Humanly Rational Choice

As a follow-up to Steve's post on how Mises can be read to have anticipated Smith's concept of "ecological rationality", and also to counter a current prevailing trend to trash rationality and embrace irrationality, I would like to point to a sentence of the philosopher David Schmidtz discussing his essay "Choosing Strategies".  "To identify genuine, substantive connections between being rational and being moral, I now believe, we will have to work with humanly rational choice and humanly moral agency as they really are, not with the mathematically tractable idealizations of them."*

Schmidtz's book Rational Choice and Moral Agency is brilliant, and in fact one of my favorite books in philosophy.  The fact that Schmidtz has always seemed to carve out the intellectual niche in his profession that moral theorizing and political theorizing is impotent unless it takes seriously the world in which we live both in terms of social relations and the constraints of the natural world has always struck this political economist as the most level headed in the genre.

I am fond of saying that "praxeology puts parameters on people's utopias" and David's use of humanly rational choosers in his approach to moral and political philosophy is consistent with this understanding of praxeology even though David would not necessarily draw that same conclusion.  The basic point is that study of purposive human action need not entail maximizing agents, but cannot make use of such a stripped down model of man, if an accurate picture of man and society is to come out of the analysis.  Why else do you think Mises spends so much time and mental effort in the first 100 pages of Human Action, trying to establish for the reader the uncertainty of the environment, the ignorance of the future, and the open-ended nature of the human choice problem?**  Mises is a rational choice theorist, but one that focused on "humanly rational choice" and not "mathematically tractable idealizations" of choice.  And his entire analysis from the logic of human action to the dynamics of the entrepeneurial market process is dependent on this human, as opposed to mechanical, perspective.

If this is so, then what does this have to say to those critics of the excessive ambitions of rational choice theorizing in the social sciences and how would the social sciences be reconstituted given both the critique and this avenue for reconstruction?

---------------------
*Schmidtz, Person, Polis, Planet (New York: Oxford Univesity Press, 2008), p. 5.

** William Jaffe in his classic paper on the marginal revolutionaries argued that Menger differed from Jevons and Walras and was not vulnerable to the Veblen-type criticism of the "lightening calculator" because to Menger man was forever caught between alluring hopes and haunting fears.  The poetry of Jaffe aside, he hit the nail on the head with respect to the Austrian tradition of economic theorizing --- undestanding "humanly rational choice" not "mathematical tractability" has always been the goal of the enterprise, and sees its most complete defenses in the works of Mises and Hayek.

Comments

[Greg Ransom]

Don't try getting a Ph.D. or tenure doing this -- tractable math constructions or quantitative research are the the pretty much non-optional for the young scholar seeking entry into the professional academic club.

"Humanly rational choice" and "humanly moral agency" are the for tenured only -- and usually by then new thoughts and new ways of thinking like this are impossible -- and are assumed to be anti-intellectual and anti-scientific, and just plain muddle.

Look at how most tenured academics treat Wittgenstein -- or Hayek.

Schmidtz writes:

"To identify genuine, substantive connections between being rational and being moral, I now believe, we will have to work with humanly rational choice and humanly moral agency as they really are, not with the mathematically tractable idealizations of them."*

[Greg Ransom]

The Ph.D. / tenure process demand a binary / on-off metric for determining who's "smart" and who "understands" the stuff and who has "moved the ball".

Only math / logical constructs and quantitative research provide an objective / universal metric for establishing this -- and its EASY.

That's why this is the metric which is used to provide academic credentials, all the way through.

So there is an "essential tension" between the demands of credentialization in economics, political economy, and philosophy, and the road to genuine understanding and explanation in these fields.

And the way the institutions are set up, the credentialization demand almost always wins -- at the Ph.D candidate and tenure track level Hayek gets excluded or misrepresented, Burke gets excluded or misrepresented, Wittgenstein gets excluded or misrepresented, etc., at least in the main, with only a few exceptions, usually at the margins of academia.

[Brian Pitt]

Professor Boettke,

Excellent post and very well said. I think that Mr. Ransom has some very good points about the importance of math and quantitative research - for prospective scholars - for those across the human sciences interested in the rational actor analytic.

(Then again, M. Weber viewed rationality as the preeminent theoretical lens for the human sciences, focused on historical and legal scholarship, and I have yet to see anything he has written that comprises a regression equation or an equation for the slope of some curve.)

*Of course, very few human science scholars incorporated these into their projects prior to the 1930s. And, M. Weber was on the academic scene prior to Samuelson, P. Lazarfield, etc. But those scholars who followed the Weberian research program, e.g., F. Knight, R. Merton, J. Buchanan, Mises, Hayek, Roepke, etc., seemed to have excellent academic careers with very few mathematical models and/or regression coefficients.

Anyway, what do you think about what Mr. Ransom has written?

[Peter Boettke]

Brian,

At the recent IHS conference I spoke out, a question was asked whether the discrimination against free market and classical liberals was so great that a career in academia was hopeless. I responded that this idea was the most pernicious one circulating in our circles and had to be resisted. Yes, academia is a hard business, but the biggest criteria for discrimination is "bad argument" not one's ideology.

That said, Greg has a point that methodological issues do matter. They form the expectations around which an argument must be made in a certain discipline. But we probably should try to understand why the dominant methodology is in place. With respect to mathematical formalization, I would suggest that is there because ambiguity in argument causes confusion and endless debates. A major source of ambiguity is that we can use the same word to mean different things, or different words to mean the same thing. Mathematics supposedly gets rid of this problem. But in the human sciences is does not --- it provides syntactic clarity, but does not provide semantic clarity. Instead, it produces an error of "mechanomorphism" in the sciences of man.

Now admitting that part of Greg's argument, the question then becomes whether good economics can still be done within this methodological milieu. Here I think that answer is YES, you can do this work but it will be harder than alternative paths. I think of this as being a 5'10"-6'0" guard competing in the NBA. It is not common, the guys that do it are exceptionally skilled and quick, but it is not impossible. They had to learn early on how to compete as an undersized player.

But we should recognize that there are many more 5'10"-6'0" kids who dream of playing in the NBA than there are 6'7"-6'9" kids. And the probability that a 6'0" tall kid will make it is that much smaller.

Those that don't make it, were not discriminated against, as much as selected out as not being able to compete. Some of this is also luck, but most of the time it is a combination of athleticism, skill and perhaps most important ATTITUDE.

My point of this analogy is to suggest that the same is true in academia. Yes the game is tough, yes the game is tilted against you if you are not working with conventional methodological tools, but NO it is possible to have a very successful career working as an out-of-sync social scientist. But the more out-of-sync you are the better you have to be to stay in the game. Muggsy Bogues was 5'3" or 5'4", he made it all the way to the NBA from Dunbar HS in Baltimore through a great college career in the ACC playing at Wake Forest. So it is not impossible for a short player to play in the NBA, but it is rare, and he had to be amazingly skilled and quick.

Verbal economists can make it, and they need to be encouraged to try (you cannot succeed unless you try), and they need to take advantage of every opportunity that reveals itself because they might be few and far between. Wishing the world was different is not going to make things better. But running away from the world because it is what it is, is also not going to change things. And even if your goal is to set up an alternative NBA for people 6'0" and under, you might have some good games at the local recreation center, but the fans would still head to the arena and the TV contracts and Radio coverage would be limited to the "real" game being played at Madison Square Garden, or the Staples Center, etc.

Does that make sense?

[Greg Ransom]

I'd put it this way.

The difference between Hayek & Wittgenstein versus, say, Lerner and Carnap or Samuelson and Quine, is a difference in understanding of how the explanation goes, what role models/constructs play, and the place of an explainer's "background understanding" and own mind in the explanation of social phenomena like market order or natural language.

All of them have used math and logical constructs. All of them have used verbal argument. The difference is how they make sense of these things -- and how they use them to explain social phenomena, e.g. market order, verbal significance, and formal construction and proof itself.

Now, being better at the formal stuff can make one better and understanding and figure out what the significance of these things are from the Hayek or Wittgenstein point of view, e.g. new and more powerful math constructs in economics shed light on Hayek's work in different ways, as do different formal advancements in logic and mathematics in the case of Wittgenstein.

In order to be the most powerful academic you can be, by all means make yourself into a 7'0" -- but what economics and philosophy need are more people who play a double game. People who get the Hayek and Wittgenstein stuff, work on it, and can put the 7'0" of height to some productive use.

[Roger Koppl]

Pete,

I really wish we would adopt a more math-friendly attitude! Economics is about the unintended consequences of human action. You often want to refer to the meanings of the human actions that generate a spontaneous order, and that may require you to use hermeneutic methods, not mathematical models. By definition, however, the overall order has, in Schutz’s phrase, no “meant meaning.” Thus, we often can and sometimes must describe the overall pattern without the help of hermeneutic methods or categories. We often can and sometimes must use math to describe the explanandum of an invisible-hand explanation. We often say (perhaps quoting Lachmann or Nozick) that we “reduce” phenomena to the level of human action, but not beyond, because meaning exists at the level of human action. Why don’t we take our own statement seriously? If we did, we would know that meaning does not exist at the level of spontaneous order. Besides, sometimes a mathematical description of human action is sufficient, as in elementary supply & demand analysis. If you want to know what a tariff will do to the price of domestic tradables, good old partial equilibrium analysis will do the job. Or think of Gode & Sunder’s zero-intelligence traders. Thus, it’s perfectly fine if we address some problems in economics using the mathematical tools of “neoclassical” economics or mathematical complexity theory.

When we wring our hands and talk about how hard it is to be a verbal economist in a mathematized profession, I think we needlessly ghettoize ourselves. That’s just not healthy. I completely agree that the profession at large overvalues math and undervalues hermeneutics, but there is nothing “un-Austrian” about mathematical social science as such. So please, please, please, let us stop classifying ourselves as somehow extra-mathematical beings.

[Peter Boettke]

Roger,

If you think I was ghettoizing Austrian economics, I must have written a very UNCLEAR argument. I am sorry. But if you look back, I am advocating playing in the NBA, not in the Euro-League let alone the local recreation league as an analogy. So I actually don't understand your interpretation.

As for the role of mathematics in economic science. Please again look at my point about ambiguity in my comment, and my point about the first 100 pages of Human Action in my original post.

The bottom line --- humanly rational choice is not so easily captured in a mathematically tractable form. Let's recognize that and go forward.

As for our dispositions --- I follow mathematical reasoning and see its limits for the sciences of man even if done by the best practitioners, and you follow mathematical reasoning and see the potential for the sciences of man if done right. We have hashed this out before, I see in the arts & sciences divide, promise in the arts, you see the promise in the sciences. Your position is the one with more evidence on its side, mine is the miniority position.

[Roger Koppl]

Pete,

I completely get the NBA metaphor and I have chimed in with you on that in the past, as you know. I still support that attitude, of course. I confess I felt exasperated, but by something else in your post and in the comments of countless others since the Austrian revival began. You even said it just now:

“The bottom line --- humanly rational choice is not so easily captured in a mathematically tractable form. Let's recognize that and go forward.”

*That* is the ghettoizing remark IMHO. And it is not really right. At some level, sure, it’s true. But we don’t always want to capture the aspects of human rationality that have so far defied mathematical treatment. And – again – when we move from the level of human action to the level of unintended consequences, hermeneutics becomes much less important and abstract patterns – the business of the mathematician – become more important. Wouldn’t you accept my example about the price of domestic tradables? (I think I borrowed in unconsciously from Machlup BTW.)

I see “promise in the arts” too, Pete. I’ve written several papers in support of hermeneutical social science, including a co-authored piece in JEBO. I just want us to keep a foot in both camps. As you know, I harp on “computable economics” because I think that it shows us why we cannot vacate the “arts” camp altogether. There is a famous quote saying, "A little knowledge may lead a man away from God but a great deal brings him back" or something to that effect. (Couldn’t track it down on Google. Can anyone pin it down for me?) I think math is a bit like that. A little mathematical knowledge leads you away from hermeneutics, but a great deal of mathematical knowledge brings you back to hermeneutics. We should not try to keep a place for hermeneutics and the “arts” by simply planting a flag and saying “This is where I stand!” At least we should not do so when we can *mathematically demonstrate* the impossibility of “reducing” human action to “objective” language and when we can show that common mathematical models assume agents can compute the uncomputable.

As Bartholom et al. note in a recent JEBO (http://linkinghub.elsevier.com/retrieve/pii/S0167268108002060) there is this weird way in which even finite games may be uncomputable, so this stuff really, really matters for economics. It makes it really hard to sustain the old fashioned Burbaki math of Debreu and others and it makes it hard to avoid the return to hermeneutics. At least it would if we would only seize the opportunity to drive the point home.

[Lee Kelly]

"We often say (perhaps quoting Lachmann or Nozick) that we “reduce” phenomena to the level of human action, but not beyond, because meaning exists at the level of human action. Why don’t we take our own statement seriously? If we did, we would know that meaning does not exist at the level of spontaneous order" - Roger Koppl

Perhaps I am just terribly uneducated, but this is one of the most bewildering comments I have read all year--maybe a mathematical formulation would make better sense.

[Rafe Champion]

von Mises wrote in "Notes and Reflections" 'It [AE] rejects the mathematical method, not because of ignorance of mathematics or aversion to mathematical exactness, but because it does not emphasise a detailed description of a state of hypothetical static equilibrium'.

Was that his last word on the matter?

Of course times change, he also wrote somewhere that you can't do experiments in economics, he probably meant controlled experiments because in his capacity as a governemnt advisor he was doing, or least attempting, experiments in policy practically daily for years.

Taking up Roger's point, surely the thing is to know enough maths to use it when it is helpful and appropriate and to recognise when it is not helping. Perhaps the main thing is to have ideas that are worth testing, then mathematics and hermeneutics and all of the other technology of research and investigation can be seen as a part of the battery of tools to draw on.

Of course Roger and others have demonstrated that Austrians can do math as good as anyone!

http://austrianeconomists.typepad.com/weblog/2007/05/the_amplificati.html

For a crit of abuse of maths in hard science check out J Schwartz ‘The pernicious influence of mathematics on science’ in “Logic, Methodology and Philosophy of Science” eds Nagel, Suppes and Tarski 1962. One of his points is that maths demands clearly defined systems and that calls for heroic assumptions about the world. To be appropriately used in science the assumptions must be correctly chosen from a larger point of view invisible to mathematics itself.

[Brian Pitt]

That makes perfect sense. Koppl, Ransom, and Champion made some very astute points.

As Swedberg said somewhere, "Do what you do best, but learn enough [math/stats] to avoid limiting your ability to publish."

[Roger Koppl]

Lee,

Sorry to have been cryptic. Let me try to explain. Apologies in advance if I am still too cryptic. Austrians are always touting methodological individualism. We "reduce" "aggregates" to "individual action." Some critics ask "Why stop there? Why not reduce down to atoms or some other level below human action." Sometimes the question is sincere and sometimes it is a "gotcha!" question. Either way, it's a legitimate question IMHO. One standard answer to this question says that we are seeking the human meaning in events, and such human meanings exist at the level of individual action. I seem to recall reading that Lachmann gave this answer to Hicks at the conference which became the 1979 (?) Rizzo volume. I also seem to recall Pete Boettke attributing the answer to Nozick. It's a pretty good answer IMHO. But if you are willing to give that answer and you accept the idea of spontaneous order, then I think you have to recognize that spontaneous orders as such have no meaning. Only the actions that unintended generate them have meaning. The GE price vector is a result of meaningful actions, but does not itself have a meaning. The volatility dynamics of asset-price time series (as described, perhaps, in a GARCH model) have no meaning even though the individual acts that generate those dynamics are meaningful.

I hope that helps, Lee.

[Roger Koppl]

Also: Like Rafe said.

[KA]

Pete,

This is an important post- and a reaffirmation of Mises and especially Part 1 of "Human Action." I'm struck by Austrians, who know how crucial that that "humanly" proviso of rational choice theory is, yet when communicating with non-Austrians and non-economists especially drop the ball and start sounding like the robotic rat choice neo-classicals we're more insightful than. I'm thinking of all those fruitless debates bewteen economists and philosophers on the nature of rationality at conferences. Why don't our Austrians just say "no, we're not positing a view of man per se; but rather a view of man in society, pursuing and coordinating his actions within the framework of scarcity, constraints and incentives towards the end of spontenously generated social order." In other words, we're not saying anything about the nature of reason, philosophically speaking. Though our methodology is, in some sense, a socio-philosophical one, with deep insights that transcend the neo-classics cruder use of rat choice. Dave Schmidtz gets it. I think if our Austrians kept your post in mind (and Mises' Human Action), and simply stressed that "humanly" proviso when *communicating* these ideas to those not blessed with the experience of having actually read Austrian stuff, our ideas would be driven to the next level among those in the academy across disciplines, and more Schmidtz's may emerge.

[I Need Money Desperately]

"To identify genuine, substantive connections between being rational and being moral, I now believe, we will have to work with humanly rational choice and humanly moral agency as they really are, not with the mathematically tractable idealizations of them."*