|Peter Boettke|
Many years ago now the social scientist Jon Elster made a Plea for Mechanisms in his Alchemies of the Mind: Rationality and the Emotions (1998). I have always been intrigued by Elster as a thinker, in the same way that I have been intrigued by Albert Hirschman or Amartya Sen. These are not thinkers that I draw on for inspiration -- such as Mises, Hayek and Kirzner, or Alchian, Buchanan, Coase, Demsetz, North, Olson, Ostrom, Schelling, V. Smith or Tullock -- nor are they the sort of thinkers I want to currently engage such as Acemoglu, Besley, Easterly, Rajan, Shleifer, Zingales, etc. But they are intriguing thinkers whose ideas are serious consideration and fun to wrestle with. They are subtle thinkers --- social scientific theory is subtle and simple in their hands, not blunt and complicated. And that is how it should be in my opinion.
One of the major confusions in economic thinking since the development of the formal theory of general competitive equilibrium and the representation of economic principles in mathematical models has been the conflation of the mechanisms that produce an equilibrium tendency with the solution to a simultaneous equation system. Another way to put this is to confuse the cause of the equilibration process with the consequences of an equilibrated system. Or yet another way to view it, that you conflate optimality conditions with assumptions rather than view them as by-products of a process which produces them.
Lets go to Webster's Dictionary first. Mechanism is defined as -- a process or system that is used to produce a particular result. A solution is defined as -- a : an action or process of solving a problem; b : an answer to a problem : explanation; specifically : a set of values of the variables that satisfies an equation.
Note the subtle but important difference.
Now read Hayek's "Economics and Knowledge" (1937) paper with this distinction in mind. The key idea in that paper is not, as many have believed, his departure from Misesian apriorism --- even Mises didn't see that --- but his insistence that the pure logic of choice while necessary component of a theory of the market process is not sufficient to explain the logic of the market process tending toward a law of one price and market-clearing. Subsidiary institutional conditions had to be specified, and in particular the question turned on the learning properties of the different institutional environments. An institutional environment of private property, freedom of contract (to freely negotiate prices), and sound money and fiscal responsibility (so profit and loss accounting can be more or less accurate) provides such an environment to incentivize, inform, lure and discipline actors so that their stumbling and bumbling efforts to improve their situation will grope toward optimal outcomes. P=MC emerges from the process, it is not an assumption prior to the analysis of the process. Similarly, we discover what least cost methods of production are each day in our business activities, it is not again an assumption prior to that analysis.
When we conflate solutions with mechanisms we tend to lose focus on how the mechanisms actually work and the time involved in the working out of solutions. This means that the actors who compel the system to adjust and correct are neglected -- in the competitive market economy this character is the entrepreneur, but there are change agents in all walks of life that must be accounted for in social sciences and solution concepts either trivialize their role or ignore it.
Consider a basic Walrasian presentation of the market system and the first and second welfare theorems. All of the working out of the issues is done prior to the public posting of prices -- a pre-reconciliation of plans between buyers and sellers is required -- and then a price/quantity vector is revealed that clears the market. But the market process itself is a "mechanism" for the evolution toward that solution -- it is a reconciliation process through as Adam Smith put it: "It is adjusted, however, not by any accurate measure, but by the higgling and bargaining of the market, according to that sort of rough equality which, though not exact, is sufficient for carrying on the business of common life."
As student of the doings of men in all their walks of life, we wan to study human activity; not that state of affairs where activity ceases to be vital to the explanation. At best the solution defines the end point that will be reached if the mechanism is capable of doing its job. Assuming the mechanism does its job, or cannot do its job, is an abdication of the most important job a social scientist must do -- which is to think through the workings of the mechanisms under investigation.
So my warning to fellow social scientists -- yes the optimality conditions and the equilibrium state of affairs is a thing of aesthetic beauty and we should never dismiss their fundamental importance, but social science is lost when our curiosity about the workings of mechanisms that bring about those conditions is squashed in favor of assuming those results (or assuming away those results). Our focus must be on human action, and the activity those actors initiate, engage in, and work their problems out through. Ultimately, our focus must be on exchange and the institutions within which those exchange relationships are formed. It is there where we will find out whether individuals and groups of individuals will be able to realize the gains from social cooperation, or they will be stalled in that effort and instead devolve into social conflict.
So why are Elster, Hirschman, Sen intriguing despite distance from them intellectually on so many issues? Because they study mechanisms and their curiosity about mechanisms is infectious. Why do I find so many modern economists less intriguing -- even those with whom I share so many priors about how the world works -- because they conflate solutions with mechanism and thus end up ignoring the mechanisms that can produce the solutions.
That is my Smithian/Hayekian plea for mechanisms in social scientific thinking.