Leiter Reports: A Philosophy Blog

News and views about philosophy, the academic profession, academic freedom, intellectual culture, and other topics. The world’s most popular philosophy blog, since 2003.

  1. Justin Fisher's avatar

    To be worth using, a detector needs not only (A) not get very many false positives, but also (B) get…

  2. Mark's avatar

    Everything you say is true, but what is the alternative? I don’t think people are advocating a return to in-class…

  3. Deirdre Anne's avatar
  4. Keith Douglas's avatar

    Cyber security professional here -reliably determining when a computational artifact (file, etc.) was created is *hard*. This is sorta why…

  5. sahpa's avatar

    Agreed with the other commentator. It is extremely unlikely that Pangram’s success is due to its cheating by reading metadata.

  6. Deirdre Anne's avatar
  7. Mark's avatar

Cochrane v. Krugman…or the Pseudo-Science of Macroeconomics Implodes

Brian Leiter

We noted Krugman's attack (which didn't go far enough) on the "pseudo-science" (with apologies to Larry Laudan) of macroeconomics, but now one of his targets, John Cochrane, an economist in the Finance Department at the Business School here has responded, and in terms that are not wholly, shall we say, edifying.   There is a useful overview of the debate here, which notes that, "Cochrane makes the same mistake he accuses Krugman of, by caricaturing and oversimplifying Krugman's argument, and, even worse, complaining that Krugman is only interested in making personal attacks on an ever-growing 'enemies list,' while engaging in his own litany of vicious slander."  The only "personal" criticism I can find in Krugman's original piece is the suggestion that some free-market utopians are motivated by the financial rewards of touting the party line.  Cochrane is far worse, but readers can decide for themselves.  But from a philosophical point of view, the really remarkable aspect of the Cochrane reply is the start.  As this observer notes:

I am trying to read John Cochrane's comments on Paul Krugman's article on why economists got it so wrong. I tend to get upset while reading. I have managed to get through the first paragraph in which Cochrane compares Krugman to someone who denies that HIV causes AIDS and compares developments in economics to progress in the natural sciences. Now in the natural sciences, there are counter intuitive models which are consistent with all available data and have withstood many tests. Also there are once mysterious facts which are explained by theories to the satisfaction of all people familiar with the theories.

Professor Cochrane even mentions the fake Nobel Prize in Economics as a way of legitimating the epistemic credentials of the field–which, of course, was precisely the objective of those who created it.  Yet this works only if you think of that Prize as on a par with the prizes in Physics and Chemistry, rather than, say, Literature or Peace, the latter two being quite explicitly ideologically motivated more often than not.

I am curious to hear how the Krugman-Cochrane dispute looks to philosophers engaged with economics.  (I am not inviting a free-for-all discussion of Krugman's or Cochrane's views, there are plenty of places on the Internet to rant and rave about that.)

UPDATE:  Here's a more substantive, and calmer, response to Krugman by a Minnesota economist, which suggests that Krugman has misdescribed the state of macroeconomics among its leading young practitioners.

Leave a Reply

Your email address will not be published. Required fields are marked *

22 responses to “Cochrane v. Krugman…or the Pseudo-Science of Macroeconomics Implodes”

  1. David Wallace

    I think the Waldman quote is a bit misleading (having read the rest of his post). The section quoted gives the impression that he's making a rhetorical point, i.e. that economics doesn't explain anything or dissolve any mysteries. But actually he goes on to have a serious discussion about whether there any such examples, and comes up with a few, mostly from several decades ago.

    That seems more in support of Krugman's position – that economics is very hard, hasn't done remotely as well as the natural sciences, and has taken a wrong turn lately, but is still a serious epistemic enterprise with some things to show for it – than in support of the view that economics is pseudoscience.

    Reply
  2. Alan Nelson

    Krugman's piece emphasized that during the lead up to recent economics events mathematical models proved to be of limited practical value. As far as I can tell, the most substantive part of Cochrane's response is the claim that the most impressive prediction of the efficient markets hypothesis is that good predictions of prices are not possible.
    He also implies that the recent fall in prices (otherwise known as the "recession" or "meltdown", don't say "depression"!) is best understood as a random fluctuation. If that were right, then Krugman would be wrong to think that it might have been predicted or better understood by some other kind of economic science.
    It is understandable that Cochrane is miffed because people want economic science to help prevent downturns and not only mathematically ensure that "the then really does follow the if".

    Reply
  3. Brian

    David Wallace: Fair point, the quote I used suggests that the author is more skeptical about economics than he is.

    Alan Nelson: what kind of science of economics is it that can not even qualitatively predict that some event will occur, even if it can not predict the precise contours and character of the event? The EMH may 'predict' that we can't predict prices, but apparently on this line of thought, economics can not even predict the direction of prices. That's some science!

    Reply
  4. Kyle Walker

    Although I think they are open questions, I think it is possible that the presumption of rationality is wrong and I think it is possible that economics doesn't fit a demarcation criteria.

    However, the idea that this recession substantiates these points is tantamount to the idea that the soviet union's collapse proves Marx was wrong.

    It is clear that the FED's actions, intervening in the economy by cutting rates, indirectly created the bubble which collapsed last October. The reason they had to cut rates to zero is that the constant propping up of the economy by the FED forced them to incrementally stimulate more and more. No appeal to irrationality is needed here. That is the position of most Austrian economists, and I think it is a fairly obvious point.

    The observation that this was happening also allowed a number of economist (and others – see Peter Schiff's book "How to Profit from the Coming Economic Collapse") to predict that this sharp fall was going to happen. I take it that this is, in fact, a qualitative prediction.

    Reply
  5. Brian

    I don't quite follow Mr. Walker's comments. The collapse of the Soviet Union has an all-too-obvious Marxian explanation (bureaucratic control fettered the development of the USSR's productive powers, pretty obviously). In any case, no one was defending the predictve success of Marxian economics or Marx's theory of history. The worry is that the putative scientific discipline of macroeconomics failed to predict an event, and not a minor event, but an economic earthquake. It failed to even assign it a probability. That a few macroeconomists did predict it may mean they have a good, alternative scientific paradigm for the study of economic phenomena, or it may mean they were lucky. Macroeconomics is sufficiently epistemically feeble (or shall we just say, immature) that we just don't know at this stage.

    It appears to be true that the decision of Alan Greenspan, a 'true believer' in the wisdom of markets, to cut rates did contribute to the housing bubble. But what is the relevance of that? The economic collapse resulted, I thought, from the secondary markets that had been trading mortgage derivatives, not from the bursting of the housing bubble itself. The market failures here were multiple and widespread. How does this help the Austrian utopians?

    In any case, this thread will not be sidetracked into either Marx or Austrian economics. I'd like to keep the focus on the philosophical questions about the epistemic status of economics that the Krugman-Cochrane dispute might seem to raise.

    Reply
  6. Matt Drabek

    This dispute between Krugman and Cochrane really exemplifies some of the work that has been done in the last 15 years by Ian Hacking. Hacking shifts his emphasis away from social scientific explanation and toward social scientific classification.

    One of Hacking's central points is that classification in the social sciences involves interaction between the classification (in this case either 'bull market' or 'recession') and the objects of classification (i.e. – the behaviors of people living under the economy, and the work of economists).

    The central "Hacking" lesson to be learned is that prediction in the social sciences can be screwed up when people become aware of certain ideas, and then change their behavior as a result of that awareness. I think this may get to Krugman's concern about prediction in macroeconomics. It is tough to make a prediction or a classification when the very making of that prediction or classification carries the potential to change the ballgame. The prediction of a permanent bull market led economists and policymakers to enact careless or ineffective policies, and created the conditions for the undermining of the prediction.

    Reply
  7. yakov_a

    Couldn't a similar argument be made about the epistemic status of medicine and biological sciences? Why hasn't modern medicine prevented the most recent out break of swine flu or cured cancer?

    While I don't doubt that modern economics contains an odd mixture of humanities, politics and science, it seems a bit much to allege "no progress" has been made. Economic historians have shown increasingly widely spaced recessions and ones that are far less devastating than in the past.

    Further, couldn't some of our epistemic difficulty be caused by economics being a science that inherently studies people capable of understanding it? Doesn't that make prediction inherently more difficult? Atoms don't tend to object to your studying them, although they too run into problems with cats in boxes.

    Lastly, few sciences have had their terminology and concepts so readily spread through popular discourse so quickly. The entire concepts of recession, depression etc. for example. Doesn't this, combined with my last point, raise a problem about what is, in fact, common sense.

    In any event, I would have thought the link between inflation and unemployment not something readily apparent. Nor, historically, was that mercantilism wouldn't work.

    Reply
  8. Michael Drake

    If Peter Schiff "predicted" the economic collapse, he apparently did it in a language very few other economists understood. So much the worse for economics, yes?

    Reply
  9. Kevin Schutte

    I'm not a "philosopher engaged with economics", so perhaps my comment is out of place, but this "debate" reminds me of a similar one in the philosophy of language:

    Person 1: Your position can't be right, because it has made so little progress in decades!
    Person 2: Not so! Your position can't be right, because the "progress" it has made applies only to hypothetical ideal languages and not real world languages! Your "results" are either tautologies or false!
    Person 1: Not so! Your position can't be right, because it has made so little progress in decades!
    …ad infinitum.

    Just substitute the word "markets" for "languages" and you have the Krugman-Cochrane controversy. So my conclusion is that economics is only as scientific as philosophy of language is. And, so, not very scientific (but not inconsistent with science).

    Reply
  10. Kevin Donoghue

    "I'd like to keep the focus on the philosophical questions about the epistemic status of economics that the Krugman-Cochrane dispute might seem to raise."

    I’m no philosopher, but I think a better point of departure might be Keynes’s comments about the nature of economics. Although he hoped economists might one day be useful people “like dentists” he also rejected the notion that you can ever really settle disputes: “in economics you cannot convict your opponent of error; you can only convince him of it.” Yet somewhat inconsistently, he claimed that economics really was on the way to becoming a science.

    Do philosphers take Keynes seriously as a philosopher? If not, why not?

    Reply
  11. Roger Albin

    For a more scholarly discussion of the Krugman-Cochrane debate see:

    http://economics.uwo.ca/econref/WorkingPapers/researchreports/wp2009/wp2009_2.pdf

    It contains a fairly damning quotation from a speech by Cochrane (several years ago) in which he concedes that recent macroeconomic work, including his own, has little relevance to phenomena like the Great Depression or the Latin American debt crisis. This would appear to validate Krugman's point nicely.

    As a non-philosopher, I'm not sure what the "epistemic" status of the field constitutes. Krugman, apparently with some justice, is accusing Cochrane and others of being ideologically committed bad scientists. So macroeconomics doesn't have much predictive power, so what? How much predictive power did weather models have when first introduced? The first serious attempt at weather modeling was done during WWI and the result was, not surprisingly, dramatically wrong. It was many, many years and many generations of advances in computing before there were reasonable weather models.

    Reply
  12. Robert Waldmann

    I think the Waldmann quote is pretty fair and I'm paid a salary to claim that I am an economist. I would like to point out that it isn't all that bad.

    Macroeconomics is particularly difficult, since great depressions are so rare that they are hard to study.

    Economists studying things like wages, asset prices, duration of spells of unemployment have become much more like natural scientists than they used to be. Theory without data is not given much much too much respect. Empirical work is based on identifying assumptions which seem obviously true not wishful thinking.

    (identifying assumptions are assumptions needed to draw conclusions about causation from correlation. One such would be the assumption that the suicide rates of people born on different days varied from year to year depending on the order in which they were picked in the draft lottery 18 years after the birth, then probably the draft and the war in Vietnam was the causal link. This is an actual example. An economist decided to do the same thing with wages (economic) not suicide (demographic). He was hired by Harvard. I'd say that was science and shows that economsts recognise it when they see it.)

    Reply
  13. Roger Albin

    Robert – there is clearly a lot of economic work which is empirically based and scientific by conventional standards. I recently read the Goldin and Katz book on the consequences of US educational policy over the last century. Here is a case of where sensible use of simple models and analytical methods is quite revealing. Its an interesting question of how things like this macro mess occur. I suspect fields that are not well developed may be more susceptible to ideologically driven problems like this one – the racist genetics earlier in the 20th century would be an example. But even well developed fields with strong "epistemic" standards can have these problems when the field becomes entangled in major political issues. The "German physics" of Lenard and Stark in the 1920s and 1930s is a good example of a mature field suffering a collapse of critical standards. German physics wasn't, like the Lysenkoist genetics of the Soviet Union, something imposed from outside the field, it was led by distinguished figures like Lenard, who was a Nobel laureate for his important experimental work. Its certainly no accident that Cochrane et alia come from bastions of pseudo-libertarianism and that their version of macro fit in well with the general reactionary tenor of American life over the last generation.

    Reply
  14. Greg Ransom

    Read economist Mario Rizzo on Cochrane here:

    http://thinkmarkets.wordpress.com/2009/09/13/the-great-moderation-in-macroeconomics/

    And economists Bob Murphy here:

    http://consultingbyrpm.com/blog/2009/09/my-response-to-john-cochrane-who-was.html

    Reply
  15. Paul Farley

    1. Economists know far more about the economy than anyone else and in 2008-09 economists saved the world from another great depression by advocating and obtaining stimulus spending. Remember, no one knew what to do at the start of the great depression – not physicists, not sociologists, not medical doctors not anyone. It was economists who picked up the pieces and formed the theories that provided clear guidance to government in 2008/09 – no one else.

    2. Is the profession split? Its pretty clear it is in the US, although I don't think so in much of the rest of the world. I tend to put that down to the rather extreme distrust of government intervention amongst some parts of the US population. Keynes is taught in every undergrad course in Australia, with the "freshwater" stuff being taught at more advanced levels due in part to its more difficult mathematical nature. In general, economists around the world are exposed to all current schools of thought, some of which will turn out to be dead ends.

    3. If anything counts as a science (and I know the philosphers can blur the boundaries) then economics does. Broadly, it tries to relate a set of data points to each in a fairly rigourous manner, usually using mathematics. And it does make predictions – the fact that those predictions are often wrong is neither here nor there – it is an evolving science. And people shouldn't crowe too loudly about the failures of economics – unless they have a better theory that is…

    4. The fact that economics is pushing up against issues of predictability should be seen as an interesting fact, not as a reason to condemn economics. After all, if economics is pushing up against these issues, then every other social discipline is too – we are all in the same boat here…if physicists are so clever, then why don't they take over from the economists and straighten everything out? Oh, wait, they did do that and formed the freshwater schools…

    Reply
  16. Brian Leiter

    I would have thought that if "anything counts as a science," it would have been physics or chemistry, not economics. A science that makes mostly wrong predictions is an odd kind of science–or, perhaps with a nod to Laudan, just a really bad science. A science that fails to even predict the occurrence or even likelihood of the major economic event of recent decades is obviously pretty epistemically feeble.

    Reply
  17. Paul Farley

    1. But who says those predictions are "mostly" wrong? The fact is that central banks have reduced the volatility of the business cycle over the last 30 years based on a Keynesian understanding of the economy (Ben Bernanke has written on this – google "The Great Moderation"). So I don't regard economics as "epistemically feeble" at all. Ok, it doesn't make counter-intuitive predictions about the nature of space and time, but hey, what do you want?

    2. Regarding the last housing bubble, lets be clear. The default view amongst sober economists was that house prices far exceeded the level at which rental returns could make investors a profit. People were buying houses not for rental returns, but so that they could sell them for even more in the fture – that is a classic bubble. Sober economists knew this and it wasn't controversial. True, some economists/financial analysts made all sorts of conjectures about permanantly lower cost of capital etc. but these were conjectures, not orthodox economic theory. Those who followed the orthodox theory were correct. Why did so many not follow the orthodox theory? Maybe the huge profits they were making had something to do with that…

    3. Having said that, the US Federal Reserve, including its chairman, were very slow to try and do anything about the housing bubble, and Bernanke went so far as to deny one existed at one point, for which he has been rightly kicked in the pants. But the US is not the world. The AUstralian central bank, for instance, was issuing warnings about a housing bubble very early on – people in Australia have no cause to blame economists if they bought into the housing market at the peak of the cycle – they were warned.

    4. I think this point is important – any analysis of economics has to recognise that in the US there is a particularly virulent form of "free market" ideology that, to my mind, is not actually justified by economic theory and is not necessarily prevalent in other parts of the world. Economics textbooks say that the free market can result in good or bad outcomes – bad is monopoly and oligopoly, good is monopolistic competition and competition. Free markets can also suffer from information assymetries and externalities which push them away from competitive market outcomes. And it is clear that free financial markets can suffer from misaligned incentives that allow dealers to make large short term profits even as they engage in highly risky long term strategies. Economics doesn't need to learn this – it already knew it. But financial markets don't like this sort of academic talk and, in the US, Wall Street has far too much influence.

    Reply
  18. Brian Leiter

    1. Those who have studied the track record of macroeconomics, like Hausman and Rosenberg, have concluded that the predictive record is extremely poor.

    2. Paul Volcker, former Fed Chairman, denies that economics has contributing anything to banking policy.

    3. It is not true that most economists, or even a minority of economists, diagnosed the bubble or saw the crash coming.

    4. It is true that Wall Street has too much influence in the US!

    Reply
  19. Jonathan Halvorson

    One thing that the defenders of economics as a science have to contend with is the fact that there are no numerical constants in their science. There are no mathematical laws with constants, nor are there well-confirmed models of real world systems that have uncontroversial constants of even a single digit (let alone the often dozens of significant digits in theories and models in the natural sciences). This was a point that I first heard Rosenberg make, though I believe the observation is older.

    Paul Farley is much too quick to associate economics with the likes of physics, chemistry, fluid dynamics or biology (the parts that are not essentially historical or taxonomic…such as physiological inquiry).

    I do think that the Rational Expectations (Fresh Water) economists are partially right in their analysis of the poor predictive ability of economics: human beings alter their behavior as they gain new information. New information can include predictions themselves. But FW economists want to stop there and keep the elegant structures of their theorems, protecting them from any potential disconfirmation by real-world human irrationality. What they are doing, essentially, is a branch of decision theory. They are ferreting out the implications of various assumptions, and refuse to accept that irrationality of various sorts might factor into those assumptions. Their process is closer to philosophy than science. But of course they don't want to admit that what they are doing is not science in the same way that chemistry or physics is, so they lower the bar in what it takes to count as science just as John Cochrane and Paul Farley do.

    This is not to impugn, by the way, that economics has helped us to manage the world better. It clearly has. But I think that we need a two-part understanding of what this science is: one the one hand it consists of theorems that create a conceptual framework for how to think about prices, quantities, distributions, etc. Here progress has been huge and unambiguous. And here the new and post Keynesians have the clear edge, because they can accept the FW models in circumstances in which they do a good job approximating the conditions, but they also have conceptual toolkits for when the assumptions in those models fail and we get stickiness based on irrational behavior of the sort that Kahnemann and Tverskey, Thaler, and Keynes himself, among others, have described.

    The second part of economics is the more directly empirical (to use a loosely Quinean framework) and it consists of models that are temporary and grounded in statistical methods. The coefficients change. There are no constants. But you do have the ability to make decent predictions under normal circumstances using the principle that as things happened in the recent past they will continue to happen in the future. Assumptions about propensity to consume, tolerance for risk, etc., can work reasonably well in the short term most of the time. But it's all in a shifting terrain, and when the earthquakes come all bets are off. The models fall apart. That is unlike, by the way, geology, which has become very good at earthquake prediction.

    PS: I have a PhD in Philosophy but have left the field. I believe there is a current philosopher of science with the same name as me and don't want him to be falsely associated with this post.

    Reply
  20. Christopher Morris

    I have hesitated to comment as I know so little about the philosophy of science. But I do know that economists, insofar as they have any thoughts about "methodology", cite outdated stuff (behaviorist, instrumentalist). For years I have berated economists for emphasizing prediction and forgetting about explanation. And now I read many of you trashing economics because of its poor predictive power! Economists do have explanations, but it never occurred to me to think that they would be able to predict much (other than simple things, e.g., raise the minimum wage to $15/hr and expect to see the unemployment rate or the black market to grow).

    I should have thought there were problems about expecting much predictive power in the social or "human" sciences. (I care nothing for the term 'science' here and don't care to pronounce on the scientific status of economics, except that it's about the best we have in the social sciences, history aside.) Surely if one could predict crashes or booms there would be a lot of money to be made on the markets (hmm… but that might affect the truth or at least the timing of the predictions…).

    I am ignorant about the predictive powers of biology, but I'm pretty sure that geology is not known for its predictive successes (who will be blamed when the Big One catches us unaware?).

    A last comment I've wanted to make for awhile. I read Cochrane's piece quickly and don't know his work. But he could easily be wrong or out-to-lunch w/o that in the slightest rehabilitating Krugman. The latter's NYTimes columns are extraordinary in many respects; he seems to want to outdo literary critics for talking about things he does not understand and for distorting the positions of his opponents.

    CM

    Reply
  21. Paul Farley

    Jonathon, I found your post to be very well considered, but I disagree with some of the reasoning.

    I'd like to set aside this science question for a bit and concentrate on what economics can and cannot do.

    As pointed out, it would be kind of a strange world if economics could predict future output, prices, quantities etc. Those predictions would themselves change behaviour (if you know that oil will be rise to 150 per barrel next year, then you will buy oil now, and it will rise to 150 per barrel straight away – not next year).

    The thing is, economists know very well that they can’t predict the value of economic variables. We know this because economic variables are random walks, and we know that because Clive Granger did some very good work in econometrics – for which he won one of those fake nobel prizes – that gave us a methodology for determining which variables are random walks.

    But the econometricians also gave us some models that, while not allowing us to predict future values of variables, do allow us to see how variables move together. For instance, we don’t know what the future value of house prices will be, but we do know that rents and house prices typically move together and shouldn’t get too far out of whack. This is the basis on which Paul Krugman predicted the current recession in 2006 (its on youtube – see

    Now, returning to this science question and the issue of prediction, I would say that it depends on what sort of prediction you are talking about. If economics is expected to predict the future value of economic variables in order to qualify as a science, then it can never be considered a science because economic variables are random walks and it is impossible to predict them. But if economics is expected to tell us how variables move together and make predictions such as “house prices are not justified by rental prices – they will have to eventually come back into realignment but we don’t know exactly when” then I think it has more chance of being considered a science.

    As for the criticism that there are no fixed constants in economics, I don’t disagree, but I don’t see why that prohibits economics from being a science – surely what makes a science is methodology (observation, theory, prediction, not necessarily in that order) rather than the underlying thing being studied? If underlying variables change value, economists and econometricians will find out precisely because they value observation. Who is to say that the underlying variables of physics are constant? Maybe they change over many billions of years.

    Reply
  22. Dave Hanson

    Brian, the link to the Kocherlakota piece in your update is broken. Would you or one of your readers have a current link? TIA.

    Reply

Designed with WordPress