Some people say the craziest things:

500K net job creation is to be expected in a recovery:

More on Governor Romney’s 500,000 job creation target, expressed as a percentage growth rate

The scientific method doesn’t involve predicting observed data:

A Random Thought on the Scientific Method

Fiscal multipliers are typically near zero:

Birthers, (Stimulus) Deniers, and Economic Myths


There is no big trend in global temperatures:

Hottest on Record

If a variable is  I(0) by construction, then it’s best to treat it as I(0) in finite samples

The Kansas Palmer Drought Severity Index, Unit Roots, and Snakes in a Room

The budget deficit is not a good measure of the budget deficit:

Further Documentation on “Stephen Moore Is a Liar”

On Trump’s honesty:

True Believers

Trump is right – the virus is just going to disappear one day

Covid-19 Pandemic in the US: The Trump-Annotated Timeline




131 thoughts on “Howlers

  1. Bruce Hall

    I was curious what outrageous comment I might have made 5-years ago:
    Bruce Hall
    January 16, 2015 at 2:05 pm
    That’s good news. Far fewer deaths from cold weather. Longer growing seasons in the northern hemisphere. Potential greening of the Sahara region. All consistent with global warming.”

    January 16, 2015 at 3:01 pm
    Actually, it means shorter effective growing seasons because plant growth slows after 86 degrees F and shuts down entirely when temps hit the mid-90s. It’s a bit counter-intuitive, but read the literature. And it also means more deaths from heat. Like a lot of folks, you need to get beyond this idea that a warming of (say) 4 or 5 degrees Celsius simply means a few more days when you can play golf. Global warming of that magnitude (which is well within standard confidence intervals) means that the bottom third of the US will see summer temperatures roughly comparable to what Saudi Arabia sees today. Of course, the Canadians might do fairly well under global warming. Maybe that’s why the Canucks are so keen on Keystone.

    2slug rightly pointed out that “northern hemisphere” does include Canada… which was why I tried to be inclusive with my statement rather than restricting it to the USA since being inclusive is very noble, especially considering that means including Canadians. Of course, one has to qualify which crops stop growing in higher heat. Some actually do pretty well while others prefer cooler temperatures. Precipitation plays a part, too. Most crops do not like desert or semi-arid conditions. But then, if you don’t have a philosophical antagonism to GMO crops, then many crops can be bred, hybridized, or genetically engineered to be heat tolerant.

    We had quite a string of high 90s days in Michigan this year, but it doesn’t seem to have affected many crops adversely. Too cold or too hot early in the growth cycle or when flowering can disrupt crop production, but plants seem to tolerate heat as long as there is adequate moisture.

    1. pgl

      Only Bruce Hall would think he gets to cherry pick his worst comment. Come on Brucie Boy – you make dumber comments than that by the hour.

    2. 2slugbaits

      Bruce Hall one has to qualify which crops stop growing in higher heat.

      All crops stop growing in excessive heat, but I thought it was obvious that I was referring to the major food crops: corn, soybeans, wheat and rice. Rice hates very warm nights. Wheat shuts down after 95 degrees. Optimal growing temps for corn and soybeans is in the mid-80s, but like wheat it stops growing in the mid-90s.

      Of course, the Canadians might do fairly well under global warming.

      Are you suggesting that the rest of the world move to Canada?

      We had quite a string of high 90s days in Michigan this year, but it doesn’t seem to have affected many crops adversely.

      How do you know? Did you ask your crops? You should ask your agronomist or your seed company.

      many crops can be bred, hybridized, or genetically engineered to be heat tolerant.

      Of course, that would be very expensive. But as many agricultural economists have pointed out, that won’t work because there is no economically viable transition path. You can’t get there from here. The problem isn’t just genetics; it has to do with the fundamentals of agricultural economics. And there is no evidence that hybridized crops are in fact heat tolerant.
      Extreme heat is the single best predictor of corn and soybean yields in the United States. While average yields have risen continuously since World War II, we find no evidence that relative tolerance to extreme heat has improved between 1950 and 2005. Climate change forecasts project a sharp increase in extreme heat by the end of the century, with the potential to significantly reduce yields under current technologies.

      1. Bruce Hall

        It’s so wonderful to have foils like 2slug and pgl. A near record temperature for Death Valley is conflated as “we all gonna die from da heat”. The actual record was set in 1913. LOL.

        Of course the trend of high temperature extremes is rising. Oh, wait.

        Well, maybe we won’t have to move to Canada after all. Eh, 2slug?

        You guys should really stick to economics and leave climate and medicine to others.

        Oh, and this year’s crops? That despite the derecho windstorms that hit the heartland.

        1. 2slugbaits

          Bruce Hall . A near record temperature for Death Valley is conflated as “we all gonna die from da heat”

          Is that your understanding of the climate change discussion? You’re not even close.

          Of course the trend of high temperature extremes is rising. Oh, wait.

          What does the highest temperature on record by state have to do with global warming? Again, you seem very confused about the topic.

          That despite the derecho windstorms that hit the heartland.

          Sorry, but try reading your own link. It was based on the USDA report from the same day, and that report specifically said that it did not include the effects of the windstorms:
          However, it is to be noted that any damage caused by the recent derecho windstorm is not reflected in this report.

          Corn yields per acre in Iowa were actually down and are lower than Minnesota’s. Oh wait…Minnesota is further north and a little cooler than Iowa. Here’s a more current projection (from yesterday):
          Iowa corn dives to just 177.81 bu. per acre estimate while soybeans hit 1146.30 pods per 3X3 square. The state comes in second to last in corn and soybean expectations, falling behind only Nebraska and Minnesota, respectively.

      2. Bruce Hall

        2slug, not a lot specifically on Michigan crops. Search term: michigan fruit crop

        Didn’t really see anything related to hot temperatures. Perhaps it wasn’t hot enough. Midland, MI (appropriate name)

        1. 2slugbaits

          Bruce Hall What do temperatures in Michigan apple orchards have to do with global warming??? Apple trees don’t grow in the south because they require a long period of cold weather. How many apple trees grow in Florida.
          Factoid: My great-grandparents came to this country from Germany in the late 1800s. They settled in Louisburg, KS and grew apples. At one time apple production was a big deal in Kansas. Today…not so much. The climate just doesn’t support apple production anymore. But I do like apple cider and if I want to get in touch with my ancestors, I buy this:

          As the ad says, growing apples in Kansas today is “challenging.”

    3. baffling

      “But then, if you don’t have a philosophical antagonism to GMO crops, then many crops can be bred, hybridized, or genetically engineered to be heat tolerant.”
      interesting. the knuckle draggers drop creationism for evolution and genetics when it becomes convenient.

      1. Bruce Hall

        baffling, as an agnostic/atheist, I wonder how you attribute that description to me? Science is a wonderful tool, but too often interpreted by fools (I add that to keep the humorous ad hominem comments coming). The stereotyping is world class here.

        I presume your Ph.D. is in genetic engineering, so I’m anxiously awaiting your other pronouncements in that field.

        1. baffling

          ” I wonder how you attribute that description to me?”
          knuckle dragger is an appropriate description of you, bruce hall. but i won’t take credit, i believe 2slugs is the one who brought such a description onto this blog.

  2. Rick Stryker

    Wow. Looks like I’m on triple secret probation. Wonder what I did….? Of course. Menzie must be upset at the news that it is highly implausible to think that Covid will be as bad as the 1918 flue pandemic. I understand that progressives are counting on a really bad pandemic to not only defeat Trump, but also empower President Harris, after the rapidly deteriorating Joe Biden is removed from office, to enact the most radical leftist makeover of the US since FDR. It must be deeply disappointing to learn that the suffering they secretly hope to exploit isn’t in the cards. But I’m just the messenger.

    Menzie thinks of my statements as “howlers” but I like to think of them as aphorisms for a new conservative era. The ignorant would be well-advised to ponder each aphorism carefully and diligently to gain enlightenment. I’ll list them in Menzie’s order.

    Aphorism 1: Absolute levels of monthly employment increases should be adjusted for the size of the labor force if they are to be compared for different time periods
    Aphorism 2: The test of a scientific theory is not its ability to explain existing observations, but rather its ability to predict new observations that haven’t been observed
    Aphorism 3: The Obama stimulus plan wasn’t effective
    Aphorism 4: When comparing data points between periods, it’s a good idea to test whether they are statistically different before asserting that they are
    Aphorism 5: Because of potential specification errors, cointegration is not the first choice of many time series experts
    Aphorism 6: The budget deficit is not always an accurate estimate of the change in borrowing of the Treasury
    Aphorism 7: Trump is the most honest politician we’ve seen or are likely to see in our lifetimes
    Aphorism 8: It’s dishonest to quote Trump out of context in order to give his words a meaning that they didn’t really have

    1. Menzie Chinn Post author

      Rick Stryker: “flue”?
      Aphorism 1: even after adjusting for labor force, 500K was very rare.
      Aphorism 2: But you can’t prove with observations that aren’t ever observed. I have a theory God exists…
      Aphorism 3: Most Wall Street economists and most academic economists disagree with you.
      Aphorism 4: The trend is not zero.
      Aphorism 5: But everybody has got to deal with nonstationarity.
      Aphorism 6: But Moore said the *budget deficit* was a certain value, not that the Treasury borrowed a certain amount…
      Aphorism 7: Wow.
      Aphorism 8: Reading the entire quote, it’s still a crazy statement.

      Rick Stryker, you are invaluable as a font of craziness to cite in class, and I hope you keep on commenting. And I hope you will have as your epitaph:

      I believe that Trump is the most honest politician we’ve seen or probably will see in our lifetime.

      That says it all.

      1. Rick Stryker


        Obviously you disregarded my advice to ponder the aphorisms carefully and diligently. For example, aphorism 2 means that a scientific theory should make predictions about phenomena that haven’t been previously observed. Then you go out and observe whether the prediction holds. For example, General Relativity predicted that light would bend around massive bodies. No one had ever observed that. Eddington set up an experiment to see whether that prediction could be observed and it was.

        1. Menzie Chinn Post author

          Rick Stryker: Or, one could observe the seeming epicycles are inconsistent with a geocentric view of the universe, and then proceed accordingly. People had hundreds of years of observations…

          1. Rick Stryker

            You are still not getting aphorism number 2. This is basic methodology of science, Menzie. You really should have learned this in high school.

            It’s not enough to explain what you already know–those hundreds of years of observations you referred to you already know about. You need to predict something new that you didn’t already know about and then confirm that prediction with some observation or experiment.

            Global Circulation Models (GCM) used for climate change modeling can’t satisfy that basic requirement. They are much too complicated, fudge key physical processes with parameterizations, and predict too far into the future to be verified today.

            We recently saw another example of this problem in the Covid crisis. “Scientists” had very well-worked out models of pandemic dynamics and corresponding policy prescriptions based on those models. But those models were never tested by making predictions and then observing whether the predictions were correct. Once they were used in real life, we saw the models were massively wrong.

            But this is how real science progresses. You formulate a model that must, at a minimum, explain what you already have observed. The acid test is to make predictions about what you don’t know and see if the model is correct. Having failed so spectacularly, pandemic models, and their implied policy prescriptions, will now be improved.

            All of this is standard methodology of science. The reason I’ve listed this methodology point as a conservative aphorism is that progressives are constantly using unproven and untested models to bludgeon people into accepting that governmental intervention is somehow necessary. They justify their proposals by saying that “they are following the science.” The have bumper stickers that say “I believe science.” They attempt to intimidate people with the prestige of real science and brand opponents of their policies as “science deniers.”

            But untested, unproven models are not science. That is the meaning of aphorism 2. Conservatives (and this means you Greg Mankiw) need to remember that.

            Aphorism 7 is probably the hardest for you to understand, but if you did understand it you would understand why Trump defeated Clinton and why he is still such a dangerous opponent of the Left, despite the headwinds he has at the moment. Michael Moore is about the only person I know of on the Left who is smart enough to get Aphorism 7.

            And no, 7 doesn’t mean that Trump is good at lying his way into power. Just the opposite. One reason it’s so hard for people on the Left to understand 7 is that they can’t see what consummate and incorrigible liars their own side’s politicians are. They are too blind and self-involved to see it.

          2. 2slugbaits

            Rick Stryker Simple question. Can you offer a scientifically plausible model in which ever increasing CO2 concentrations will not lead to higher global temperatures? If not, then we’re just arguing over “when” not “whether.”

    2. macroduck

      So the new conservative era will rely on aphorisms rather than on facts. Essentially, some guy makes strong statements that others of his tribe are required to accept, no questions asked. Got it. But, but…how is that different from conservatism today?

      In a movement in which strong statements are required to be believed, the head aphorism-writer is a big poohbah. Once again, we should not take our trolls’ comments here as intended for regular readers. They are meant for either an innocent by-stander who may be made to doubt the facts (junk science tactics) or to impress current conservatives poohbahs so the writer may be elevated to poohbah status. Ricky is trying to prove to his masters what a good boy he is. Doing so while Trump’s odds of re-election decline as time passes is meant to prove he is loyal to his masters.

      1. macroduck

        So Menzie, maybe elevating Ricky’s profile by devoting a post to his howlers is not in the best interest of the country. Howling is the whole point. Ricky’s poohbah prospects improve when you take note. Just look at where “Jane Gaunt” is today.

        1. Menzie Chinn Post author

          macroduck: Well, maybe. But I’m waiting for the day that, like John Barron, his identity is revealed so he gets directly the attention he so richly deserves.

          1. baffling

            the last time menzie hinted at his identity, rick took a months long hiatus from the blog. his return hints at a possible change of employment status?

    3. Dr. Dysmalist

      “Rick Stryker”:

      Your ‘models’ you espouse here cannot be evaluated because you keep them hidden. However, based on your discussions of them, I am highly convinced that they are not math, they are ‘mathiness’ that’s intended to put an extremely thin veneer of rigor on a steaming pile of bullshit. The ‘analyses’ that you have presented here for many years are neither rigorous nor intellectual. They are ‘intellectualiness,’ as transparent and bereft of rigor as Stephen Colbert’s ‘truthiness.’

      Dr. Chinn and several of the commenters expect a high death toll from the pandemic because they have looked all of the data with intellectual rigor and intellectual honesty, two concepts with which you are manifestly unfamiliar, based upon your long commenting history.

      The fact that you accuse our host of being “upset” by a lower death toll than from the 1918 pandemic is indefensible and inexcusable. For even suggesting it, you are a monumental asshole, and for that you should be banned, you and CoRev both. Obviously, you are intellectually and morally stunted and twisted, just like your unceasing defense of Trump, a man with absolutely zero redeeming qualities as a human being. You have forfeited the level of respect you have been accorded thus far, even though, in truth, intellectual dishonesty and gaslighting deserve no respect at all.

      If you’re going to insist on learning nothing here and maintaining your belligerent ignorance, you should just go away.

      1. Rick Stryker

        Dr. Dysmalist,

        How did I hide my models? When I estimated R0, I provided a link to the paper that describes the method I used. Also, when I made an argument about the ultimate covid death toll, I showed all the calculations. If you can’t understand any of that, that’s not my fault.

        1. 2slugbaits

          Rick Stryker As I said before, I don’t particularly disagree with your claim that the current Rt is ~1.02. That means it is increasing. Of course, that’s just a mean estimate and the CI bounds could be higher or lower. But there’s tension within your argument. If you are calculating the Rt based on confirmed cases, then for the sake of consistency you should also be counting the CFRs based on confirmed cases. And that gives you a CFR around 1.4%, not the 0.6% or 0.3% you calculated. In other words, in order to get to a 0.3% CFR, you have to generate an Rt much higher than 1.02. You can’t have it both ways.

          But it’s this statement that bothers me: Probably the numbers will never really get as high as these calculations would suggest because 1) R0 will not stay up above 1 permanently if deaths and cases are increasing
          You may very well be right that the fatality numbers never got as high as what you calculated, but if that’s the case then it happens because people internalize the very actions that you are opposing. If the case count gets high enough, then people will very likely social distance, wear masks, stay home, and otherwise slow down the economy. Which is exactly the same result you’d get if the government told people to social distance, wear masks, stay home, and otherwise slow down the economy. Again, you can’t have it both ways. If you truly believe in the herd immunity approach, then you have to go all in. And if you go the full herd immunity approach, then you’re really depending on people NOT socially distancing, and NOT wearing masks, and NOT staying home and NOT behaving in ways that slow down the economy. It seems you want to have your cake and eat it too. Or more likely, you want to convince rubes like CoRev and Bruce Hall to go out as though COVID is just like the flu in order to pump up your portfolio. Do as a say, not as I do.

          BTW, you really didn’t dispute my claim that hospitals were overwhelmed in places where the disease ran rampant. The fact that some hospital in Boise, Idaho wasn’t overwhelmed is irrelevant to whether or not hospitals in Milan or Paris or Detroit or Phoenix or NYC were overwhelmed. And hospitals in some departments because of the virus, so it’s not the case that the virus didn’t affect hospital operations overall. Had we not taken strong lockdown actions, then there is every reason to believe a lot more hospitals would have been overwhelmed. The whole point of issuing warnings is to get people to change behavior so that bad things don’t happen. And then when people heed that advice you can’t turn around and say those warnings were wrong because bad things didn’t happen. That’s just stupid.

      2. 2slugbaits

        Rick Stryker In order to keep things compartmentalized, I’m using a new post to comment on your understanding of the SIR model. The R0 is the initial transmission rate. Here’s the NIH’s more precise definition of the difference between R0 and Rt:
        Exploring these parameters and their implications further, the difference between R0 and Rt is related to the proportion of individuals that are already immune (either by vaccination or natural infection) to that pathogen in that population. So another way of calculating Rt for a pathogen in a given population is by multiplying R0 by the proportion of that population that is non-immune (i.e. susceptible) to that pathogen.6 Hence, R0 will only equal Rt when there are no immune individuals in the population (i.e. when all are susceptible). This means that any partial, pre-existing immunity to the infecting agent can reduce the number of expected secondary cases arising.

        The R0 is sometimes referred to as the basic reproductive number and Rt is the effective reproductive number. The R0 never changes.

        Now, as to the basic SIR model, there are two key parameters. One parameter captures the intuition of Rt and represents the rate at which someone infects people per day. So if there are 100 individuals in the population and 1 person is contagious and infects 1 person per day while contagious, then that parameter (call it “beta”) is 1 / 100 = 0.01. The second parameter represents the time a person is contagious. If someone is contagious for 5 day, then each day they are 20% of the way through their contagious period. If contagious for 10 days, then each day is 10%. Call that percentage “nu”. The herd immunity threshold is simply “nu” divided by “beta”. So if “nu” is 0.20 and “beta” is 0.01, then the threshold value is 20. That means the rate at which people will continue to become infected each day will keep increasing until only 20 people are still in the susceptible category. Notice that this is exactly the same model that is used in advertising to model product growth. Also notice that there are two parameters that are important, not just the “beta” parameter. If instead of a “nu” parameter of 0.20 we used 0.10, then the new threshold value would be .10 / .01 = 10, meaning the rate at which people become infected would not begin to fall until 90 people had become infected and only 10 people were still susceptible. My point is that I don’t think you took the “nu” parameter into account in your calculation. It matters.

    4. pgl

      “Menzie thinks of my statements as “howlers” but I like to think of them as aphorisms for a new conservative era.”

      Since aphorism is just a fancy way of saying pithy statement – I take it you are now endorsing that conservatives flat out lie on a daily basis in a pithy sort of way! Maybe with that viewpoint – Kelly Anne Conway can have you write Trump’s acceptance speech.

  3. joseph

    Stryker: “Trump is the most honest politician we’ve seen or are likely to see in our lifetimes”.

    You can stop right there. There simply no point in arguing with Stryker because he has no absolutely no concept of truth or falsity. You’re wasting you time.

    Worse, as Mark Twain said “Never argue with a fool. Onlookers may not be able to tell the difference.”

    1. noneconomist

      Most honest politician we’ve seen in our lifetime.
      That explains a lot: Rick Stryker IS blind.

  4. 2slugbaits

    Speaking of howlers, how about this gem. It’s this morning’s interview on CBS This Morning.

    When caught lying about supposed voter fraud in Indiana he tried his motor mouth routine that works so well on Fox News. But the CBS interviewer wasn’t having any of it. But Mike “Empty Suit” Pence kept running at the mouth. But the worst was Pence’s dishonest claim that he didn’t know anything about QAnon. Pence likes to pretend that he’s a super moral Christian, but I started to worry that his nose was going to grow right through my television screen. I have to agree with Politico’s Tim Albertson when he (a Republican!) lamented how Pence has become the most pathetic and morally compromised member of Team Trump.

    1. baffling

      trump allies should take notice how he simply threw bannon under the bus when he was busted for fraud. the walls continue to crumble around trump and all who participated in this madness. this is how all grifters come to an end.

    2. macroduck

      As a born Hoosier, I have pretty good access to Indiana scuttlebutt. Pence, who was headed for defeat in his 2016 re-election bid, is generally recognized as just not that bright. Trump rescued him from obscurity on the condition that the Pence who was disappear, replaced by the Pence who would be. The Pence who would be would appeal to the religious right (check) but never outshine Trump (check – don’t make the Bannon mistake) and do whatever dancing is needed to make Trump’s vices into virtues. The only authority higher than Trump is Mother. God? God is handy as long as his kid’s beatitudes don’t come into it.

  5. baffling

    just like 2slugs, i thought menzie perhaps served a higher authority which kept rick on this site to educate him. but just like 2slugs, it becomes apparent that menzie keeps rick on the site so that he can periodically beat the little critter into a pulp. rick stryker better return to his day job as the gay porn star, because as a commenter on an econ blog, he fails miserably.

    1. Rick Stryker


      You are revealing your homophobia Baffling. I’m not surprised. I’m sure racism is lurking in there too, but you keep it hidden.

      1. Menzie Chinn Post author

        Rick Stryker: I think baffling was highlighting hypocrisy, which is certainly germane when contemplating the Conservative movement in 2020, or Trumpism specifically. I saw no criticism of homosexuality per se.

        1. Rick Stryker

          Baffling did not mention any hypocrisy in his comment. He merely said I should go back to my day job as a “gay porn star.” Not just a porn star. He needed to throw in the “gay” part. I know homophobia when I see it. Disappointing to see you defending him.

        2. Rick Stryker

          Baffling did not mention any hypocrisy in his comment. He merely said I should go back to my day job as a “gay porn star.” Not just a porn star. He needed to throw in the “gay” part. I know homophobia when I see it. Very disappointing to see you defending him.

          1. Menzie Chinn Post author

            Rick Stryker: Quoting baffling verbatim: “or instance, rick stryker is a gay porn actor (google it) who believes it is defensible for somebody to discriminate against others in the gay community.”

          2. baffling

            in the past, i assumed rick was ignorant of his gay porn star moniker, so i gave him the less controversial nickname of dick striker, being the conservative that he is. however, his comments seem to suggest he was very aware of the name. on further analysis, his moniker was probably a freudian slip. exactly what videos have you been watching online, rick stryker?

    2. Barkley Rosser


      Actually I think it is partly to provide examples of people not understanding econometrics that he can use in his classes, and he has essentially admitted having done soon more than one occasion, which is certainly excellent pedagogy for which he should be congratulated. There is also simply the matter of shoeing tolerance and open-mindedness, even to people who are regularly spouting silly and propagandistic things.

      As regards Rick’s own aphorisms, indeed the supreme howler must the #7, the claim about Trump being “the most honest politician.” I mean really, the latest count on his verified lies ot at least substantial mistepresentations since becoiming POTUS now exceeds 20,000, with this rate now exceeding 23 per day. probably Rick will argue that most, if not all, of these reports are “fake news,” and maybe some of them are themselves misrepresentatons or exaggerations. But let us say that as many as 4/5, the vast majority, are not true lies. That still leaves over 4,000 with a recent rate exceeding 5 per day.

      Heck, we have seen GOPists go mightily after Dem politicians for one or two lies decades old, see Hillary’s 1995 false claim about having landed in Bosnia in 1995 in a helicopter in the middle of war, which I believe she did quietly admit did not happen eventually. And now we have them whupping it up over Biden’s plagiarism in 1988 of Neil Kinnock material in his campaign, the revrelation of whic hforced him to drop out of the race. In Acasemia we are seriously down on people plagiarizing other people’s work, but this was a campaign, essentially copying lines from another campaign, not remotely the same thing, and not actually a lie. Indeed, I think they are having trouble catching him in even one lie, and if 1988 counts, well, that was 32 years ago. Even at his lower rate, Trjump lies more in one day than we have Hillary and Biden doing together in their entire careers.

      Really, Rick, how ridiculous do you want to look here?

      1. Rick Stryker


        I’m not doing my job if progressives don’t think I’m ridiculous. I think you guys are ridiculous too for:

        a) falling for the fact-checking scam, in which partisans disagree with Trump and pronounce his statements false or misleading because they disagree with him. Any conservative could do the same to any Democratic politician and get a list as least as long as Trump’s
        b) not recognizing that Trump often exaggerates or engages in hyperbole to manipulate the news media into talking about something he wants them to talk about. The press smugly records Trump’s hyperbole in their fact-checking book while Trump laughs at them in the White House. Trump’s supporters don’t feel mislead in any way and are laughing with Trump.
        c) not seeing that Trump is incredibly honest where it really matters: Trump keeps his promises. He never lies about what he intends to do and everyone knows exactly where he stands. You can’t say the same about any other politician in our lifetimes, Democrat or Republican

        And by the way, Menzie has never gotten an example of not understanding econometrics from me. I’ve deigned to correct his econometrics more than a few times, when I felt like it.

        1. 2slugbaits

          The Rick Stryker from December 2015 would have a different opinion about Trump’s honesty and fitness for the job.

        2. Barkley Rosser


          Oh dear, I fear arguing with is as big a waste of time as arguing with Moses Herzog, who has been spouting Trump propaganda about Joe Biden being senile, which even most Fox News bigshots were last night admitting is clearly not the case after his smash hit acceptance speech at rhe Dem convention.

          a) Well, I am wrong that the sum total of lies of Biden and Hillary is less than what the reduced number, 5, of lies Trump says daily, but not by a whole lot. I have not checked on Hillary, who I suspect has more of them, but I have done a deep check on Biden lies. It looks like he has made as many of 7 in his nearly half century career, maybe 8. This is still about a third of what is the current unadjusted daily rate of Trump lying. So, no, Rick the list of Dem lies does not even get beyond one day’s worth of Trump lying.

          b) is just incoherent.

          c) Ah, so here we have the claim that Trump keeps his campaign promises. Well, there are many he has not, such as building his wall, which he has added a big 3 miles to, although sort of upgrading portions of what is there already, although now we have this embarrassment of Bannon getting busted for ripping off suckers who want the non-delivered wall.

          But then we have indeed Trump fulfilling promises that are complete disasters, even as his media bubble praises him for doing so and his followers have any idea what a big disaster he has engaged in. The really big one is his removal from the Iran nuclear deal, which has led to Iran moving closer to getting a nuclear weapon. His policy has led to the greatest humiliation of the US at the UN in its entire history. Only the Dominican Republic supports his disastrous policy. This is horrendous, but I have no doubt next week his sycophants will lie totally that this is great.

          I should warn you, in case you do not know, that this bs by Trump is personal to me. I personally know individuals who remain at this time in the Riyadh Ritz-Carlton, put there by MbS with the support of Trump and his worthless son-in-law. This is directly connected with Trump’s humiliating removal of the US from the JCPOA. I warn you that this is way over your head. This is a global screwup by Trump.

          As for your final remark, you should not make a bigger fool of yourself with such nonsense. At least 3 of Menzie’s anti-aphorisms specifically were on econometrics issues, where you made a complete fool of yourself. i have never seen you successfully find a fault with any econometric analysis by Menzie.

          As for me, I mostly stay away from these matters, but I both personally know the deep details but have published on how UW-Madison became one of the world’s leaders in econometrics, with Menzie a current example of that dominance. This line goes from Sewall Wright, whom I knew, one of the formulaters of the neo-Darwinian synthesis of evolutionary theory that relied upon clarifying the statistical foundations of population genetics, his protegee, the late James Crow, who brought him to Wisconsin. Because of them, the son-in-law of Wright’s co-formulater of population genetics, R.A. Fisher, that would be George E.P. Box of Box-Jenkins and Box-Cox and more, went to UW in its Statistics Dept. His and the presence of Wright and Crow brought Arnold Zellner to the econ dept and then after that Art Goldberger, who should shared the Nobel with Klein, and ran the dept for decades (and with whom I studied econometrics). I suspect that Menzie himself, and even Jim H., much less Bob Flood, know all this.

          Anyway, Rick, you have zero credibility claiming that you ever caught Menzie in making an econometric error. Sorry.

          1. 2slugbaits

            Barkley Rosser which even most Fox News bigshots were last night admitting is clearly not the case

            It’s worse than that. A new book will be coming out next week based with some of Fox’s most prominent on-air personalities. Even the loudest pro-Trump on-air cheerleaders at Fox privately believe Trump is off his rocker and worry about his mental health.

    3. pgl

      Darn Menzie challenging us to Google the part about Rick Stryker and your gay porn star tag. Dumba$$ me took the challenge and look what popped up:

      Every Wich Way (1990) – Jeff Stryker Jade East Brad Carlton Mercedes Coy Rick Stryker VHS

      The picture for this video was enough to make me gag. But who is Jeff Stryker?

  6. Ulenspiegel

    baffling wrote : “rick stryker better return to his day job as the gay porn star, because as a commenter on an econ blog, he fails miserably.”

    In that kind of day job premature intellectual ejaculation is no show stopper, I have to admit your advice economically really makes sense.

    On the other hand a statement like “Trump is the most honest politician we’ve seen or are likely to see in our lifetimes” may indicate a species with a very short life span, e.g. a thunderblight. It’s complicated, really complicated.

  7. pgl

    Protect the Vote? Now this is a real HOWLER!

    LeBron James calls out political ad campaign for misconstruing tweet

    LeBron James fired back at a digital ad campaign for misconstruing one of his tweets about voter suppression, tweeting Friday: “Nobody should be able to use my name (or anyone else name) to lie and deceive about the election.” James was addressing a story in The Washington Post about a website called Protect My Vote that, according to the report, has been circulating ads on Facebook that paint mail-in voting as unreliable. One of the Facebook posts features a June tweet from James where he called polling closures in Kentucky a sign of “SYSTEMATIC RACISM and OPPRESSION.” The Facebook post misconstrues the tweet, suggesting that James was linking the closures to the expansion of mail-in voting ahead of this year’s general election. A longtime adviser to James, Adam Mendelsohn, told the Post that the ads were “shameless” and “reprehensible,” saying lawyers were examining the matter. James also tweeted Friday: “Not sure what we can do legally but definitely trying to figure it out!”

    What to say about Protect the Vote? Racists trying to suppress minority voter turnout? Check. Shameless Trumpian liars? Check. Stupid as a rock to go after LeBron? Definitely!

  8. joseph

    Ah, so Rick Stryker claims that Trump is not a dishonest inveterate liar. Instead he is a vary stable genius playing 11 dimensional chess with his lies to manipulate the press.

    Nope. Nothing nearly so glorious. Trump is just an insecure little child that has to lie about even the smallest things to protect his fragile ego.

    Think about the absurdity. A man is just elected president of the United States. And what is the most important thing on his mind with this awesome new power? He spends the entire first week in office lying about the size of the crowds at his inauguration.

    This is a person who is not mentally well. And a Rick Stryker who looks at this mentally unwell person and can only see genius — he must also be mentally unwell. It’s like a Jim Jones cult of blind devotion.

      1. pgl

        I do not think you even remotely understood what Michael Moore was doing here. First of all – this was about the 2020 election and not the 2016 election. But the key was the last line:
        “Fahrenheit 11/9” opens in theaters September 20.

        Moore was hyping his movie. BTW – this was September 2018 almost two years ago. Way to dust up some two year old piece where some movie director was promoting his latest film. Rick Stryker – dumber than a rock.

        1. Rick Stryker

          Of course it was about 2020. Why would I quote him about the 2016 election?

          You don’t get it though. Good.

  9. Rick Stryker


    What will it take to get to 100% uncertainty? You must be close to an unmasking since you went to all this trouble to list my alleged howlers.

    Some years ago, I woke up from a coma with no memory of my former life. I am also eager to find out who I am. Do I have tenure? Publications? Maybe I’m not an academic at all.

    1. Moses Herzog

      I would be slightly worried if you are afraid of the connection being made between your name and the blog comments. If you asked Menzie for anonymity, I think he is sportsmanlike/gentlemanly enough he would grant it. Daring him to unveil your real name when he verifies~~~ I wouldn’t be walking on that patch of ice if I were you.

      1. pgl

        The Rick’s writing style somehow reminds me of the chief Trumpian blow hard Stephen Miller. But then I would not want to accuse The Rick of being a flaming racist.

        1. Moses Herzog

          It all boils down to two things if you wanna get down to it, the first thing being Menzie’s abilities to recognize patterns in former colleagues writings or possibly former Treasury Department folks, or behaviors of people who worked with him before. I tend to think it’s not in that group because I don’t think Stryker’s econometrics or stats abilities are that high. What’s more apt to be the case is he’s a Stephen Moore type or National Review writer type. Lacking real analytical tools but feeding at the trough of Republican welfare for adulating patronage.

          The 2nd thing, would be if Rick Stryker has writings in his real name in the public forum. If he does, there is computer software available which could tag him relatively easily. How much that software costs to run I don’t know. But my guess is, if it gnawed at Menzie that much, he has the resources he could get hold of that software and run it. It becomes even easier to pin down if he has 2-3 real people names (and their public writings) in mind, and can concentrate the software on those names to get the percentage probabilities it is who he thinks it is.

          For those who care to know, Jstylo is an “open source” program. Which means if you know how to run the code, it’s FREE.

          I wonder if anyone with a PhD and a fair amount of friends, knows anyone who could cross reference suspected names through that Jstylo open source program?? If I was a sharp-minded man living around Madison Wisconsin, that’s a question I might be asking myself right about now.

  10. Rick Stryker


    In your comment above, you attempt to explore my understanding of the SIR model. I know that model well, much, much better than you do. Before you start lecturing me, I’d like you to show some basic understanding by doing the following two tasks, task that I can complete in 15 minutes. I’m not asking anything out of left field; I’ve mentioned both items in my comments, so you if you are qualified to comment on me you should be able to handle this.

    1) Estimate ‘R0 for Dane county in Wisconsin, so that Menzie can know what covid conditions are like locally. You can’t just read it from a website. Tell me what data you used and precisely how you did it. I’d like to see some code snippets.

    2) In one of my previous comments to Menzie, I used the equation I = -1/R0*ln(1-I), where I is the percentage ultimately infected and R0 is as normally defined. Derive that equation.

    These are both easy questions for anyone who is familiar with pandemic modeling. If this takes you hours to respond, I will know that you are desperately googling. If you don’t respond at all, well…..If you come back quickly with correct answers, I might start to take you seriously.

    1. Moses Herzog

      @ Rick Stryker
      Rick, I’d be lying to you, if I said, very much in spite of myself, that I didn’t find you a fascinating character study.

      For the record, I have found 2slugbaits to be a very knowledgable person, and willing to share that knowledge to benefit others when it’s of no advantage to him personally. I think that’s one of the gifts of the internet and the gifts of this blog, is commenters like 2slugbaits who are willing to extend a hand for a lift upwards towards the next block up in learning. My observations over a relatively long period is that 2slugbaits fits the highest definition of an educated man, one never too certain in his own stance (always willing to adjust to new data or a “wider data set”) and a man willing to have civil disagreement, even when not returned the same courtesy by the person he’s bandying words with. Instead of punching back to the lower gut, 2slugbaits offers his opponent an extended hand wearing a focus mitt and tells him what’s wrong with his punching technique.

      What I think of also when reading your venom here Rick Stryker, is Joe Biden’s words on “the light” and “the dark” during his DNC convention speech. Which element do you think you belong to Rick??~~”the light” or “the dark”??

      1. Rick Stryker


        What venom? 2slugs said he was going to examine my understanding of pandemic modeling, which is pretty high-handed and offensive, especially since he’s not shown any knowledge whatsoever. I merely did the same. I asked him some questions about my comments I know he can’t answer. If he knew anything about it, he should have been able to come back in 15 minutes.

        Same with Barkley. He made his patronizing comment that Menzie is using my econometric comments as examples of fallacies. I had given Barkley the benefit of the doubt, but once he made that comment I knew he was fake. He may know his narrow area, but he doesn’t know anything about econometrics.

        Maybe I’ll post the solutions to the problems this evening if I have time so that 2slugs and Barkley can learn something.

  11. Rick Stryker


    In my previous comment to 2slugs, I’ve decided to call him on his pretending to know things he doesn’t. Your comment above has inspired the same reaction in me. I think you are a #fakeconomist that doesn’t know anything about econometrics. Wisconsin has some of the finest econometricians, blah, blah, blah. Are you kidding? That’s your reason? If you are going to tell me I’m wrong, it’s time for you to stop appealing to authority and to step up and say something actually substantive, which you’ve never done. So, I have two questions for you.

    1)I said “Because of potential specification errors, cointegration is not the first choice of many time series experts” You, along with Menzie, apparently think that statement is wrong. Explain why this statement is wrong. Notice that Menzie in his “rebuttal” did not do that.

    2) In my comment on one of Menzie’s regressions, I said: “The variables are trending obviously. If there is really no relationship between them but you nonetheless estimate a linear regression, then we know asymptotically that the coefficient on Y goes to the ratio of the drifts of the series, R2 goes to 100%, and the t-statistic goes to infinity. The regression is spurious and there is no basis to decide between levels or logs and no reason to examine heteroskedasticity. It’s not a question of the regressions not being “adequate.” If there is no underlying relationship, then both regressions are nonsense and the question of which is preferred is meaningless. ”

    I want you to demonstrate mathematically that I am either right or wrong that running the regression that Menzie did would result in a spurious regression if the variables are trending with unit roots. Specifically, demonstrate that I’m either right or wrong that the estimate goes to the ratio of the drifts, R2 goes to 100, and the t-stat goes to infinity.

    This is also an easy question. If you are not a #fakeeconomist, you should be able to do it. Otherwise, don’t comment on econometrics.

    1. Barkley Rosser

      Oh dear, Rick, why are you going out of your way to make such a fool of yourself? It looks like you have been taking Moses Herzog’s comments about me here too seriously. As it is, I would hope you noticed, I have been one of those who has not indulged in speculating about your past or your day job or your true identity, regarding all of which I do not care. But if you wish to start claiming I am a “@fake economist,” well, unlike you I have an open and clear public record. My cv is readily available, so you can easily see what my over 200 publications are along with my other professional activities, and my record of Google Scholar citations is also public, although two of my most cited items are missing from it. It should be a good 1000 higher or so. I also have a Wikipedia entry that last time I checked was not too inaccurate.

      Regarding your silly demands, both are simply small scale howlers. On 1), well, it is in fact the case that cointegration is sometimes a fallback when one finds specification problems in alternative approaches. And as for 2), well, among its other infelicities I shall simply note that if there is no relationship then R squared is not going to go to 100% “asymptotically.” This is just blubbering gibberish. Please, do not attempt to debate econometrics with me further. I shall not respond, no matter how much you try to imitate Moses and indulge in childish challenges. Not worth my or anybody else’s time, although I responded this time (I shall send Dilbert Dogbert after you to throw you out of the bar if you do so).

      BTW, it appears that Howler #9 was this silly claim by you to have successfully criticized Menzie on econometric issues. You are the one who should be staying away from commenting on such matters.

      1. Rick Stryker


        That wasn’t the question on 1. I asked whether it is true or false that “Because of potential specification errors, cointegration is not the first choice of many time series experts.” You did not answer the question directly, implied that it is true in your answer, but nonetheless pronounced it a “minor howler.” My statement is true.

        On 2, you asserted that my statement is false. You provided no mathematical demonstration, as I suspected you wouldn’t. My statement is true, however.

        As I thought I made clear, I don’t care about anyone’s vita or what institution they belong to. I only care about whether what they say is true or false. You shouldn’t claim that my statements are howlers if you can’t back that up. It is you who are making a fool of himself by refusing to answer because according to you you are too important too answer. The truth is you can’t answer.

        1. Barkley Rosser


          If you had asked me, I could have helped you rewrite your two comments to make them somewhat more accurate. But if were to do so, I would be wasting everybody’s time, and Dilbert Dogbert would hit me over the head with a whiskey bottle and throw me out of the bar.

          OTOH, of course you are right that I am a very narrow “@fake.economist.” After all, if you dig through my over 200 publications you will find ones that appear in academic journals of math, physics, computer science, biology, psychology, philosophy,sociology, management, finance, and a few other disciplines. Obviously I have no ability to focus on economics at all and so clearly just a fake.

          Furthermore, I am now President of the Society for Chaos Theory in Psychology and Life Sciences, obviously a further sign of my disorientation, distraction, and complete lack of any discipline to focus on economics. About a month ago I chaired its annual conference, which was online, where aside from papers by psychologists and biologists there were ones by mathematicians, physicists, computer scientists, philosophers, organization theorists, and some other types, even a few economists.

          I even presented a paper that had some material on obscure matters of time-series econometrics, such as the relationship between ergodicity, stationarity, and homogeneity in time-series. I am sure you are aware of the related theorems by Birkhoff and von Neumann on ergodicity that assume such systems are stationary, and it has been generally accepted by most time-series econometricians that indeed all ergodic time series are stationary, even though it has long been well known that there are stationary time-series that are not ergodic, such as limit cycles. However, I discussed some special cases recently appearing in the math literature of systems that indeed are ergodic but not stationary, with these occurring in certain special infinite dimensional chaotic systems. But then, I am sure you are well aware of these developments and could lecture us about them at length and in detail, given that somehow an “@fake.economist” like me is somehow aware of them.

          1. Rick Stryker


            I didn’t need any help re-wording my statements since they are accurate as written. You said they weren’t and I challenged you to show me why, which you did not do.

            I wrote out the argument below for you why my second technical statement, which you dismissed as “blubbering gibberish,” is in fact a true statement. You still haven’t shown me why my first statement is false.

            Like I said–your list of accomplishments is irrelevant. Either you have the facts on your side or you don’t. In this case you don’t, but I would not have reacted as I did if you had not been so consistently condescending. Listen to yourself in your latest comment: “If you had asked me, I could have helped you rewrite your two comments to make them somewhat more accurate. But if were to do so, I would be wasting everybody’s time”

  12. CoRev

    I have a different candidate for perhaps the worst howler ever. It is PGL’s 2 point analysis of US daily death rate.

    August 14, 2020 at 3:21 am

    Bruce Hall tells an enormous number of lies about this virus problem but the central theme of his incessant parade of lies is that the daily death count is declining. Really?

    On August 12 – the death count jumped to over 1500 for that day alone. Of course, the death count for the next day was only 1076 so I guess to a confused little liar like Bruce Hall – this confirms his parade of lies.”

    Even his own atrocious analysis proved Bruce correct. Even funnier, later in the thread he admitted that the data showed the US daily death rate diminishing.

    Incidentally, now a short 10 days later than PGL’s August 12 data reference, the 7 day moving average from that very reference continues to confirm Bruce’s comment and not PGL’s howlingly poor analysis.

    1. Anonymous

      My 2nd HOWLER example candidate is a 2slugs comment. Several years ago he and I were debating the global Temperature data, and <b< provided a time series of Global

  13. baffling

    rick stryker, you seem to be rather testy when challenged about your part time gig as an econometrician. are you suffering from performance anxiety at your day job as the porn star? i still find it fascinating that rick stryker, who tries to cover all his bases like a lawyer, overlooked the fact that he named himself after a gay porn star. it simply had to be a freudian slip. no worries rick, we welcome folks from all backgrounds onto this blog. but the irony still amazes me.

  14. Rick Stryker

    Barkley Rosser refused to support his statement that “well, among its other infelicities I shall simply note that if there is no relationship then R squared is not going to go to 100% “asymptotically.” This is just blubbering gibberish. Please, do not attempt to debate econometrics with me further.” Apparently his vita speaks for itself and he has no need to defend his ex cathedra pronouncements. I see no need to debate either since Barkley knows nothing about econometrics. But I’ll give Barkley a free econometrics lesson. But if he wants more, he’ll need to sign up for my free market econometrics class at Wossamotta U.

    Note: If you don’t want to go through this argument, I would suggest you can prove to yourself that R^2 goes to 1 by generating longer and longer realizations of two unrelated series with a unit root and a drift. Run OLS on each realization and you’ll see the results I have claimed. You can do the same for the other results I’ve claimed.

    Proceeding more formally, suppose we have two trending series x(t) and y(t) that both have unit roots. These series are not related to each other. So we can write:

    Y(t) = a + Y(t-1) + e(t)
    X(t) = b + X(t-1) + u(t)

    e(t) and u(t) are innovations.

    Let’s suppose we regress these two unrelated series on each other:

    Y(t) = alpha + BetaX(t) + mu(t) for t = 1 to T

    What are the properties of R^2 as T goes to infinity?

    We can re-write the two series as:
    Y(t) = at +sum(e(i)) + Y(0)
    X(t) = bt + sum(u(i)) + X(0)

    In this notation, sum(e(i)) = sum of the innovations e(i) from i = 1 to t

    We can also write these processes as deviations:
    y(t) = at + sum(e(i)) + Y(0) – sum(Y(i))/t
    x(t) = bt + sum(u(i)) + X(0) – sum(X(i))/t

    R^2 = [sum(x(t)y(t))]^2/[sum(x(t)^2)*sum(y(t)^2)]

    In the expression for R^2, the sums go from t =1 to T

    The key is to multiply the numerator and denominator out, using the fact that sum(t^2) = (T^3)/3 + (T^2)/2 + T/6
    If we multiply the numerator out, we’ll get a^2*b^2(T^6)/9 + lower order terms. Similarly, if we multiply out the denominator, we will also get a^2*b^2(T^6)/9 plus lower order terms.

    Divide the numerator and denominator in the expression for R^2 by T^6 and take the limit as T goes to infinity. All the terms go away except a^2*b^2/9 in the numerator and a^2*b^2/9 in the denominator.

    Therefore, R^2 goes to 1 asymptotically as T goes to infinity.

    Using the same technique, we can show that asymptotically the OLS slope estimate beta_hat goes to a/b just as I asserted.

    1. Barkley Rosser

      OK, Rick, I am going to be in danger of being hit over the head by D.D. with a whiskey bottle, but I shall make this one and only reply to your long effort to try to prove your statement #2 is correct. I simply note that here you imposed the condition that the series have unit roots, but that was not in your original statement. That is all I am going to say about this, other than to note that it is not just Donald J. Trump who lies about what he says.

      1. Rick Stryker


        No, the unit root condition was in my statement To quote from my statement above:

        “I want you to demonstrate mathematically that I am either right or wrong that running the regression that Menzie did would result in a spurious regression if the variables are trending with unit roots. Specifically, demonstrate that I’m either right or wrong that the estimate goes to the ratio of the drifts, R2 goes to 100, and the t-stat goes to infinity.

        I was very clear and specific. And I formulated the problem that way to apply to Menzie’s regression case in his post, since C and Y are typically found to be integrated of order 1. Your response was “And as for 2), well, among its other infelicities I shall simply note that if there is no relationship then R squared is not going to go to 100% “asymptotically.” This is just blubbering gibberish. Please, do not attempt to debate econometrics with me further.”

        As I have demonstrated, you were completely wrong. You are not qualified to comment on econometrics.

        1. Barkley Rosser

          No, Rick, you are lying. The specific comment you wanted me to comment on says nothing about unit roots. I realize you are desperate to somehow get at Menzie by somehow trying to score some kind of points on me. But just outright lying is not going to do it. All it will do is emphasize that you are indeed a true follower of the massive liar, Donald J. Trump.

          1. Rick Stryker

            How can I be lying Barkley when I linked to my comment, quoted myself, and then bolded where I specified the unit root condition. And of course, the unit root condition makes sense in the context of the regression that Menzie was running in which both variables are typically to be I(1). I asked you very clearly and specifically if my statement was false under those relevant conditions. You said my claim was “blubbering gibberish.”

            Calling me a liar when you are so clearly wrong adds to the evidence that you are a #fakeeconomist.

            But I’ll give you a chance to redeem yourself. Let’s suppose the variables in Menzie’s regression did not have unit roots, but were in fact trend-stationary instead. Are you saying that my statement is false in that case (even though that’s not a very relevant case to look at in the context of Menzie’s regression)?

          2. Barkley Rosser

            I apologize for that, Rick. It was written before I found your “unit roots” words. I had been focusing on your gibberish preceding paragraph, which I have commented on below.

            When I mess up, which I do plenty, I quickly admit it and apologize. Moses H. has denounced me at length for my incompetence at linking to sources, as well as for my misspellings, mostly due to just not proofreading before posting, although reflecting plain ignorance in the case of “McKinnon.” He also accuses me of many things that are baloney, but when I goof, I readily admit it. These blunders on this blog do not stop people from citing my research frequently, and my failure to see your “unit roots” and my unfortunate comments following that, will not stop people from citing my research on econometrics, even if you think somehow nobody will ever take anything I have to say about it seriousl ever again.

            The problem for you, Rick, is that Menzie has completely nailed you with these 8 Howlers, with you also guilty of #9, despite your half-baked efforts to claim you actually have successfully critiqued his work or remarks here, and your pathetic efforts to defend yourself on some of them have gone exactly nowhere. So you have thrown up dust by trying to get into fights about basic econometrics with me and 2slugbaits, but this will not work as there is simply nothing there. I suggest you emulate me and just admit it when you are caught being wrong, but, no, I suspect that like CoRev and a few others here, this is beyond you. Too bad.

  15. Rick Stryker

    I gave 2slugs some time to see if he could google his way into verifying my assertion that in the standard SIR model, we can estimate the percentage of people ultimately infected using the relation:

    i* = -(1/R0)*ln(1-i*)

    where i* is the percent ultimately infected. I didn’t think he could do it since I haven’t been able to find the derivation anywhere although this equation is sometimes quoted. I’ll now give 2slugs a free mathematical epidemiology lesson. But if he wants more, he’ll need to sign up for my free market epidemiology class at Wossamotta U.

    Let S(t) be the number of susceptible people at time t. S(0) is the original population. I(t) is the number of people infected at time t. The two governing differential equations are:

    S'(t) = – beta*I(t)*S(t)
    I'(t) = beta*I(t)*S(t) – alpha*I(t)

    S'(t) and I'(t) are the time derivatives

    Dividing the first equation into the second, we get (dropping the t dependency since I don’t feel like typing it)

    I’/S’ = (beta*S*I -alpha*I)/(-beta*I*S) = -1 + alpha/(beta*S)

    Now we can separate the variables:

    I’ = (-1 + alpha/(beta*S))S’

    Integrating, we obtain

    I = -S + (alpha/beta)*lnS + C where C is an arbitrary constant

    We know that the limit as t goes to infinity of I(t) = 0 while the limit as t goes to infinity of S is some number S_inf. Thus we may write

    I(0) +S(0) -(alpha/beta)*ln(S(0)) = C


    I(0) + S(0) – (alpha/beta)*ln(S(0)) = S_inf -(alpha/beta)*ln(S_inf)

    Now we can re-arrange the above as

    beta/alpha = ln(S(0)/S_inf)/(S(0) + I(0) – S_inf). Since

    R0 = S(0)*beta/alpha, multiplying both sides by S(0) we get

    R0 = ln(S(0)/S_inf)/(1 + i(0) – s_inf)


    i(0) = I(0)/S(0), which is very close to zero
    s_inf = S_inf/S(0)


    i* = s(0) – s_inf. If we ignore i(0) since it is close to zero, we have

    R0 = ln(S(0)/S_inf)/i*

    We can write ln(S(0)/S_inf) as -ln(S_inf/S(0)) = -ln[(S(0) – I_inf)/S(0)] where I_inf is the total number infected. Since i* = I_inf/S(0), we have

    R0 = -ln(1 – i*)/i* or

    i* = -(1/R0)I*ln(1 – i*)

    That concludes this evening’s lesson. If 2slugs were to ask politely, I will show him how to estimate current R0 for Dane county.

    1. joseph

      It’s like Trump needing to brag that he passed a dementia test — PERSON WOMAN MAN CAMERA TV.

      You know, Rick, just like Trump, your attempts to prove that you aren’t crazy just make you sound even more crazy.

    2. 2slugbaits

      Rick Stryker I can run EpiEstim too, but I really don’t care about Dane County’s Rt. Pointless in any event since students are returning and it would be OBE in a heartbeat. FWIW, rtlive shows Wisconsin’s Rt leveling at 1.13, but that will likely change as students return. As to your proof, you simply copied what was already out there. The transformation uses an e* in the Jones paper that you linked to several months ago rather than an “i*”. Yes, I read it. The derivation for the “i” appears as appendices in several papers by Hetcote1976 and 1989 and again in a 2000 paper referenced in the Jones paper. Your log version was also referenced in an rtlive note several months ago.

      Also, in your original post you cited this equation:
      I = (-1/R0)*ln(1-I)

      which you repeated in your “challenge”:
      I used the equation I = -1/R0*ln(1-I),

      However, in your latest post you cited this equation:
      i* = -(1/R0)I*ln(1 – i*)

      See the error?

      But as interesting as all that is, it’s not really the point of the discussion. We’ve been discussing a couple of issues. First, I pointed out that R0 and Rt are not the same thing. The CDC worries that a lot of models written by mathematicians are confusing the two:

      Inconsistency in the name and definition of R0 has potentially been a cause for misunderstanding the meaning of R0. R0 was originally called the basic case reproduction rate when George MacDonald introduced the concept into the epidemiology literature in the 1950s (17,19,24,25). Although MacDonald used Z0 to represent the metric, the current symbolic representation (R0) appears to have remained largely consistent since that time. However, multiple variations of the name for the concept expressed by R0 have been used in the scientific literature, including the use of basic and case as the first word in the term, reproduction and reproductive for the second word, and number, ratio, and rate for the final part of the term (13). Although the frequent use of the term basic reproduction rate is in line with MacDonald’s original terminology (9), some users interpret the use of the word rate as suggesting a quantity having a unit with a per-time dimension (7). If R0 were a rate involving time, the metric would provide information about how quickly an epidemic will spread through a population. But R0 does not indicate whether new cases will occur within 24 hours after the initial case or months later, just as R0 does not indicate whether the disease produced by the infection is severe. Instead, R0 is most accurately described in terms of cases per case (7,13). Calling R0 a rate rather than a number or ratio might create some undue confusion about what the value represents.

      If I’m understanding you correctly, in your example you concluded that herd immunity would be achieved at 37% with an R0 of 1.25. It’s not clear what other parameters you used. When I use an R0 of 1.25 I get a herd immunity of 20%. But that just means it’s the point where the rate of new infections slows down. In the basic SIR model eventually everyone becomes infected if the R0 is greater than 1.00.

      But here’s the crux of the matter. We can all agree that behavior changes in response to a virus. That will tend to lower the effective transmission Rt. And we see that in the data. The rtlive site is good at showing this. What we see is a wave pattern in just about every state. Things are bad, then people (or governments) start reacting and the Rt falls. Then people get careless and the Rt climbs. This is a slow walk to herd immunity in the sense that sooner or later everyone will become infected. It’s also very inefficient. A hard shutdown could have given us the chance to extinguish the virus, but this oscillating behavior guarantees that the economy will go through several partial shutdown cycles for a very long time. It doesn’t matter whether a partial shutdown is directed by governments or if it’s self-directed by people refusing to shop or refusing to go to work. The net result is the same level of economic damage. Making COVID-19 a permanent part of life forever after is not the kind of return to normal that people want. We had an opportunity to extinguish COVID-19, but it seems that keeping bars open was more important.

      Finally, it’s beginning to look like the basic SIR model is inappropriate. This morning we learned of the first positive case of someone having been reinfected. That means something like an SIS model (which describes the common cold) might be more appropriate. It also means we’ll likely need semi-annual vaccines.

      1. baffling

        “This morning we learned of the first positive case of someone having been reinfected. That means something like an SIS model (which describes the common cold) might be more appropriate. It also means we’ll likely need semi-annual vaccines.”
        this is also consistent with the other known coronaviruses that contribute to respiratory infection and the common cold. it means covid19 remains troublesome into the future.

        1. CoRev

          Baffled, link please. Some questions that need answers: 1) How diagnosed (Tested or hospitalized both occurrences), 2) Hospitalized 2nd occurrence, 3) False positive (test and/or diagnosis), 4) How contracted in each incidence, 5) Same version of Covid-19 or mutated, 6) If mutated what is different?

      2. Rick Stryker


        No you can’t run epiestim or anything else, not without great difficulty and a learning curve, otherwise you would have done it. You had plenty of time if you already knew how to do it and were conversant with pandemic modeling.

        Moreover, I gave you a lot of time to google to see if you could find out how to derive that relationship. Why didn’t you come back quickly with the derivation and a reference? I don’t need to Google as I derived it myself when I saw it referenced. But I wanted to see if you could at least Google, find it, and then understand it well enough to spit it back. Obviously you weren’t able to do that.

        You say you found the derivation on the internet. The only reference I found was Heathcot (2000), which doesn’t contain a complete proof and does something different from what I did. If you found something, link to it. I link to everything. Why can’t you?

        1. 2slugbaits

          Rick Stryker I have run EpEstim. I’m the one who provided you with the reference, remember? And it’s not difficult. The pain-in-the-ass part is dumping the county level data.

          The only reference I found was Heathcot (2000)

          I think you mean Hethcote.

          which doesn’t contain a complete proof and does something different from what I did

          Reread my comment. I didn’t say it was in the 2000 paper. I said the proofs were referenced in the 2000 paper.

          You say you found the derivation on the internet.

          Where did I say the papers were available online?

          Completely false.

          Completely true. An R0 greater than 1.00 always means the epidemic is growing…at least with any plausible parameters. And by “plausible” I mean a “nu” parameter that doesn’t remove the victim almost immediately. The only way you suppress the virus is to get the Rt below 1.00. Are there variants of the basic model that capture behavioral changes such that Rt falls below 1.00 and the virus stops growing? Sure. But not in the basic plain vanilla SIR model. In the basic SIR model the contact parameter never changes, so there is nothing to stop the spread, except of a very, very swift death. The fact that the contact parameter never changes and is therefore very unrealistic is one of the many major problems with the basic plain vanilla SIR model. Here’s about as basic plain vanilla as it gets:

          pars <- c("beta"=.010, "nu"=.1)
          times <- seq(0,100,1)
          y0 <- c(100,1,0)
          sir <- function(t,y,p) {
          yd1 <- -p["beta"] * y[1]*y[2]
          + yd2 <- p["beta"] * y[1]*y[2] – p["nu"]*y[2]
          + yd3 <- p["nu"]*y[2]
          + list(c(yd1,yd2,yd3),c(N=sum(y)))
          + }
          sir.out <- lsoda(y0,times,sir,pars)
          + ylab="Compartment Size")

          Copy and paste it, run it. Show us were the Susceptible compartment doesn't eventually asymptote to zero. And it matches reality. Can you modify it to include the effects of policy interventions on "beta"? Sure. But then it's not the basic plain vanilla SIR model.

          Don't want to explain how you came up with that 37% value?

          1. Rick Stryker


            You did not tell me about epiestim. I’ve been using it for months now. It’s what I used to provide the 1.02 estimate for Menzie. You mentioned the R software after I did that estimate, not before.

            I do not believe that you can do this estimate using epiestim. Getting and loading county data is very easy if you know what you are doing so that’s hardly a reason not to do it. Like I said, I’ll show you how to do it if you ask politely.

            You said I copied my i* proof from references. Specifically you said: “As to your proof, you simply copied what was already out there.” How do you know that? Did you read those papers you referenced. If they are not online, did you go to the library to get them? I didn’t copy my proof from anywhere. I gave you that question specifically because I didn’t think you could find it in any reference. I never saw it derived in any reference.

            I know you didn’t find the derivation in any reference because you don’t understand the derivation even though I spelled it out for you in detail. One of the fundamental discoveries of mathematical epidemiology is that not everyone gets infected when R0 >1. Your statement that everyone gets infected if R0 > 1 is FALSE. Also, if you understood the derivation of the i* equation, you’d know where the 37% came from. It’s not herd immunity. That’s a different concept.

            I’m happy to explain all this to you if you are genuinely curious.

          2. 2slugbaits

            Rick Stryker I don’t recall your ever recommending EpiEstim, but I’ve had it and used it from time-to-time for a couple months.

            As to the 37% figure, here’s what you said:

            If we assume, counter to experience, that R0 goes up and stays up so that on average R0 = 1.25 over the life of the pandemic, then we can solve that equation for the per cent ever infected. We get I = 37%.

            Assuming an infection fatality rate (IFR) of 0.6%, we can estimate the total number of fatalities then to be .37 X 330 X 0.6 = 735K, about 1/3 of 1918. I used an IFR of 0.6% for that calculation, but I believe that IFR is too high. I think it’s really 0.3%. If I’m right, then we’d get a total death toll of 368K.

            In other words, you calculated 37% of the population would become infected. Based on the English language I understood you to claim that 37% was the upper limit that would become infected. I was asking about what parameters you used to come up with 37%.

      3. Rick Stryker

        “If I’m understanding you correctly, in your example you concluded that herd immunity would be achieved at 37% with an R0 of 1.25.” Nope, that’s not herd immunity. Go back and look at the derivation I provided for you.

        “In the basic SIR model eventually everyone becomes infected if the R0 is greater than 1.00.” Completely false.

  16. baffling

    and now kellyanne conway bails on the white house. the ship is sinking and the rats are jumping.

  17. baffling

    “Using the same technique, we can show that asymptotically the OLS slope estimate beta_hat goes to a/b just as I asserted.”
    is this something you have published rick, or have you “asserted” somebody else’s work? that would be plagarism you know. copying somebody else and claiming it is your knowledge. just curious.

    1. Barkley Rosser

      This is true when properly stated, which RS was sloppy about to the point of it becoming blubbering gibberish, as I stated. However, baffling, you are right that the basic result is trivial and not remotely new, although RS did not claim to be showing anything new.

      An easy way to see what lies behind this is to go backwards from trivial math. So, if one draws two lines, then the rate at which they converge or diverge, depending on which way one is going, is determined by the ratio of their slopes. Needless to say, if when one is running regression estimates of those two lines, and the conditions hold that were not stated b RS, then there will be a tendency for the estimates to converge on the “true” lines, with those slopes associated with the appropriate regression coefficients, so as the estimates approach the true lines, again assuming the various conditions hold that allow for this (easily found in any elementary textbook), the estimate of this coefficient that essentially measures that rate of divergence or convergence of the lines will approach a value equal to the ratio given.

      Given that probably he has induced some confusion on the part of some people, I should probably help clarify further on the other matter. The true element in the earlier part of his garbled second statement is that in general, again under certain unstated by him conditions, there is a tendency for R squared to rise for a given regression estimate as more observations are made and used in the regression, which is the point implied by the overly strongly and unclearly atate result he claimed. I have already noted that he left out the unit root matter in the initial statement he made that he then dragged in later.

      Another crucial confusion is that when he piled on with his equations he introduced as the meaning of “asymptotic” havint time go to infinity. But in fact what is involved is the number of observations going to infinity, the result here being more general than simply time series regressions, even if he again left out various important assumptions for this to hold. Indeed,it is perfectly possible to have a time series where even though time may be gong to infinity, before one gets to infinity the series may simply stoply stop having observations, the time series end, in which case taking time to infinity does not take R squared to 1.

      Even if the number of observations goes to infinity as time does so, there are still glitches that can arise that RS was too careless to rule out. So, it is useful to keep in mind that R squared is basically total variance over explained variance. If either of those goes to either zero or to infinity as the number of observations goes to infinity, then RS’s equations do not work out so well, which can happen. But then, this is the sort of annoying thing that an “@fake.economist” who knows no econometrics would point out, quite aside from some deeper issues about how to do the underlying statistics at all, matters I was just writing about in the book I am currently writing for Springer, although undoubtedly only things in the very narrow area RS recognizes that I might know something about might be correct in this book, senile as I am, as our good Moses has clearly established, although I am sort of curious what RS thinks is this “narrow” area of economics that I might know something about. I am not sure what it is, given that I spend too much time writing papers for journals not even in economics at all and barely even know how to walk out a door here in my obscure corner of Virginia where I lurk.

      1. baffling

        “I gave 2slugs some time to see if he could google his way into verifying my assertion that …”
        barkley, he certainly did assert this was his result. i agree, it is not really plagiarism. and if rick wants to google a topic and then pontificate on a blog, i guess that is his right. but as you said, he was a bit sloppy in his normally meticulous preparation. i guess that day job as a porn star is starting to impact his moonlighting as a econometrician.

        1. Rick Stryker


          I didn’t Google a result and then pontificate. I chose the equation I did just because I mentioned it before and I didn’t think 2slugs would be able to find it by googling. I derived it myself. If 2slugs is an knowledgeable as he says, and wants to critique my knowledge of pandemic modeling, then he should have been able to derive it too. As you saw, he failed, as I knew he would.

          You guys on the Left think you are much, much smarter than you really are. This conceit is much of the reason that you don’t think you need to justify your beliefs. You think it’s enough to mock or dismiss people you don’t agree with, because you imagine you are so much smarter and superior morally. Or, if you are like Barkley, you condescend and hrumph about your many accomplishments as a justification for not having to defend your insulting comments. Menzie brought up alleged “howlers,” which exemplifies perfectly this Liberal conceit. So, I thought this might be a good time to teach a few of you a lesson in intellectual humility.

          You, on the other hand, are hopeless Baffling. You don’t know enough for me to teach you any lesson.

      2. Rick Stryker


        I wasn’t sloppy at all. I was very clear in my statement to include the unit root condition. Specifically, I said:

        “I want you to demonstrate mathematically that I am either right or wrong that running the regression that Menzie did would result in a spurious regression if the variables are trending with unit roots. Specifically, demonstrate that I’m either right or wrong that the estimate goes to the ratio of the drifts, R2 goes to 100, and the t-stat goes to infinity.

        I was very clear and specific. And I formulated the problem that way to apply to Menzie’s regression case in his post, since C and Y are typically found to be integrated of order 1. Your response was “And as for 2), well, among its other infelicities I shall simply note that if there is no relationship then R squared is not going to go to 100% “asymptotically.” This is just blubbering gibberish. Please, do not attempt to debate econometrics with me further.”

        As I have demonstrated, you were completely wrong. You are now pretending that this is all trivial and I wasn’t specific enough. But I was. And if this is all so trivial, why didn’t you say something like, “Well, that may be true if there is a unit root but…” You didn’t say that because you didn’t know it until I gave you a free econometrics lesson. and you are now pretending to have known it all along.

        You should stop commenting on econometrics so as not to embarrass yourself further.

        1. Barkley Rosser

          OK, Rick, I did not see your “unit roots.” I was focusing on the paragraph before it, which is still a pile of blubbering gibberish. There remains a basic problem with your claim. It is irrelevant to anything Menzie said or presented, who is your target with all this, trying to overcome him nailing you with your various howlers. It gets to this: none of what he has looked at is in an asymptotic condition with infinite observations. So what holds true in that condition is simply irrelevant to what holds when not in that condition.

          I bungled on the matter of claiming you did not mention unit roots when you did, but your demonstration does not demonstrate anything that undercuts what Menzie argued. You are still stuck with your revealed howlers. Sorry.

          1. Rick Stryker

            Well, I guess you saw my other comment and belatedly realized that my statement is also true in the case that the time series are trend–stationary. It’s not about unit roots necessarily. You learned something at least, but that realization has forced you to shift your argument to be that my statement is not false but rather irrelevant, and that’s your new reason it is gibberish. Ironically, you used some gibberish about asymptotics to justify your shift in position.

          2. Barkley Rosser


            I stayed away from commenting on that later part of that first paragraph in your second comment other than to describe it as “blubbering gibberish,” and that is what it is. It is just plain false, not only that, it is stupid. The regressions are not spuriouis at all and indeed it is useful to look at heteroskedasticity. You have not scored anything on Menzie at all with this worthless garbaged.

            Oh, and writing “R2” is silly, althoug trivial. Of course, after i put it as “R squared you then accurately put it as “R^2,” when you started spouting your irrelevant equations from some elementary textbook.

            Again, Menzie ia dealing with finite data sets whete things that hold asymptotically are simply irrelevant. Sorry, I messed up by not seeing “unit roots” in your next paragraph, but this point simply does not get you off the hook of having failed to find fault with Menzie’s economstrics, Howler #9 for you.

    2. Rick Stryker

      I have asserted a fact that is well-known among people competent in econometrics Baffling.

      1. baffling

        and here i thought you were going to teach me something new, that i could not learn in a textbook. i had high hopes that you would produce something innovative rick, showing that you don’t simply have a warped mind but a creative one as well. i am sorely disappointed that all that was necessary was a little google search for somebody else’s contributions.

  18. joseph

    and now kellyanne conway bails on the white house.

    She says it is to spend more time with her family.

    No word yet from her family whether they would like to spend more time with her.

    1. Moses Herzog

      Kellyanne’s kids phoned the White House and said they were afraid ICE agents were going to kidnap them and put them in a cage. She had to calm them down by reminding them they were white and that only the “dirty” people from below the tilting from erosion southern wall Kellyanne affectionately refers to as “Bannon’s gimpy cash register” have their children stolen from them and put into cages.

      Kellyanne has told various journalists she’s been inspired by the fears of her own kids to do a series of children’s books, current working title “Don’t Worry White Kids, Only the ‘Dirty’ Kids Are Kidnapped by MAGA and Put in Cages”. She plans on giving the proceeds to the local Stormfront chapter in Atco New Jersey.

  19. Rick Stryker


    Thanks for estimating R for Dane County. Part of the reason I asked for an estimation of R for Dane county is that my estimates are very different from other sites like rtlive. My estimate as of yesterday for Dane was 2.37 and it was a little over 2 yesterday. You reported you got 0.923. That’s quite a difference. I think we need to reconcile as I’d like to know if those high numbers I’m getting are legitimate or there is a problem somewhere.

    Have the students come back to Madison? An R0 over 2 is not good if true.

    1) What method did you use in epiestim?
    2) What dates did you use for the estimation? What data source did you use?
    3) Can you give a few data points on a few selected dates so I can compare with my data?

    On your question about 37%, it comes from solving the equation i* = -(1/R0)*ln(1-i*). With an R0 of 1.25, i* = 0.37. That’s different from herd immunity which is estimated as 1 – 1/R0 = 0.2

      1. Rick Stryker


        Thanks Menzie. I noticed that R in Dane county has been rising over the last week and I thought students returning might have been related to that. Could be my estimate is just wrong. But seems a bit alarming if true. Is the University testing the students, quarantining, etc?

        1. Menzie Chinn Post author

          Rick Stryker: The university has indicated it will provide testing, and has requirements for limiting interactions within dorm rooms. Whether this will prove successful remains to be seen; my priors, given observed student behavior, is that it will be difficult to keep hybrid operations going – especially against a backdrop of moderately high cases. But the test will be when classes begin, next week.

        2. Moses Herzog

          It goes without saying Menzie can tell you better than me. This would give you a rough outline:

          Rick, as you might imagine a dirty liberal saying, I think students returning to school is highly inadvisable (with the possible exception of offering physical classes for international students so as to avoid being deported for not attending physical classes, which I think trump changed his mind on when he realized he was the only dumb ass in favor of this).

          That being said, as much as one can get college age to control their hormones, I am sure UW-Madison students will carry themselves well.

  20. Rick Stryker


    I didn’t say the regression was spurious. I was commenting on the general methodology used. When you are doing a regression with trending variables, if they are not related, they will appear to be related because of the reasons I previously mentioned. But the regression will be spurious. On the other hand, if the trending variables are related, then the regression estimates the cointegrating vector. Those estimates are consistent (super-consistent actually). But hypothesis tests are in general invalid (except for a few special cases) and there is no point reporting HAC standard errors, etc, etc. If you want to test hypotheses about the estimates of the cointegrating vector, then you’ll need some special procedures. In either case, spurious or cointegrating, you can’t do conventional hypothesis tests. That was my point and it’s a valid one.

    1. Barkley Rosser

      Look, Rick, I have admitted I messed up by failing to see that you said “unit roots.” Now you say, “I didn’t say the regression was spurious.” But I have just double checked. You did say it. You are now just out and out lying.

      I realize you have become desperate because you have failed to overturn a single one of Menzie’s examinations of your Howlers. Really, it is time to stop squirming and just accept that you have been exposed. Outright lying simply shows what a follower of that liar of liar Trump, and your bizarre effort to claim he never lied remains probably the wildest and most extreme of all your lies.

      1. Rick Stryker


        Once again, you accuse me of lying but you don’t provide any link. If you think I said the regression was spurious, then provide a link. And make sure you provide the entire quote, not just some cherry-picked part. As I said, the variables that Menzie used are typically found to be I(1), which is why I asked you to demonstrate why I was wrong in the case that the variables had unit roots with trends. Given that C an Y are likely I(1), it wouldn’t have been a spurious regression in the particular case he was running. But I was making a general point about being extra careful when running a regression on trending variables.

        By the way, I use ecconbrowser to get material for my own classes. The post we are talking about I used as an example at Wossamotta U of an econometric fallacy to watch out for. I didn’t point them to econbrowser directly but rather just used Menzie’s example. I’ve used quite a few examples from this blog, from both Menzie and some of the commenters. I wouldn’t point my students to Econbrowser however since one of them, afflicted with the liberal disease, might be tempted to email Menzie about my identity. I am using you as an example of an economics professor that doesn’t know anything about econometrics. The students think all economists are competent but they need to realize that’s not the case.

        As you may have noticed, I have not tried to argue against Menzie’s alleged fallacies. Instead, I have opted to teach a few of you a much-needed lesson in intellectual humility. You think because you are progressive that must mean that you are smarter and better than other people. I’m here to disabuse you of that nonsense.

        1. Barkley Rosser

          Rick and anybody who is paying attention and thinks Rick is not lying about this,

          Go to your comment on Aug. 22 at 11:05 AM, paragraph 3, lines 3-4, the following sentence appears, which was the basis for my describing what you wrote as “blubbering gibberish.”

          “The regression is spurious as there is no basis to decide between the levels or logs and no reason to examine heteroskedasticity.”

          1. Rick Stryker


            You ignored my admonition: ” And make sure you provide the entire quote, not just some cherry-picked part.” Here is the whole quote, where I bold some important words:

            ” Before you can even begin to make this choice, you’ve first got to insure that you have a valid regression. The variables are trending obviously. If there is really no relationship between them but you nonetheless estimate a linear regression, then we know asymptotically that the coefficient on Y goes to the ratio of the drifts of the series, R2 goes to 100%, and the t-statistic goes to infinity. The regression is spurious and there is no basis to decide between levels or logs and no reason to examine heteroskedasticity. It’s not a question of the regressions not being “adequate.” If there is no underlying relationship, then both regressions are nonsense and the question of which is preferred is meaningless.

            Moreover, if there really is a relationship and those regressions are cointegrating, then hypothesis tests such as t stats are invalid anyway, except in special cases.

            Just as I been saying over and over, I was saying that whether there is or isn’t a relationship between the trending variables, you can’t use conventional hypothesis tests. That’s my very valid point.

            But notice how Barkley attempted to misquote me out of context. And notice how he didn’t provide a link so that readers could go back and see the whole context.

            And he accuses me of lying?

          2. Barkley Rosser


            Sorry, but I have really had it with your lying. A few comments above, today at 11:57 AM you said in the first line, “If you think I said the regression was spurious, then provide a link.” I did so.

            This followed yesterday at 3:04 PM which comment began with “I didn’t say the regression was spurious.” But you did, which is utterly clear. I am definitely not going to comment further on your lies.

            Although now I am curious. You are suggesting that you actually teach. Do you actually publish? How many GS citations do your publications have? I think you can tell us that without outing your precious identity.

  21. Rick Stryker


    This comment of yours “Oh, and writing “R2” is silly, althoug trivial. Of course, after i put it as “R squared you then accurately put it as “R^2,” when you started spouting your irrelevant equations from some elementary textbook.” shows again your general ignorance of econometrics.

    Because you don’t know the subject well at all, you just assumed that my statements about regressions on trending variables were false. Once I disabused you of that, you now think I must have copied them from “some elementary textbook.” That’s not a very good defense. If my point is widely discussed in elementary textbooks, what does that say about you for not knowing it?

    In fact, my claims and derivation are not in any elementary textbook I know of. My statement and derivation are not in Stock and Watson. They are not in Wooldridge. They are not in Johnston. They are not in Judge et al. If you think my statement and the derivations are in some elementary textbook, then tell me what book.

    I never saw my statement and derivation about regressions with trending variables in any text book or anywhere really. I am sure they are probably somewhere though.

    1. Barkley Rosser

      Some of your statements are correct. Some of them are not correct. I hope I have made clear which ones are and which ones are not. Again, what is true fails to establish any meaningful critique of anything Menzie has said here. And now I am really quite tired of you endlessly rambling and lying here. You are doing too much of an imitation of CoRev now, who is also notorious for just going on and on with yet more and more nonsense as his arguments keep getting shot down.

      1. Rick Stryker


        Ironically enough, you again accuse me of lying, even though you claimed that I copied my derivation from an elementary textbook. I asked you to tell me what textbook, and you now change the subject.

        I never heard of the community college at which you teach, but I thought perhaps that my derivation might have been in a textbook in your department, so I checked the so-called “advanced econometrics” course and the textbook was Stock and Watson, which doesn’t have my points or the derivation. I’m beginning to think you just made that up about the elementary textbook.

        Speaking of elementary, I noticed at your community college that there is a course offered called Fin 380, “Elemental and Derivative Securities.” That must be a typo—Elementary and Derivative Securities would make more sense. You might want to let them know.

        You have plenty of practical programs at your community college, which I like. However, where are the programs in diesel mechanics, welding, etc? I didn’t see them but perhaps I missed it. People still do need training in those fields.

        1. Barkley Rosser

          No, Rick, the lie was not in your trivial “derivation,” it was in a sentence demanding that I perform derviations here. As it is, you have yourself above reproduced the actual sentence you denied you wrote. I think you think that surrounding the sentence with a lot of emboldened words somehow fixes it, but it does not. The whole pile remains blubbering gibberish.

          This is just another example of you doing a bad imitation of Trump. First you lie, “I did not say that!” Then when you are caught in your lie, you go “Well, it is OK given all of the things I said around it!” Gag.

  22. Rick Stryker

    I asked 2slugbaits the following questions about his estimate of R in Dane County:

    1) What method did you use in epiestim?
    2) What dates did you use for the estimation? What data source did you use?
    3) Can you give a few data points on a few selected dates so I can compare with my data?

    I selected Dane County to estimate R0 since I knew that when I did the estimate I got a number very different from rtlive, using the method I was using at least. A number close to rtlive, which is publicly available on a website, would be suspicious. And wouldn’t you know it, 2slugs, after days of prodding from me, comes back with an estimate suspiciously close to rtlive. And when I ask him how he got his estimate, I get silence. Just silence. Makes you wonder doesn’t it…..?

    All I need is the data set he used and the method he used.

  23. joseph

    Rick Stryker is still hung up on trying to prove he isn’t demented by doing arithmetic. Just like Trump trying to prove he isn’t demented by repeating over and over to a journalist PERSON WOMAN MAN CAMERA TV.

    They say, ‘Nobody gets it in order,'” he said, referring to the doctors. “It’s actually not that easy, but for me it was easy. And that’s not an easy question.” “They say, ‘That’s amazing,'” he added. “‘How did you do that?’
    Here’s Trump:

    Sounds just like Stryker. It’s really sad.

    You don’t have to play his game. Stryker revealed everything you need to know about him when he said this: “Trump is the most honest politician we’ve seen or are likely to see in our lifetimes”.

    If there is no such thing as truth, there’s no point arguing about elementary arithmetic.

    1. Rick Stryker


      Your comment reminds me of the old adage that to a man with a hammer every problem looks like a nail. And to a man who understands nothing more than arithmetic, everything quantitative looks like arithmetic.

      We’re not talking about arithmetic here. This discussion is way over your head. (And apparently way over Barkley’s head judging from his last comment.) Best to confine your comments to sizes of crowds and other counting, adding, and subtracting matters.

  24. baffling

    “If 2slugs is an knowledgeable as he says, and wants to critique my knowledge of pandemic modeling, then he should have been able to derive it too.”
    actually rick stryker, based on your comments on this blog, you have zero credibility in the area of pandemic modeling. a while back you came on the blog and criticized medical professionals for their pandemic models-something they do for a living-after playing with the models for about a month. but i have yet to see any credibility from you in the area of pandemic modeling. do you have any publications in the area of pandemic modeling? do you serve as a resource to government agencies in the area of pandemic modeling? does the press ask you for comment regarding pandemic modeling? have you ever received a grant for pandemic modeling? besides picking up some reference material a couple of months ago and playing with it on the computer, exactly what contribution have you made to the field of pandemic modeling that gives you any credibility rick stryker?
    rick, you have as much credibility in pandemic modeling as corev has in global climate change. you need to return to your day job as a gay porn star, and stop pretending to have expertise in areas in which you have no credibility. or show me the evidence.

    1. Rick Stryker


      Well, I did derive that equation I used to size the final size of the epidemic. And I also used a standard epidemiological method to estimate R for the US and Dane country.

      And I am Professor of Free Market Epidemiology (as well as Law and Economics) at Wossamotta U.

      1. baffling

        well rick stryker, you completed a couple of the questions in the week 1 problem set from the intro to pandemic modeling online course you are taking. that is a LONG way from having any credibility in the field of pandemic modeling. as i stated previously, you have NO credibility in the field of pandemic modeling. if you did, you would have provided the evidence. you are not even an amateur in the field, rick. stay in your own lane, rick.

  25. Rick Stryker


    My advice to you is “never teach econometrics.” Maybe you know the area of economics you specialize in, but I’m starting to have my doubts. I gave you the whole quote and then bolded the relevant sections. My quote from my past comment was just what I said it was. I did not say that Menzie ran a spurious regression. Again, I was commenting on interpreting hypothesis tests when variables are trending. I was saying that the regression is either spurious or cointegrated (implicitly making the relevant assumption that the underlying variables are I(1)), and in either case conventional hypothesis tests are not valid. This is a simple point. I’m not sure why you can’t understand it.

    1. Menzie Chinn Post author

      Rick Stryker: The original context was looking at some diagnostics when running a log levels regression (consumption on income). Some statistics have nonstandard distributions that can signal a spurious regression — e.g. Jarque-Bera, others. So, you are correct that standard hypothesis tests involving calculated standard errors can’t be conducted; however, not all diagnostics are uninformative.

      1. Barkley Rosser

        Of course the convergence claim by “Rick Stryker” depend on a variety of assumptions he did not list, and I did not point out, although I effectively did in one of my comments above, with this getting really tiresome. But since the possibility of “nonstandard distributions” has been raised by Menzie, I shall simply note that indeed among those unmentioned assumptions hiding behind his derivation is certain degree of “standardness” of distributions. But this is getting really tiresome, especially as “Rick Stryker” continues to pretend that he did not out and out lie about what he said here. This is more Trumpishness, one minute claiming not to have said something, but then when it is clear you did, attempting to argue that how people interpreted this thing you claimed not to have said is incorrect. Truly pathetic, but truly Trumpish.

        1. baffling

          you need to understand how rick works. he came onto this blog criticizing the pro’s for their accuracy, knowing full well that his criticism was dishonest. the pros did the best they could with the data available back in february. rick reruns using conditions known in june, and then says see, i am better than them so my voice should be heard on this issue. its akin to criticizing the national hurricane center because they missed the mark by a few hundred miles in florida while predicting the path of a hurricane that formed just off the coast of africa. it is dishonest.
          but rick stryker has also been dishonest in the way he has shifted the argument-a common trick on his part. he has shifted away from his credibility and expertise in pandemic modeling-in which he has none-into the area of econometrics. but then he tries to use some expertise in econometrics to reinforce his supposed credibility in pandemic modeling. but again, he has absolutely no credibility in this area. but this slight of hand is used by him to try and fool the casual reader into thinking rick has some creditability in the area of pandemic modeling-of which he has none.
          rick used to be a more active poster on this blog than in the recent year, so some of you are not really aware of his tactics. but they really are devious, with sleight of hand used to redirect an argument and reframe it to his advantage. he will try to take a small “gotcha” moment and extend it beyond its application. his intentions are not good, which is why i am not in favor of “rehabilitation” but banning his commentary. the hosts of this blog disagree, and i fully understand their position. but as long as he remains on this blog, i will maintain a voice to point out his intellectual dishonesty-because as we have seen with trump, that does have consequences.

        2. Rick Stryker


          It seems that now you are now saying that my claim rests upon some assumptions. This is a blog, not a journal, so I wouldn’t think we’d like to get into such details. But ok, fine. Why don’t you tell me what my unstated assumptions are and then comment whether they are broad enough for my claim to be valid for applied econometrics?

  26. joseph

    Barkley Rosser: “But this is getting really tiresome …”

    That’s because the whole thing is just a way for little Ricky to say “Look at me. I’m smart too!” There is no other point to it.

    PERSON WOMAN MAN CAMERA TV “They say, ‘That’s amazing,’” he added. “‘How did you do that?’

    It’s just a performance. It has nothing to do with the subject of the original post and the fact that Rick Stryker is a nitwit, always wrong about everything. And willing to do anything to debase himself in service of his Dear Leader.

    1. Rick Stryker


      I’m not trying to prove I am smart. I’m giving a few of you a much-needed lesson in intellectual humility.

      You never say anything substantial so it’s impossible to give you a lesson.

Comments are closed.