Social Science on Wax Wings?

Francis Fukuyama reviews Bruce Caldwell's recent book on F.A. Hayek for the Wilson Quarterly, emphasizing Hayek's hostility, not merely to state economic planning, but to empirical social science in general. It's nice (if a little rich) to see the man who once confidently declared the "end of history" acknowledging the unpredictability of emergent orders. But the critique actually seems too strongly put. It's doubtless true that we can't expect to "fully model" a system, like the market or a human culture, whose "higher functions cannot possibly be inferred from its physical substratum." But the impossibility of a "full" model in that sense, while probably a good reason to be wary of attempts at intervention, scarcely seems to make the case against positive social science per se. After all, complex physical sciences like organic chemistry don't proceed by straight deduction from the laws of particle physics. And increasingly sophisticated econometric and statistical tools make it possible to control for more and more confounding factors. That doesn't mean we can plan economies, but we're not confined to ignorance either. Which is a good thing, because many of the most potent arguments for free societies rest on the very measurable benefits they produce.

Comments

[Jason Ligon]

joe:

Egad! We didn’t just agree on something, did we?

I would think the lesson would be that regulations issued with low levels of confidence in underlying assumptions should be avoided – ergo many regulations on the books should be avoided. I note that your conclusions are different. 😉

Planners exhibit humility these days, eh?

[Gary Gunnels]

Rick Barton,

I can think of one rule; the so-called “iron rule of oligarchy.” It states that any organization, no matter how democratic, etc., will in the end be run as an oligarchy. This rule is used a great deal in interest group study and theory.

[thoreau]

I’m a physicist who minored in economics as an undergrad. I always enjoyed economics classes.

Economics, if properly understood, is a very powerful tool to inform decision-making. Economics provides a lot of quantitative methods for sharpening your thinking, asking good questions, and analyzing information. Economists have compiled a lot of information to formulate some pretty good general observations, and have plenty of useful examples and case studies illustrating the behavior of markets. If you use it to achieve a better understanding of a situation before making a decision it is an excellent tool.

Of course, some people (including some physicists, believe it or not) think that economics can provide predictions based on deterministic rules. I actually had a physics professor who was trying to apply all of his sophisticated mathematical tools to predict the stock market as well as macroeconomic variables like GDP, inflation, etc. I don’t know the latest status of his efforts, but I do know that the founder of Ikea is the richest man in the world, which means that my former professor is not.

Interestingly, I knew very few economics professors who pretended to be physicists. Most of them seemed keenly aware that their field of study is most useful for informing decisions rather than predicting the future. It seems to me that it’s usually the media that portrays economics as being more like physics than meteorology.

Lacking much exposure to other social sciences I will refrain from commenting on the attitudes of other social scientists. However, I wouldn’t be at all surprised if most of them are actually quite aware of the limitations of their work, and that the media overplays their alleged predictive abilities.

[Rick Barton]

Gary Gunnels,

The constants that I was saying have never been discovered in any of the social sciences are of the quantitative variety, such as the Gravitational Constant, the Plank Constant and The Speed of Light in a Vacuum.

thoreau,

From my question in the post 10:21 AM: Do you know if; in the natural sciences are complexity and constants mutually exclusive in some manner?

[thoreau]

Rick-

I don’t know how to approach your question. Fundamental constants are the stuff of physics and to some extent chemistry: Planck’s constant, the mass of the electron, etc. They are the stuff that you find when you strip things down to the most fundamental level.

In more complex systems, however, people find all sorts of numbers that seem to be largely independent of the underlying interactions. For instance, people have found that a wide variety of systems have noise that obey a power law with the same exponent. People have found all sorts of patterns involving the Fibonnaci sequence and the Golden Ratio. And so forth.

[joe]

JL,

“I would think the lesson would be that regulations issued with low levels of confidence in underlying assumptions should be avoided.”

Actually, regulations issued with low levels of confidence aren’t the problem – they usually contain some sort of safety valve to avoid absurb outcomes.

It’s the regulations issued with a high levels of confidence that have done the most damage. Requiring roads to be built with enough pavement for two fire trucks to pass when cars are parked along the sides (even when cars never park on that street) comes to mind.

[Rick Barton]

…emphasizing Hayek’s hostility, not merely to state economic planning, but to empirical social science in general.

From the review:
One can never fully model and predict complex phenomena such as the spontaneous orders produced by the interactions of simpler agents.

A critical point here is that unlike the natural sciences, a constant has never been discovered for the social sciences.

Question: In the natural sciences are complexity and constants mutually exclusive in some manner?

[Jason Ligon]

When social scientists begin acting more like the weathermen they are than the physicists they pretend to be, we will all be better off.

[Shannon Love]

“And increasingly sophisticated econometric and statistical tools make it possible to control for more and more confounding factors”

Well, actually no, they don’t. Reductionism in general does not work well for complex dynamical systems like economies. Learning more about the individual units of a system often tells you nothing about it’s global behavior. In the study of real world human systems, prediction is also defeated by two different phenomenon.

(1) The Red Queen Race: “Now, here, you see, it takes all the running you can do, to keep in the same place.” While our measurement tools and predictive models keep improving the complexity and rate of change in the economy keep increasing apace.

(2) Human awareness: Unlike gas particles or ecosystems, human beings alter their behavior in response to attempts to predict their behavior. Even if we could create a perfectly predictive snapshot of the global economy at anyone time, peoples awareness of the existence of the snapshot would cause them to change their plans. Any predictive model would collapse under a “we know that they know that we know that they know…” sought of feedback loop. Better economic measurement and models become themselves confounding factors to prediction.

Hayeck railed against those in every generation who hold forth that they possess predictive models of large scale human behavior like the economy. Economist, sociologist and the like, have repeatedly portrayed themselves as engineers able to produce any desired outcome at will given enough political power. They lied.

Honest economist have dedicated themselves to mapping the boundaries of our ignorance, separating what we can predict from what we cannot. This is terribly important. As Will Rogers said, “It isn’t what we don’t know that gives us trouble, it’s what we know that ain’t so.”

[joe]

Rick, a chaotic system can result from as few as three interacting variables, according to the James Gleik book.

“When social scientists begin acting more like the weathermen they are than the physicists they pretend to be, we will all be better off.” Hear hear! It took Urban Planning way too long to come to this conclusion, and the country suffered 30 years of Urban Renewal madness because of it. It actually took a group of hippie planning students taking over the national planning conference to get the charter amended to admit that planning is not primarily a technical field.

This is not to say that nothing can be learned or directed by social sciences like planning and economics, but the awareness of these limitations does foster an appropriate flexibility and humility.

[Steve Horwitz]

Julian wrote:

“But the impossibility of a “full” model in that sense, while probably a good reason to be wary of attempts at intervention, scarcely seems to make the case against positive social science per se. After all, complex physical sciences like organic chemistry don’t proceed by straight deduction from the laws of particle physics. And increasingly sophisticated econometric and statistical tools make it possible to control for more and more confounding factors. That doesn’t mean we can plan economies, but we’re not confined to ignorance either. Which is a good thing, because many of the most potent arguments for free societies rest on the very measurable benefits they produce.”

I’m not so sure I buy this. First, Hayek’s vision *does* leave room for positive social science. In his discussions of complexity, he carves out room for what he calls “explanations of the principle” as the sort of predictive work we can do with complex systems. He also used the term “pattern predictions” in much the same way: explanations of the principle enable us to predict patterns. In Bruce’s book, he make an explicit defense of “economic reasoning” in particular by contrast to “economic modelling.” The former is what we can employ to the complex system of the market; the latter is to fall victim to forms of scientism.

Julian’s faith in the increasing sophistication of econometric modelling is not one I share. There is MUCH that can be done with data, both quantitative and qualitative, but constructing evermore sophisticated models is likely to bump into Hayekian limits to reason.

In any case, Caldwell’s book is, in my view, the best book ever written on Hayek and one of the best books on Austrian economics in the post-1974 era. I’ll have my own joint review of that book and Lanny Ebenstein’s second book coming out in the Journal of the History of Economic Thought sometime later this year.

[Quotes on Economics]

One definition of an economist is somebody who sees something happen in practice and wonders if it will work in theory. — Ronald Reagan