Thursday, January 26, 2012

What future for economics?

Martin Wolf writes on the Financial Times' The World blog (also known as Gideon Rachman's blog) about a forum he moderated in Davos, featuring several promiment salt-water economists. To summarise, he has 10 propositions about economics. Since I took the liberty of writing something on this topic in the introduction chapter of my dissertation, I thought I might comment.

1. orthodox economics had, in the years leading up to the crisis, become more a cult than a science, particularly with the assumption that what exists in competitive markets has to be the best possible outcome, since, if it were not, it could not exist. So, if crises are not predicted, it is because they cannot be: they are the result of unexpected shocks, by assumption.

Statements such as this I can only evaluate based on my undergraduate experience, given that my grad school days only started when the crisis was already almost upon us, and were spent in any event among Public Administration people, not economists. All I can do is recall that when I studied Finance & Banking in Groningen, we did spend quite a bit of time on the (heterodox) theory of Behavioural Finance, including the (relatively orthodox) theory of rational bubbles. (Orthodox because the model involves quite a bit of rationality.)

2. let a thousand flowers of thought bloom. There cannot be just one general model of the economy or just one approach to economics. Among the blooms discussed were behavioural economics, neuroeconomics, computer based modelling of processes over time. Participants recommended talking to political scientists and even sociologists. They also recommended looking at the causes of inequality, the economics of happiness, the role of institutions, the importance of culture, and the effects of power. Fortunately, economists are creative people. A great deal of imaginative stuff is going on.

This is where I put in my usual warning about the difference between interdisciplinary research and multidisciplinary work. Each of the social sciences should play to their strengths, and the idea that individuals optimise something within certain constraints is the key strength of economics. Individuals might have limited knowledge, limited computing ability, or they might care a great deal about the welfare of others, but there is a meaningful concept of utility that individuals optimise in some meaningful way. If we give that up, for example by taking on too many psychology-inspired notions of optimising, we end up like a car-boat, which, despite these pretty pictures, is not a very good car and not a very good boat either. Instead, each discipline should do what it is good at, while at the same time trying to figure out how to aggregate our results.

3. the sociology of the profession – the need to define and defend a core discipline that can be taught to students and so determines what it means to be an economist – militates against such heterodoxy. There is a fundamental tension here. But cross-disciplinary co-operation is one way of out.

This is true in every academic discipline. Without a core that can be taught to undergraduates, an area of research doesn't even deserve to be called a discipline. But then, I don't see the problem with economists teaching undergraduates the basics of utility- and profit maximisation under uncertainty in micro-economics, about IS-LM etc. in macro-economics and about Heckscher-Ohlin in international trade theory. As far as I know, that is still the basis of what we do.

4. human beings are not rational calculating machines. Their mood and approaches to decision making varies with the circumstances.

Straw man. No one in economics would dispute this. The point is simply that in many cases it can be useful to talk about them as if they are rational calculating machines.

5a. time matters in economic processes, which are, in general, not reversible and not characterised by any sort of equilibrium.

This one I can only agree with. My neo-institutional dissertation, above-mentioned, devoted a whole section to how the outcome we observe is supposed to move towards the optimum predicted by the theory, arguing that this is a serious weak point of both Williamson and Hart c.s. Later on, one of the "benchmark theories" I examined was (forum participant) Brian Arthur's path dependency, which has its own strengths and weaknesses, but which was valuable enough to earn a spot in my story.

5b. economics suffers from physics envy. It seeks to be an exact science, which is impossible.

Rubbish. People who make this criticism tend to be mistaken in two ways: First of all, physics is a lot more difficult than they think it is. There's more to it than Newtonian mechanics. Secondly, this "physics envy" is what drives us forward. While we should be cautious that we do not overdo it, and sceptical about the dominance of calculus in economics, physics envy as such is what has made economics superior to sociology, its 19th century sibling.

6. the world is not computable. It is far more sensible to think in terms of irreducible uncertainty than computable risk. This fundamental point made by John Maynard Keynes was lost in the subsequent so-called “neoclassical synthesis”.

I would have added Frank Knight's uncertainty/risk distinction, but OK. Absolutely agreed. This is why I used some of my afore-mentioned introductory chapter to advocate for modesty on the part of economists. Our work only describes one aspect of society, and even then we tend to overlook the uncertainty that surrounds us.

7. being a study of complex human behaviour, in which the world is created by human understand and motivations, economics is hard.

Indeed. And physics is hard, too.

8. in theory it is right and proper to abstract in order to focus on a specific phenomenon. In addressing policy, this is irresponsible. Policy must be informed by an understanding of everything that might bear on the problem in front of one. This makes economic policy really hard to do well.

Like I said, economists should focus on what they're good at. When they're done, they should combine their results with everybody else's. Only then can you make policy recommendations. Failure to do so makes the theoretical basis of the recommendations dangerously narrow, and has a tendency to ignore the is/ought distinction. Unless they are checked by other disciplines, economists have a dangerous tendency to equate efficiency with desirability.

9. even though economists get much wrong, they still have much to offer to non-economists who tend to assume that economic problems are far more simple than they actually are.

Amen. I would only add that more often than not, the contribution that economists make is to show that problems are simpler than they are perceived to be. Economists are good at cutting away the dead wood, of asking "why?" a few more times than anyone else. That is what is so awesome about the work of Ronald Coase: His 1960 article outlines a number of tort law hypos that lawyers are very familiar with, but instead of asking what the rule of liability is, he asks why the rule of liability is what it is. "Why?"

10. there is a great danger that in rejected the most simplistic pro-market mantras, economists and policymakers will embrace even more dangerous and naïve statism.

No comments: