05 Aug 2013

Thoughts on My Debate With David Friedman

David Friedman, Economics, Rothbard, Shameless Self-Promotion, Steve Landsburg 97 Comments

After reviewing our debate (which I’m reposting below), I realized there were a few points where I misunderstood what Friedman was saying. It will not surprise you to hear that I think there were also several points where he misunderstood me. (For one obvious example, of course I know that modern physicists think the universe is non-Euclidean. That has nothing to do with my point: Even in light of that fact, if a student in 7th grade tried to “prove” theorems in geometry by appeal to experiments, he should be marked down heavily for totally misunderstanding how geometry works.)

Rather than re-litigating every twist and turn in the debate, let me try a new way of motivating my position (and I’ll send this post to David to see if he’s interesting in pursuing the discussion). I claim that the people who today consider themselves free-market economists got that way not because of empirical research, but because they read ingenious thought experiments from the likes of Bastiat, Say, Henry Hazlitt, Milton Friedman, Ludwig von Mises, Friedrich Hayek, Murray Rothbard, David Friedman, and Steve Landsburg.

Take free trade in particular. Nobody believes in free trade because of regression analyses, or of “natural experiments” in which some country dropped tariff barriers as an offshoot of some other, unrelated event. Rather, people believe in free trade because of brilliant explanations such as David’s own “Iowa Car Crop” (which I first read in Landsburg’s Armchair Economist, and which Steve correctly described as “a thing of beauty” that can “change the way we see the world”).

Going further, to the extent that you might abandon your teenage belief in free trade, it would again be due to a thought experiment, a simplified tale showing how (say) a big country could affect the world price of cotton and so maybe a small tariff could, in principle, be welfare-enhancing. Even here, the disputes would be more philosophical rather than empirical. Nobody would convince someone to change his mind on free trade, by pointing to empirical evidence. For example, I remember when Pat Buchanan and some others pointed to the strength of the US economy when there were high tariff barriers. That didn’t faze me a bit (and still doesn’t), because there are a million moving parts in these historical analyses. I think the US would have been even richer had they adopted free trade back then.

Another consideration: Intro textbooks or even pop books on economics often start with a section about, “How to Think Like an Economist.” These principles or “laws” of economics are almost always non-falsifiable. They are things like, “There are always opportunity costs,” or, “There’s no such thing as a free lunch,” or, “Decisions are made on the margin.” You don’t go out and “test” whether these statements are true, instead you decide a priori that you are going to use these principles to parse the evidence.

Yes, if we’re trying to assess whether the Great Recession was more due to supply-side or demand-side “shocks,” then that is an empirical question. But it’s not a “law of economics” to ask whether the recession from 2007 onward was due to the supply or demand side.

In contrast, the “laws of physics” are entirely empirical. In principle, they might be different in 200 years, as physicists learn that Nature obeys a different set of rules from what we currently believe. In contrast, when the “marginal revolution” occurred, it wasn’t because of new data, it was because economists came up with a better mental way of organizing and explaining what humans had known for centuries.

97 Responses to “Thoughts on My Debate With David Friedman”

  1. David Friedman says:

    What you are saying may well be correct, but I don’t see its relevance to the question of the Chicago approach vs the Austrian approach since, as I tried to explain several times over, the Chicago approach starts with the theoretical argument. But when you go beyond questions that are reasonably simple and familiar, you end up with theoretical arguments that depend on plausible simplifying assumptions and you need empirical evidence to decide whether you have kept enough of what is relevant to the problem to get the right answer. In addition, when doing original work, the need to find testable predictions of your theory forces you to think through the theory itself more carefully.

    Consider, as an example, my theory of the size and shape of nations, webbed on my site. It’s economics, not economic history, because although I am testing it against historical data, the theory itself is as relevant to the future as to the past. Or consider my Cold Houses in Warm Climates article, also webbed. The conclusion depends on at least two simplifying assumptions, either of which might be wrong. I formed the theory on the basis of casual empiricism, and I would be more confident it was correct if I, or someone else, had done a more careful examination of the data on thermostat settings as a function of local climate.

    With regard to a different point you made somewhere recently on our exchange, I didn’t focus on the natural rights vs consequentialism issue because I don’t think it has anything to do with the Chicago vs Austrian issue, which was supposed to be our topic.

    • guest says:

      From how you spoke about your father, it seemed he was willing to give a lot of himself to you. It seemed you were pretty close.

      I thought that was cool.

    • Matt Tanous says:

      “But when you go beyond questions that are reasonably simple and familiar, you end up with theoretical arguments that depend on plausible simplifying assumptions and you need empirical evidence to decide whether you have kept enough of what is relevant to the problem to get the right answer.”

      I am curious as to how you intend this. Because if you mean, say, examining a historical event, you might need empirical evidence to determine which economic laws were in play – what events occurred and what effect they would have. But if I am analyzing price controls, it is not simplifying assumptions I am making, but a thought experiment in which I hold everything else constant but what I am examining philosophically. I KNOW price controls cause problems because of this analysis, and an empirical result where (for instance) rent control was passed and the accordant shortage did not manifest would not faze me. I’d KNOW some other factor was involved, as my ceteris paribus analysis is not dependent on empirical data.

    • Jeffrey says:

      “But when you go beyond questions that are reasonably simple and familiar, you end up with theoretical arguments that depend on plausible simplifying assumptions and you need empirical evidence to decide whether you have kept enough of what is relevant to the problem to get the right answer.”

      I would argue that the plausible simplifying assumptions you make in, for example, your theory of the size and shape of nations, are peculiar to time and place. Thus, these simplifying assumptions are simply aids to enable you to apply (correctly or incorrectly) the universally valid laws of economics proper to particular conditions. This would put your theory of the size and shape of nations in the discipline of economic history.

    • Jeffrey says:

      “In addition, when doing original work, the need to find testable predictions of your theory forces you to think through the theory itself more carefully.”

      I do not dispute this point. However, economics proper is afforded no such luxury because, as Bob and others have pointed out, the empirical data generated in the social sciences is categorically unfit to prove or disprove economic theory.

      It is true, some empirical evidence may provide the impetus to an economist to rethink his theory, but the resultant rethinking/revision of the economic theory would be completely internal to the theory itself, since, again, empirical evidence cannot prove or disprove economic theory.

      The Austrians maintain this position not because of any desire to escape the “rigor imposed by empirical data” or to arrogantly proclaim the superiority of pure reason over empiricism. It is simply a correct recognition of the true state of affairs with regard to the relationship that exists between economic theory and empirical data. The Austrians, it is true, say that empirical data cannot disprove economic theory. Note, however, that they also say that it cannot prove economic theory, either.

    • Ivan Jankovic says:

      David Friedman: “Consider, as an example, my theory of the size and shape of nations, webbed on my site. It’s economics, not economic history, because although I am testing it against historical data, the theory itself is as relevant to the future as to the past”

      This has nothing to do with economics, although it might be interesting or clever or illuminating. Economics in Misesian approach is not a predictive science at all: demand curve slopes downward, marginal utility is diminishing, the rate of time preference is positive, and so on. those are supremely important principles of economics but they do not yield any “testable predictions”.

      Of course, it is conceivable that some of these principles could be incorrect, but you cannot ascertain that by an appeal to “empirical evidence” pertaining to the isolated statements following from them: the only thing you can do is to change your axiomatic foundation for deduction, by pragmatically changing some of the axioms if you think they are wrong or useless for the purposes you want to achieve. But,, once you had done that it does not make any sense to “test empirically” the more complex claims which are the logical derivations from your axioms. Those complex claims are not “testable predictions” at all, but apodictic truths, as long as you accept the axiom or axioms from which they follow. Of course, you can use empirical evidence at the level of deciding which axioms to choose (Rothbard accepts this), but not in evaluating the truth of the theorems. The theorem that minimum wage increases unemployment follows logically from the simple supply-demand analysis, and once you accepted that demand curve slopes downward and supply curve upward you cannot say: now, I don’t like this specific consequence of those axioms pertaining to minimum wages, and want to to “test it empirically”. That is just like going outside and measuring the real triangles in order to check if the angles add up to 180 degrees, AFTER you accepted the Euclidean axioms from which logically follows that they do! Of course, you can revise the axioms by adding the non-Euclidean ones and thus get a different description of reality, but again you cannot “test empirically” this new description either, after you revised the axiomatic foundation of your science.

    • Bala says:

      In addition to Jeffrey’s replies above, I thought it fit to point out that unlike what you say here,

      “But when you go beyond questions that are reasonably simple and familiar, you end up with theoretical arguments that depend on plausible simplifying assumptions and you need empirical evidence to decide whether you have kept enough of what is relevant to the problem to get the right answer.”

      the Austrian explanation does not depend on plausible simplifying assumptions. It starts from axiomatic propositions and concepts and goes on to derive the laws of economics using deductive reasoning. There is therefore no role for empirical validation. I have not studied the Chicago School’s work, but from your statements, I infer that it makes assumptions, builds models, uses the latter to make predictions and then tests the models themselves using empirical data that is compared with the predictions. For this reason alone, I would consider the Chicago School approach deeply flawed. Unless you can explain why the Chicago School approach is epistemologically sound, it would be difficult to accept your arguments.

    • guest says:

      Nobody caught this one, yet?

      Consider, as an example, my theory of the size and shape of nations, webbed on my site. It’s economics, not economic history, because although I am testing it against historical data, the theory itself is as relevant to the future as to the past.

      He doesn’t have access to data in the future, so he is actually attempting to validate his theory with logic, not data.

      • Tel says:

        No, he is presuming that the mechanism that operated to produce past data will be the same mechanism that operates to produce future data. That’s not proven either, every law in the universe might change tomorrow, but so far this presumption has proven solid.

        • Ken B says:

          Exactly.

          This is what I find so curious about your Darwinism posts Tel. Because the basic logic is the same, using history to test hypotheses e, and then hypotheses to explain history.

      • Garrett Watson says:

        The theory is confirmed if the predictions made by the theory turn out to be correct, and falsified if incorrect. The logic of the theory is not what validates or refutes it.

        Instead, he is differentiating between economic history and economics – economic history focuses solely on past data points, while economics uses theory to make predictions while also verifying that the theory can adequately explain past events.

        • guest says:

          The theory is confirmed if the predictions made by the theory turn out to be correct …

          No, the theory is only exemplified at a specific instance in the future, if the prediction comes true. But what about next time?

          This is why you can never prove a theory with data.

    • TheDjinn says:

      “But when you go beyond questions that are reasonably simple and familiar, you end up with theoretical arguments that depend on plausible simplifying assumptions and you need empirical evidence to decide whether you have kept enough of what is relevant to the problem to get the right answer.”

      Your use of the phrase “right answer” makes it clear you view Economics as a physical science rather than a social science. Social sciences aren’t tested by making predictions, and neither is economics.

      Economic theory itself won’t help you make a prediction, and empirical “data” doesn’t falsify economic theory, because the so-called “data” is rarely unquestionable. In order to predict in social sciences, one must use a heuristical process and attempt to assign weights to various potential causal factors using judgment. There is no scientific manner to generate economic predictions from economic laws in the same manner that there is no scientific manner to formulate physical laws using physics predictions. In both cases, the transmutation of knowledge from objective laws to specific cases and vice versa is a heuristical and ad hoc process.

      “In addition, when doing original work, the need to find testable predictions of your theory forces you to think through the theory itself more carefully.”

      Arguably true, but testable predictions in economics are practically impossible because you are not in the position of being able to hold most or even much of any variables constant and thereby perform asuch tests.

      For reference, consider Rothbard in this link:

      http://www.mises.org/rothbard/schuller.pdf

      He makes a lot of great points. To expand on his position, the distinction between physical and social sciences is the following:

      Physical sciences mostly revolve around empirically testing hypotheses to derive laws, and when an empircal test suggests a scientific law that appears illogical or unintuitive, the test itself is not necessarily questionable as it is quite easy to formulate such a test. More likely, the law itself is poorly formulated or not fully supported by the test itself, or perhaps simply misunderstood by those attempting to test it. Anyone purporting to debunk well-formulated empirical tests through appeal to sense, consistency, or a priori axioms has most likely failed in their analysis or argument in some subtle way or generated axioms that are not as tautological as they first appeared.

      Social sciences mostly revolve around deriving laws through logic which can be used to create empirically testable hypotheses, but when a hypothesis doesn’t appear to be supported by any evidence this does not necessarily debunk the law. More likely, the test itself is flawed, due to the difficulty of generating said test and isolating for the variable in question. Anyone purporting to debunk a well-argued economic law through way of empirical test has most likely failed to formulate a true test and rarely, if ever, is it worth re-examining the law.

      Obviously, there is some potential for cross-over; an economist can use a casual examination of empirical evidence to aid in the generation of laws, and a physicist can use an appeal to consistency or sense to formulate a hypothesis. The point is, the scientific method is not the be-all and end-all of science. It is simply an effective method to generate scientific knowledge when and where empirical tests are easy to perform. A priori logic is another way to derive scientific knowledge, and should be used where most appropriate (that is, when the a priori axioms can be stated as truisms or tautologies with utmost confidence and surety).

      When neither axioms can be stated nor empirical tests be created, then no knowledge is possible.

  2. GeePonder says:

    It was an awesome debate. Great job by both sides.

  3. Bala says:

    I still think you are missing the point David Friedman was making and therefore still not getting to the response you should have given. The issue is not how the theory is drawn or explained. It is how it is validated. David’s repeated insistence, the implication of which was that that is what makes the Chicago School superior to the Austrian School, was that the Chicago School makes testable predictions about reality which it then tests by collecting data and validating theories. The emphasis he placed on coming up with conjectures that are validated using empirical evidence was very high.

    The key, therefore, is not whether the theory is based on data but whether it makes sense at all to talk of validating economic laws through empirical research. The real rebuttal should have been that it is meaningless to talk (as David does even in his response out here) of validating propositions drawn through deductive reasoning starting using empirical evidence. The very approach he speaks of is flawed and erroneous and is based on the mistaken notion that a field of study cannot be considered a science that informs us of the real world unless it makes testable predictions about the real world that are then validated with data. This error (I think) of empiricism is what you should have been pointing out.

    • Dan says:

      It sure seemed to me that Dr. Murphy was making this exact case during the debate.

      • Bala says:

        It didn’t seem to me that way at all. That’s precisely what I have been trying to point out. As Ivan Jankovic pointed out here, Bob did not address David’s repeated claim that the proper job of economic science is to make testable predictions and the correct way forward in economic science is to test these testable predictions against data.

        • Dan says:

          Yeah, I understand what you guys are saying. I’m just saying that I disagree. I think he Murphy did precisely what you guys are saying he didn’t.

        • Jaycephus says:

          I get how you can gather materials and perform a physical experiment, fairly repeatably.

          I don’t see how you can establish an economic laboratory and initialize conditions repeatedly, while supplying identical inputs for multiple runs at the experiment, and then do it again at some other location in the world, and hope to have anything proven when you’re done.

    • Bob Murphy says:

      Bala wrote:

      The very approach he speaks of is flawed and erroneous and is based on the mistaken notion that a field of study cannot be considered a science that informs us of the real world unless it makes testable predictions about the real world that are then validated with data. This error (I think) of empiricism is what you should have been pointing out.

      Bala, I really don’t know what to say. That’s what I was doing throughout the whole debate. That was the point of my geometry example. That was the point when I said the methods of the natural sciences don’t work well in the social sciences, and so we should use something different. That was the point when I said just look at how we’re arguing over the causes of the Great Depression still, so this proves empiricism doesn’t work. That was the point when I said how there have been cul de sacs in the social sciences but not in the natural sciences.

      Obviously I didn’t put a neon sign on it and categorize what I was doing, the way you wanted me to. But I hope you will agree that most of my time was spent talking about the very thing you are asking me for.

      • Bala says:

        Bob,

        It’s not about what I wanted to see. It was about what I could see him doing. I just felt that your attack was way too subtle. He just didn’t get it and kept repeating the point about economics being a science that tries to come up with conjectures that are to be validated through (the holy grail of) data.

        It is my opinion that when an opponent fails or refuses to get the subtleties of your argument, you should throw the neon sign at him. Failing that, he gets to go home thinking he has done a great job and enough members of the audience would go home thinking that he did indeed destroy you. His response to your point on geometry is a typical example. You (obviously) think you have presented a fantastic argument to which he responds telling you that physicists today think the real world is not Euclidean. What is a partially-informed listener supposed to infer from this? My fear is that many would infer that Bob Murphy has been demolished by David Friedman.

        I am just saying that you had the opportunity to decimate his epistemological foundation in public and you didn’t do it.

        • Bob Murphy says:

          Bala said:

          You (obviously) think you have presented a fantastic argument to which he responds telling you that physicists today think the real world is not Euclidean. What is a partially-informed listener supposed to infer from this?

          Bala, look at my response starting at 28:30.

          If you don’t like the way I made the point, OK, fair enough, but you’re chiding me for ignoring the issue, when that’s mostly what I talked about throughout the debate.

          Also, when I gave my geometry example and Friedman accused me of not knowing the physical universe is non-Euclidean, I truly decided at that moment, “Well, if that’s how he’s going to play this, and he wins people with that kind of response, I am happy to let him have those people. I’m appealing to the people who are listening to what I’m saying and can see the point.”

          • Bala says:

            “but you’re chiding me for ignoring the issue, when that’s mostly what I talked about throughout the debate.”

            No, Bob. I am slapping my forehead as hard as it can take it and screaming “WHAT A GOLDEN OPPORTUNITY WASTED!!!!”

            🙂

          • Bala says:

            OK, Bob. I’m getting a clearer sense of why I am finding your response wanting after listening to the segment from 28:30 (as you suggested) to 32:10. Your response was a defence of the Austrian approach, not an attack on Friedman’s (and the Chicago School’s) approach. Your response would have been clear to an Austrian listener well acquainted with Mises’ writings but not to others. That’s just my assessment, by the way.

        • Dan says:

          Bala,

          Maybe you should make a YouTube video or something where you can show people the proper way to handle this debate, because I thought Dr. Murphy did a great job explaining the inferiority of the empirical approach to economics.

          • Bala says:

            Dan,

            I’ve already done that on this board. It’s upto Bob to decide whether what I am saying – getting aggressive in attacking others’ flawed positions and approaches rather than defending your own position – makes sense or not.

  4. Bala says:

    Bob,

    This post of yours highlights a point that I have always been hinting at but probably never got down to laying out explicitly.

    I think many Austrians (quite unintentionally) end up presenting far weaker arguments than they could because they do not attack the epistemological unsoundness in the arguments presented by many from opposing school of thought. This debate is just one example.

    To take another example of what I mean, see how we take on arguments (by various schools of thought) based on GDP data. The correct approach (as I see it) is to remind our GDP-using opponents that the very concept of GDP is, economically speaking, deeply flawed and hence that their arguments based on GDP data are equally deeply flawed. We need to be pointing to gross savings and the savings-consumption ratio as the real measure of economic activity. Instead, we try to point out inconsistencies in their theories based on their own assumptions, an approach whose ineffectiveness is made glaring by just how long the “debate” is going on. You are never going to win that argument because every time they are painted into a corner, they are going to paint a door on the wall, open it and get out.

    I wish Austrians (like you) get more aggressive on the epistemological front, which is where their real strength is vis-a-vis other schools.

    • Dan says:

      There is a ton of material out there that does exactly what you are asking for, but taking on someone on their own terms can be beneficial, as well. Dr. Murphy is great at showing that even on their own terms his opponents are wrong. Tom Woods is great at using the constitution against his opponents even though he’s an ancap and doesn’t have any love for it. We need to be able to take on our intellectual opponents on all fronts. Division of labor, and all.

      • Bala says:

        Dan,

        I do know there is a ton of material out there (I too learnt by reading that material), but what about the semi-ignorant masses who believe in the pseudo-intellectual pronouncements of the mainstream? Wouldn’t you want them to be on your side (at least eventually)?

        The problem is that all these pseudo-intellectuals have the masses believing that the Dipsy Doodle world they are painting for them to see is the real world indeed. That’s what enables them to paint a door to open whenever cornered. Unless you destroy that illusory world, you wouldn’t make too much headway in these discussions. I present, as my evidence, Lord Keynes.

        Incidentally, I have the same problem with arguments for anarcho-capitalism based on the constitution. I usually switch off from such discussions because I am quite certain that they will end up being pointless.

        • Dan says:

          OK, since you know where to find the material that says what you think needs to be said, then you can link the ignorant masses to that material, right?

          Seems like you have a very narrow way on how you think these issues should be discussed, and your upset that other Austrian/libertarians use different strategies. My advice would be to become famous and do it your way.

          • Bob Murphy says:

            My advice would be to become famous and do it your way.

            Dan you can get away with saying that…

          • Bala says:

            Dan,

            My evidence just proved me right again. Just scroll down to see for yourself.

  5. Guillermo Barba says:

    Friedman said that if you want to get a deeper and more powerful knowledge, supposedly to ‘fix’ the economy, you will need to use empirical scientific methodology. Hayek called that: ‘scientism’, or the pretense of knowledge.
    In terms of predictions, which methodology has proved more powerful: econometric modelling or Austrian?
    We just need to remember the predictions of Fisher, Samuelson, M.Friedman, Greenspan, Bernanke, etc.

  6. Bob Murphy says:

    David (Friedman), if you’re still reading, this was the email from Mike when we were talking about the logistics:

    =====
    Matt,

    The agreed format is as follows:

    10 minutes each – overview of each of your “schools”
    20 minutes total – back and forth discussion to challenge and clarify key points
    40-60 minutes total – Q&A
    5 minutes each – closing statement

    This is meant to be a friendly discussion more than a hard-hitting debate. Both sides are anarcho-capitalist, but David will be explaining his Chicago School-inspired pragmatic approach to economics and ethics, while Bob will be focusing on the Austrian School-influenced axiomatic approach.

    Best,
    Mike
    ============

    But maybe you were not in the earlier emails where Mike was more explicit that I would be defending natural rights anarcho-capitalism, and you would be defending your consequentalist approach (which you had already laid out at Porcfest).

    Anyway, it looks like there was genuine miscommunication on this point, but I never would have wanted to spend 90 minutes debating just methodology at Porcfest, since I think those people would be more interested in anarcho-capitalism.

  7. james says:

    The funny thing about the “Iowa Car Crop” is that the Japanese in that example are following precisely the opposite of what the ideologue who wrote it recommends.

    I guess they’re not quite as stupid as US “economists” who find satisfaction in facile little thought experiments.

    • Bob Murphy says:

      James can you explain exactly what’s wrong with the logic in his argument? It should be easy for you.

      • james says:

        that should have been “doing precisely the opposite” rather than “following”..

        The Japanese protect and develop their domestic industries.

        The Iowa farmer who sends his wheat abroad whilst Detroit crumbles is feeding foreigners who have little interest in ‘free trade’ dogma, and instead see their industries as important national assets worthy of support.

        Whether you agree with that approach or not, the funny thing is that in the ‘Iowa Car Crop’ example the people sending over the Toyotas in exchange for crops don’t really give a damn about ‘free trade’. They preserve their automobile industry whilst the US makes their cornflakes.

        Actually this is precisely what the US did during the ‘austrian’ golden age of the 1870s-1900s. The dirty little secret of all the austrian economists who praise the ‘long depression’ is that the US imposed import tariffs throughout that period, which managed to keep domestic prices of many goods low whilst exporters gained higher prices from their imports and developed domestic industries.

        • james says:

          *whilst exporters gained higher prices for their exports*

        • guest says:

          Import tariffs CAN’T keep domestic prices low. The whole reason to import is because it would be cheaper to do so.

          Import tariffs keep domestic prices higher than they would be without them, because they don’t have to compete with foreigners.

          • james says:

            you would have thought that, and in most cases that might be true. But in the US at that time import tariffs coincided with low domestic prices and higher export prices for the same goods.

            • guest says:

              Since import tariffs restrict competition, it follows logically that prices would be higher than they would be without the tariffs, regardless of how low those higher prices were.

              • james says:

                not necessarily. The fact that exporters were able to get higher prices abroad meant that they could keep prices low at home whilst still making a profit, perhaps. A large influx of foreign goods might have destroyed domestic industries, reducing incomes and productivity.

              • guest says:

                A large influx of foreign goods might have destroyed domestic industries, reducing incomes and productivity.

                You don’t need as high an income or as much productivity, if you’re getting cheaper goods in return.

                The cheaper imports cause everyone’s standard of living to rise.

                If you’d be inclined, this video addresses this issue well:

                Defending the Undefendable (Chapter 23: The Importer) by Walter Block
                [WWW]http://www.youtube.com/watch?v=PTT_WHyzZ54

              • Ken B says:

                guest, is this a logical possibility:

                in return for a whole range of tariffs I agree to make one of my products available cheaply. I do this to help re-elect the crooked politicans implementing the broad range of protections I seek.

  8. skylien says:

    Can someone explain what is meant with “the physical world” is non-Euclidian?

    I read a bit about it, and as far as I got it, it just seems to be a restriction which says that Euclid’s laws apply on plain surfaces only (and if I understood it correctly on cylindrical ones as well). Only on the surface of spherical ones they don’t hold. But Euclid never even defined them that way, did he? I mean a “straight” line on the surface of a sphere is no straight line. On the surface of a sphere there are just no straight lines at all (else it would not be a sphere). Only through the sphere you have straight line.

    So since you still have plane surfaces and proper straight lines in the physical world I would argue that the physical world is neither exclusively Euclidian nor exclusively non-Euclidian but both. You just need to use the right geometrical laws for the specific application you currently need. So it seems to me it is not to be about a new superior theory replacing an old wrong one (like it sounds to me when people speak of “Euclidian” VERSUS “non-Euclidean”), but just an addition to the old theory, which means you need both to understand the world. Am I wrong with this understanding?

    • Bob Murphy says:

      No Skylien that’s not the claim they’re making. You’re right, if people 500 years ago took a piece of cloth in the shape of a right triangle, and then laid it onto a bowling ball, then it might appear to violate the Pythagorean theorem, but only because it would no longer be a right triangle. So that’s not what people mean when they say the actual physical universe is non-Euclidean.

      It’s more like, if the universe is a closed system, then if you went out in a straight line, eventually you’d end up right where you started. That makes no sense in a Euclidean universe; you should just keep going forever.

      • skylien says:

        Ok, thanks. That makes more sense now. So it just means that the universe as it may exist might be in some way “spherical” like the earth and we can only move on the surface of it but not through or away from it. So what appears to be a straight line to me is in fact no straight line at all, if you could observe it from outside of this closed system (which we can in terms of the earth but not the universe).

        If this were true then I think it is hard to measure it as well. I mean I can’t see how someone would develop measurement instruments that operate outside of our universe (What point of reference would you have to calibrate it to?). And if I really had the ability to travel in a real absolute straight line, I would necessarily leave the universe to wherever that may be then, maybe “non-universe”.

        In short “non-Euclidean” means we are all moving in a big world of PackMan, which PackMan can’t leave as well. However in so far as PackMan is concerned it is still correct from him to say 180° is the sum of the angles in a triangle on what appears to him to be a plane surface. That is exactly what he will measure; even if space is bent also his set square is bent well, right?

        • skylien says:

          *..it is still correct for him to say…*

    • Bob Murphy says:

      Skylien try reading this article.

      And here I’ll paste in an email from Marco B., reprinted here with permission:

      I watched your debate with David Friedman and found it very interesting. I want to add something regarding the “dispute” on geometry as analogy to praxeology.
      Your Euclidean geometry argument :
      “In geometry you go through and you learn stuff about Pythagorean Theorem and so on and that’s all done from axiom and you deduce it logically. If a student in class said “well let’s go out and test to make sure that triangles’ angles add up to 180°” or to say “let’s make sure that given these assumptions the Ipotenuse adds to what we think that should do” the teacher would have to explain that “nono you’re misunderstanding what geometry is, that’s not how geometry works”
      David Friedman response:
      “I was particularly struck by his mistake in physics, the fact the apparently believes that the truth in Euclidean geometry can be known without empirical tests where as in fact where is not known by empirical tests whether or not the universe is flat; think of the surface of a sphere where plane geometry doesn’t work, add another dimension and you get a three dimensional model of the universe in which the three angles of a triangle do not add to 180° and in fact there are parts of the universe where it’s true so that’s a nice example where you can form very plausible conjectures on the basis of an elegant system of pure theory called Euclidean geometry but you need empirical tests whether to find out if they’re correct”

      Well, you’re correct and David is not. Mathematicians did not develop non Euclidean geometries because someone went out, measured the sum of the angles of a triangle drawn in a spherical surface and found that it didn’t add up to 180°. What they did was something else.
      When Euclid wrote his Elements he started with four postulates:
      1. A straight line may be drawn between any two points.
      2. A piece of straight line may be extended indefinitely.
      3. A circle may be drawn with any given radius and an arbitrary center.
      4. All right angles are equal.

      Then he had to add a fifth one
      5. If a straight line crossing two straight lines makes the interior angles on the same side less than two right angles, the two straight lines, if extended indefinitely, meet on that side on which are the angles less than the two right angles.

      Or in the equivalent form
      5. Through a given point, only one line can be drawn parallel to a given line.

      But he had mixed feelings about that “postulate” and actually he tried to avoid using it in his proofs as much as possible. That postulate was attacked from the start since, as one of his commentators, Proclus, wrote “This postulate ought even to be struck out of Postulates altogether; for it is a theorem…”
      But if it’s really a theorem then it requires a proof and so a lot of mathematicians tried to prove it (in the original or an equivalent form)… and failed. (e.g. Saccheri quadrilateral proof)
      But then it gets interesting. Since the theorem/postulate cannot be proven true, some mathematicians tried to develop a new geometry starting with its negation. The first one was Gauss who wrote in a letter:

      “The assumption that (in a triangle) the sum of the three angles is less than 180° leads to a curious geometry, quite different from ours, but thoroughly consistent, which I have developed to my entire satisfaction.”
      He did not published his works so we had to wait till Lobachevski who developed his geometry on the assumption that:
      Through a given point, there are more than one parallel line to a given line
      This formulation is equivalent to Gauss’ one. No empirical study – Lobachevski did not measure angles in a hyperbolic space – just theoretical work.
      And finally we have Riemann and the elliptical geometry:
      If l is any line and P is any point not on l , then there are no lines through P that are parallel to l
      Most important, Riemann wrote:
      “The value of non-Euclidean geometry lies in its ability to liberate us from preconceived ideas in preparation for the time when exploration of physical laws might demand some geometry other than the Euclidean.”
      And in fact, when Einstein’s general relativity challenged the physics model used till then, the mathematics and geometry to describe it was already there, developed on pure theoretical, deductive reasoning, not empirical testing.
      Kind regards from Italy,
      Marco Bolletino

      • skylien says:

        Thanks!

      • skylien says:

        After reading this article, I guess I really have to understand relativity theory to understand what is being said there.. It seems that they can “see” with telescopes that the univers is non-Euclidean.. So maybe the set square is not bent as well?

        • Ken B says:

          Read Gravitation, available from Dover by Gibilisco. It’s a good entirely non technical discussion. Or Flatland, by Abbott, which is a lot of fun, and not about relativity but illuminates some of the ideas related to geometry.

          Anything to keep you from reading Rothbard skylien 🙂

          • skylien says:

            Thanks for the tip.

            I just read Keynes and there is for a long shot no Rothbard in sight since I already understand where he is coming from and it doesn’t make much sense to only read stuff you agree with or already understand (which doesn’t mean I entirely agree with Rothbard nor that I entirely understand everything I read of Rothbard so far). Although there is the urge to just confirm you own bias I know that I need to read stuff I don’t agree with or don’t understand yet, if I want to be smarter some day.

      • Ken B says:

        Nothing in this completely accurate recapitulation of the history contradicts what David said.

        • TheDjinn says:

          He views “believing that the truth of Euclidean geometry can be known without empirical tests” is a “mistake in physics”; except the “accurate recapitulation of history” notes the truth of Euclidean geometry and even non-Euclidean geometry was discovered without empirical tests, but rather solely with thought experiments.

          He concedes that “you can form very plausible conjectures on the basis of an elegant system of pure theory called Euclidean geometry” but then asserts “you need empirical tests whether to find out if they’re correct”… except according to Mr. Bolletino, non-Euclidean geometry was developed when people began to suspect the truth of one of it’s axioms and not when any empirical test was performed.

          The truth of theoretical systems can’t, by definition, be debunked via empirical tests. One can perhaps question the practical applications of the theorems if empirical evidence calls into question one of the fundamental postulates… but if the postulates are questionable from an empirical standpoint (and the goal is to model reality ) then they aren’t proper postulates.

          It’s not really an empirical critique but a logical one: postulates need to be tautological (1. A straight line may be drawn between any two points) or definitional (2. A piece of straight line may be extended indefinitely) if you intend to ascertain any knowledge of reality. If your goal is just to explore a hypothetical system under intentionally untrue postulates, then you might be doing something intellectually intriquing but that shouldn’t be confused with acquiring knowledge through deductive reasoning.

  9. ChrisW says:

    It seems to me that Mises dedicated a big part of Human Action to answering the a priori question because he had been and knew he would continue to be challenged on it. Now it’s been close to 60 years since he published it and I don’t know how many since the original German version and the same arguments still pursue Austrians.

    I would flip the arguments around and say that we are happy to believe that the core tenet of praxeology starting with purposeful action and continuing through all it’s derivative corollaries are good enough for us because in our minds they are derived from airtight thought experiments, But for those who can’t bring themselves to give genuine consideration to, what they consider, our heterodox opinions, then challenge them to falsify the praxeological conclusions. Tell them that they are free to consider them just opinions and theories rather than a priori and to go to town experimenting and challenging and doing what they do.

    David Friedman did not care to attack praxeology on its merits, He only wanted to argue that the conclusions should be experimented. His mind works and has been trained that way and he has what he believes to be too much experience and positive outcomes from thinking that way. Nothing we say can talk him out of that. When confronted with a different idea I have to analyze and process it in my own distinct way and I think everyone else does the same.

    We don’t lose anything by encouraging him, or anybody else that wants to, to test our premises any way they want. It doesn’t mean we have to. We can instead just challenge and then steer their results based on our core understanding. We may all gain something from the exercise.

    • Ken B says:

      “We don’t lose anything by encouraging him, or anybody else that wants to, to test our premises any way they want. It doesn’t mean we have to. ”

      Those not.

  10. Kyle says:

    It sounds like you guys agree: an economist needs an a priori framework before he can grapple with real world events and figures. Where the disagreement seems to lie, I think, is how to refine that a priori framework. Mises and Hoppe seem to suggest pure rational thought is the only way, making sure the logic stemming from the action axiom is sound. Friedman seems to be suggesting that one should empirically refine this apriori framework, but only with regards to those simplifying assumptions which go beyond his “rationality assumption”. I can’t tell from his words if that is on the table for empirical reevaluation. On the other hand, Austrians don’t tend to make any further assumptions beyond the action axiom.

  11. Lord Keynes says:

    “Going further, to the extent that you might abandon your teenage belief in free trade, it would again be due to a thought experiment, a simplified tale showing how (say) a big country could affect the world price of cotton and so maybe a small tariff could, in principle, be welfare-enhancing.”

    The counterarguments against Ricardo are not pure “thought experiments”: they are heavily based on empirical evidence and demonstrations of how the real world does not conform to Ricardo’s theories and assumptions.

    • TheDjinn says:

      ” they are heavily based on empirical evidence and demonstrations of how the real world does not conform to Ricardo’s theories and assumptions.”

      Which is why they’re debatable on methodological grounds to this day, whereas Ricardo’s original thought experiments (or perhaps, the subsequent reformations and revisions by him and other authors) are largely inarguable.

      Your assertion that “the real world does not confirm to [Ricardo’s] theories and assumptions” begs the very question of the debate: whether you have a valid tool of measurement with which to obtain objective data regarding the “real world” with which to address his particular claims through an empircal process rather than on its own merits with a deductive process.

  12. The Austrian Brotherhood says:

    How’s that Tull treatin’ ya?

  13. Ken B says:

    The real dichotomy is between those willing to be wrong, and those not. Science is marked by a willingness to be wrong. This is not because science is empirical; so is mathematics, which is not empirical. Mathematicians respect counter examples; mathematicians are willing to wrong. Gender equity theorists not so much.

    Many of the commenters here fall firmly into the ‘those not’ camp. They deny that anything can refute their notions. So do many of the religious believers here, who even when faced with clear absurdities or contradictions just shrug and intone “with faith all things are possible” or “we cannot comprehend god’s plan.” In other words, logic does not apply; I cannot be wrong.

    • Bob Murphy says:

      Ken B. wrote:

      The real dichotomy is between those willing to be wrong, and those not.

      This struck me as hilarious, since I’ve often wanted to ask you, “Ken, have you ever been wrong in your life?”

      • Ken B says:

        Some might see things a bit differently Bob. If I challenge your claim it is because I am implicitly acknowledging your claim might disprove mine.
        (That yours rarely do is a happy circumstance.)

    • skylien says:

      Ken B,

      How do you determine if someone is willing to be wrong?

      I mean since you say “many of the commenters fall firmly” here, you must have a way of knowing this…

      • Ken B says:

        Well I could be wrong. Lets just take the believers. But Tanous for example was willing to have Jesus crucified twice rather than admit the gospels differ on the day, and Murphy explicitly uses the “we can’t fathom the mind of god” stuff, which in plain English translates ” yup you caught me in a contradiction, so what?”

        • skylien says:

          I mostly don’t follow the Sunday post discussions, since I don’t see a point in (non)believing something you just can’t know anyway, and even if you could never prove to someone else. I assume that whoever has a strong position in such a topic necessarily must be caught up in a contradiction this way or another since there is nothing to prove or self-evident or axiomatic, there is just no basis what so ever.

          And since especially religion is a very emotional and personal matter I would not interpret too much in it. I would not take a discussion of religion as prove that someone is generally not open to be wrong on different subjects.

          • Ken B says:

            Well you are right about the different subjects.

        • skylien says:

          And I have to add with “Well I could be wrong” you did relativize the tone of your former comment significantly..

          😉

    • Tel says:

      Mathematicians respect counter examples; mathematicians are willing to wrong.

      I would have thought that the whole idea of mathematics is that once a theorem is proven, it stays proven and you can move onto other stuff. Admittedly mathematicians regularly argue about what constitutes proof, but they agree that proof is final.

      • Ken B says:

        Yes, but my point is different. It’s about an approach that does not privelege one’s own intuition and insight. Mathematicians submit to a pretty clear standard. No mathematician will put a stake in the ground and clim his idea is right unless he can prove it, and will accept a demonstration that he in fact erred. Landsburg had a couple major mathematicians who claimed (absurd) results and then, when the error was specified backed down. Andrew Wiles withdrew and later corrected his proof of Fermat’s Theorem. The point is that unlike some in say Literary Theory, mathematicians, like chemists, are prepared to be refuted.

        • skylien says:

          I think that the logic in math is so rigorous and definite that a prove or disprove is much more clear than say in economics. I mean look at 2+2=5. It would be really insane to argue this was the really true result. But tariffs are bad or good doesn’t look that conclusive right from the start, does it?

          So I think the difference is not to be found in the character of the people doing economics or math. It is the subject that is vastly different and therefore leads to different results in debates and also debating styles.

          People aren’t more smug in economics because they are, but because the subject allows them to get away with being more smug without looking that foolish/self-righteous and ignorant than say in math.

          So mathematicians submit to pretty clear standards because it is possible in math not because they are better persons on average.

          • skylien says:

            *… real true result..* …

          • Ken B says:

            I didn’t say better. I said willing to be wrong. And MOST economists are also willing to be wrong. But some are not.

            I am dichotomizing people not subjects. But different types of people choose different subjects. Compare the approach to evidence and argument in a group of literary criticism students and a group of computer programmers and you’ll see what I mean.

  14. Demosthenes says:

    I can’t believe I’m saying this… but someone needs to read Mirowskii’s More Heat than Light.

  15. neutrinoide says:

    Nobody believes in free trade because of natural experiments. Errr I do!

  16. Bill Karr says:

    As it happens, I am a libertarian who is very interested in Austrian school economics and I’m also a PhD student in mathematics. My particular field of study is Lorentzian geometry, the mathematical foundation of general relativity.

    Bob, you ONLY mentioned geometry when you first described your position and David, of course being the empiricist, brought up physics. To do *physics*, one must formulate some sort of empirical hypothesis of the form, “our universe is modeled by a geometry based on such and such axioms” and then test that theory. But to actually DO geometry has absolutely nothing to do with going out into the universe and start measuring things.

    So we should be sure not to confuse mathematics with physics. And perhaps analgously, not to confuse economic theory/law with economic history? But then I think the point of an empiricist approach to economics is to say that if you develop your theory on an axiomatic system which has absolutely no bearing on what we observe in the economy itself, then it is tantamount to mental gymnastics.

    I don’t know. I love this type of debate. Fun stuff to think about.

    • Major_Freedom says:

      “To do *physics*, one must formulate some sort of empirical hypothesis of the form, “our universe is modeled by a geometry based on such and such axioms” and then test that theory.”

      Wait, doesn’t a theory, which is necessarily proposed sometime in the past, that is “tested” in the present, imply that what was true in the past about the universe is still true today? Where did that idea come from? It’s not capable of being falsified or confirmed using the empiricist approach, because empiricism itself is a priori structured to assume constancy over time. But if that’s the case, then doesn’t empiricism contradict itself?

      “But to actually DO geometry has absolutely nothing to do with going out into the universe and start measuring things.”

      Who says we have to go out? We are already there by virtue of thinking. Perhaps we can go out by identifying necessary truths about thinking as such.

      “But then I think the point of an empiricist approach to economics is to say that if you develop your theory on an axiomatic system which has absolutely no bearing on what we observe in the economy itself, then it is tantamount to mental gymnastics.”

      So the assumption that what is true in the past is still true today, which is a truth about the structure of empiricism, that idea is “mental gymnastics”?

  17. Bill Karr says:

    Bob, I loved the walkie talkie example! Great stuff. I’m going to use that myself next time this topic comes up in conversation.

    • skylien says:

      Yes I agree, I’d loved to hear from David F. if that was a sufficient explenation or not, and of course if not then why not..

  18. Keshav Srinivasan says:

    Bob, I have a question about praxeology, since I know pretty much nothing about it: you keep making comparison’s to Euclidean geometry, but has any Austrian attempted to present their reasoning as precisely as Euclid did, I.e. with axioms, theorems, and rigorous step-by-step proofs, as opposed to just vague arguments which they claim can be made precise? Has either Mises or Rothbard done this? Which

    Also, has anyone managed to put praxeology in symbolic form, the same way that mathematics has been?

  19. Epiphyte says:

    What do you think would happen if taxpayers could choose where their taxes go? I’ve asked David Friedman the same question and would be really interested to compare your answer to his.

    • Dan says:

      I’d choose for my taxes to go right back in my pocket.

      • Epiphyte says:

        Dan, we know what the demand for milk is…right? This is because people have the freedom to demonstrate their preference for milk. But do we know what the demand for coercion is? Nope. Why not? Because people do not have the freedom to demonstrate their preference for any “public goods”…including coercion.

        So if we implemented tax choice, how many taxpayers would give their taxes to the IRS? A few, many, most?

        If many, or most would give their taxes to the IRS then Rothbard was really wrong about how many people want to rob their neighbors. Good thing we didn’t implement anarcho-capitalism.

        If few people would give their taxes to the IRS, then couldn’t the argument be made that the benefit was too narrow for coercion to be considered a public good? Yup. Therefore, the result would be anarcho-capitalism.

        So the first step is giving taxpayers the freedom to demonstrate their preference for coercion. Then we’ll take it from there.

  20. Julien Couvreur says:

    By the way, we’re having a post-debate discussion on methodology on David Friedman’s site as well.

    http://daviddfriedman.blogspot.com/2013/08/my-porcfest-talks.html

  21. Michael Duff says:

    I finally watched the Friedman/Murphy debate in full, and I’m trying to understand the distinction between the two positions.

    But all I see is two guys talking past each other.

    Friedman seems to be saying that Chicago School folks use a combination of theory and data to reach conclusions while Austrians just use logic.

    Of course, Chicago people have theories that they rely on, and Austrians consider data when they make arguments.

    I’m a word guy, so I think of axioms in terms of definitions instead of math.

    Austrians say the fundamental laws of economics are axiomatic, kind of like the definition of a word like “chair” is axiomatic.

    A chair is something you sit on. That is the logical definition of a chair.

    But you still need to look at data to determine if your definition applies to any particular object in the world.

    There’s an object in the corner over there. We need to look at it and determine if it fits the Austrian defintion of “chair.”

    It looks like a chair, but it may be a cardboard cutout. It may be a mirror image. It may be a hologram. Or it may be so poorly constructed that it falls apart when you try to sit on it.

    Any particular object may fail to meet the definition of “chair” but that does not change the axiom. That does not change the fundamental definition of what a chair IS.

    If you can’t sit on it, it’s not a chair.

    Similarly, when Supply goes up, price goes down. If you see a chart where that does not appear to be true, that doesn’t change your definition, it means there is something else going on.

    You’re not turning a blind eye to the real world, you’re just starting from a logical definition and identifying when reality deviates from that law.

    For example, Paul Krugman is telling everybody that is a chair, but an Austrian will point out that it only has three legs.

    If someone tries to sit on it and falls, that doesn’t change the definition of a chair. It means this object failed the chair test.

    And here’s one more point about the predictive power of economics. I compare the current financial crisis to a medical diagnosis.

    If you see a guy who’s 500 pounds, binging on donuts day and night, you can say logically, this person is engaging in dangerous behavior and is likely to have a heart attack. But you can’t predict the exact day and time of the heart attack. There are too many variables.

    You know something bad is going to happen because of this dangerous behavior. You can point to history, or you can just make a logical argument based on how the human body works.

    I know this is philosophy 101 here, but I felt like this was a philosophical debate that got bogged down trying to be an economics debate.

    Does my chair example work as an analogy for Austrian axioms?

  22. Will Porter says:

    Alright, just my 2 cents here, but I’m under the impression that Mises’s entire point forever about economics and scientific prediction was that you **cannot** take statistical data about past human behavior and use it to make predictions about the future.

    Since human action is guided or propelled or navigated by human knowledge and valuation (subjective), it cannot be predicted by any hypotheses which assume a Causal regularity in the concatenation of phenomenon (in Mises’s words). Human action is not a Causally-guided phenomenon, but a Teleological one, guided by “purpose”.

    The only scientific understand about economics we can come to must be via a priori axioms, which can be used to parse the data and figure out what happened. We can do ceteris paribus thought experiments and reason out the logical conclusions, and then take these and apply them to historical economic data.

    We *illustrate* the theory in history. But if something occurs which is at variance with our logical conclusions, we must assume that all other things were, of course, **not** equal.

    Here also lies the seemingly impossible task of constructing an economic laboratory in which you test economic propositions. Friedman has said video games and World of Warcraft, but I dunno, I remain skeptical on that front.

    Anyway, maybe I said something cool, maybe I’m completely insane. The jury’s out.

  23. Colombo says:

    It seems to me that academics tend to create problems by insisting that there can be only one valid methodology.
    I understand that one may prefer to use only one methodology for his own work, perhaps tweaking the methodology as you go. But other people may want to use different methodologies. Humans use different calendars, languages, literary topics, poetic forms, numeration systems, religions and theologies, mythologies, laws and sources of law, familiar structures, cooking, clothing, production of food and goods. Why should humans use only and only one scientific method, especially in a field such as economy?

    More options can be a good thing in scientific research and philosophy. There is no need to restrict creativity in thinking.

    In my opinion, all philosophical controversies look like a fight of bucks over a doe. The doe of philosophers is clearly the awe of lay people, and perhaps some form of tenure. It is not knowledge or truth what they want, but the feeling of being useful and being admired. I’m not attacking either Friedman or Murphy, my indictment is directed toward every single philosopher, scientist or economist in history who have wasted time in futile endeavors.

    • aby says:

      Are you a time traveller?

      • Colombo says:

        Time travelling does not require Government passport!

Leave a Reply to skylien

Cancel Reply