they’ll stop their hikes even though inflation won’t be on track to get all the way down to 2%.”

    As to what this implies for output:

    “…we expect a 2% fall in GDP for central banks to get inflation down to what the economy can comfortably produce now.”

    The combination of “…inflation won’t be on track to get all the way down to 2%” and this train wreck of expression “…inflation down to what the economy can comfortably produce now” leaves some room for doubt about BlackRock’s outlook, but not much. It’s a forecast, after all, so the point estimate is not the most important thing.

    1. Ivan

      The main thing to me is that BlackRock understand that the Fed’s desperate attempt to hammer a nail in with a screwdriver will end up pretty ugly. They presume that the Fed is just doing this for show (to intimidate economic growth downwards) and will stop before any real damage. That has to be the presumption if we don’t presume they are crazy. However, the damage they have done to housing could be more than they can control – and hard to reverse. Time will tell.

  12. ltr

    January 8, 2023

    a ) There were 5,226 coronavirus deaths in China on May 26, 2022.
    b ) There were no coronavirus deaths from May 26, for nearly 6 months, through November 19.
    c ) From November 20 through January 7, 2023 there have been 41 coronavirus deaths in China, bringing the total from 5,226 to 5,267.

    d ) During the nearly 6 months of no coronavirus deaths in China, there were 300 to 500 coronavirus deaths each day in the United States.
    e ) Coronavirus deaths in the United States continue at 250 to 450 each day.
    f ) There have been 1,121,089 coronavirus deaths in the United States through January 7, 2023.

  13. baffling

    continued irrelevant comments by ltr, repeated insidiously on many different blog posts. often times from ccp sources. please stop this cowardly act which abuses prof. Chinn’s hospitality on this site. it is rude and unbecoming ltr. especially when you know those statistics were provided by the ccp to other outlets intentionally incorrect.

    1. Baffling

      I know. They permit commenters to use multiple names to pretend their position has support from others. Even j zanfino would agree.

Comments are closed.