Thursday, November 20, 2008

DIY economics

There I am, performing a bit of rotary sanding on top of a piece of the newspaper that I’d considered disposable (the Business pages of the Guardian), when something catches my eye. Namely, a reference to ‘the science weekly Nature’. What’s all this?

It is an article by the Guardian’s Management Editor Simon Caulkin, explaining why ‘self-interest is bad for the economy’. Needless to say, that’s not quite right, and presumably not quite intended. The economy relies on self-interest. What Caulkin is really saying is that self-interest without restraint or regulation is bad for the economy, especially when it generates the kind of absurd salaries that promote reckless short-termism and erosion of trust (not to mention outright arrogant malpractice). Caulkin rightly points out that Adam Smith never condoned any such unfettered selfishness.

But where does Nature feature in this? Caulkin refers to the recent article by Jean-Philippe Bouchaud that points to some of the shortcomings of conventional economic thinking, based as it is on unproven (indeed, fallacious) axioms. In physics, models that don’t fit with reality are thrown out. “Not so in economics”, says Caulkin, “whose central tenets – rational agents, the invisible hand, efficient markets – derive from economic work done in the 1950s and 1960s”. Bouchaud says that these, in hindsight, look “more like propaganda against communism than plausible science” (did anyone hear Hayek’s name whispered just then?).

Now, the last time I said any such thing (with, I hope, a little more circumspection), I was told by several economists (here and here) that this was a caricature of what economists think, and that I was just making it up. Economists know that markets are often not efficient! They know that agents aren’t always rational (in the economic sense)! Get up to date, man! Look at the recent Nobel prizes!

In fact I had wanted, in my FT article above, to mention the curious paradox that several recent Nobels have been for work decidedly outside the neoclassical paradigm, while much of economics labours doggedly within it. But there was no room. In any event, there is some justification in such responses, if the implication (not my intention) is that all economists still think as they did in the 1950s. These days I am happy to be more irenic, not only because that’s the sort of fellow I am but because it seems to me that thoughtful, progressive economists and those who challenge the neoclassical ‘rational agent’ tradition from outside should be natural allies, not foes, in the fight against the use of debased economic ideas in policy making.

But look, economists: do you think all is really so fine when a journalist paid to comment on the economy (and not just some trumped-up physicist-cum-science writer) not only possesses these views about your discipline but regards it as something of an eye-opener when someone points out in a science journal that the economy is not like this at all? Are you still so complacently sure that you are communicating your penetrating insights about economic markets to the world beyond? Are you so sure that your views are common knowledge not just in academia but to the people who actually run the economy? Maybe you are. But Nobel laureates like Joe Stiglitz and Paul Krugman aren’t.

Thursday, November 06, 2008

What you don’t learn at school about the economy

[Why, you might wonder, would I want to expose myself to more flak from economists by writing this column for Nature’s online news? Well, if any economists get to see it at all, I hope they will recognize that it is not an attack at all but a call to make common cause in driving out simplistic myths from the world of economic policy. I was particularly taken with the article by Fred Argy on the use of economic theory in policy advising, and I guess I am moved to what hope is the irenic position of recognizing that his statement that ‘economics does not lend itself to doctrinaire policy assertions’ should apply too to criticisms of traditional economic theory. The world is too complex to become dogmatic about this stuff. My goodness, though, that seems unlikely to deter the pundits, from what I have seen.

Below, as usual, is the ‘long’ version of the column…]

When the sophisticated theories of economics get vulgarized into policy-making tools, that spells trouble for us all.

The column inches devoted to the global financial crisis must be running now into miles, and yet one could be forgiven for concluding that we are little the wiser. When one group of several hundred academic economists opposes the US Treasury’s bank bail-out and another equally eminent group supports it, how is the ordinary person supposed to decide about what is the right response to the financial mess we’re in, or indeed what caused it in the first place?

Yet some things seem clear. A largely unregulated market has clearly failed to deliver the optimal behaviour that conventional theories of economic competition promise it should. Most commentators now acknowledge that new types of financial regulation are needed, and the (for want of a better word) liberal pundits are seizing the chance to denounce the alleged superiority of the free market and to question whether Adam Smith’s invisible hand is anything but a libertarian fantasy.

Behind all this, however, is the question of why the free market hasn’t done what it is supposed to. In the New Statesman, economics Nobel laureate Joseph Stiglitz recently offered an answer: “For over a quarter of a century, we have known that Smith’s conclusions do not hold when there is imperfect information – and all markets, especially financial markets, are characterised by information imperfections.” For this reason, Stiglitz concludes, “The reason the invisible hand often seems invisible is that it is not there.”

Now, some might say that Stiglitz would say this, because analysing the effects of imperfect information is what won him his Nobel in 2001. In short, the traditional ‘neoclassical’ microeconomic models developed in the first half of the twentieth century assumed that all agents have ‘perfect’ information: they all know everything about what is being bought and sold.

This assumption makes the models mathematically tractable, not least because it ensures that all agents are identical. The idea is then that these agents use this information to deduce the actions that will maximize their ‘utility’, often synonymous with wealth or profit.

Under such conditions of perfect competition, the self-interested actions of market agents create an optimally efficient market in which asset prices attain their ‘correct’ value and supply matches demand. This is the principle notoriously summarized by Gordon Gekko in the movie Wall Street: greed is good, because it benefits society. That idea harks back to the French philosopher Bernard Mandeville, who argued semi-humorously in 1705 that ‘private vices’ have ‘public benefits.’

Stiglitz and his co-laureates George Akerlof and Michael Spence showed what can go wrong in this tidy picture when (as in the real world) market agents don’t know everything, and some know more than others. With these ‘asymmetries’ of information, the market may then no longer be ‘efficient’ at all, so that for example poor products can crowd out good ones.

Now, as Stiglitz says, this has been known for decades. So why isn’t it heeded? “Many of the problems our economy faces”, says Stiglitz, “are the result of the use of misguided models. Unfortunately, too many [economic policy-makers] took the overly simplistic models of courses in the principles of economics (which typically assume perfect information) and assumed they could use them as a basis for economic policy.”

Economist Steve Cohn, now at Knox College in Galesburg, Illinois, echoes this view about the failures of basic economic education: “More than one million students take principles of economics classes annually in the United States. These courses will be the main contact with formal economic theory for most undergraduates and will influence how they think about economic issues. Only a few percent of all students studying introductory microeconomics will likely use a textbook that seriously challenges the neoclassical paradigm.”

And there lies the problem. When people criticize economics for its reliance on traditional models that, while occasionally applicable in some special cases, are just plain wrong in general, the usual response is that this is a mere caricature of the discipline and that of course economists know all about the shortcomings of those models. Look, they say, at the way people like Stiglitz have been rewarded for pointing them out.

Fine. So why are the traditional models still taught as a meaningful first approximation to economics students who may never encounter the caveats before graduating and becoming financiers and policy advisers? Why has free-market fundamentalism become an unqualified act of faith among many pundits and advisers, particularly in the US, as this year’s Nobel laureate Paul Krugman has explained in his 1994 book Peddling Prosperity? Why are models still used that cannot explain crashes and recessions at all?

Robert Hunter Wade of the London School of Economics agrees that the sophistication of academic economics tends to vanish in the real world, despite what its defenders claim. “Go to the journals, they say, and you find a world of great variety and innovation, where some of the best work is done on issues of market failure. And they are right as far as they go. But one should also sample economics as it is applied by people such as World Bank country economists when they advise the government of country X, and as it is hard-wired into World Bank formulas for evaluating countries' policies and institutions. In this second kind of economics the sophistication of the first kind is stripped away to leave only ‘the fundamentals’.”

As Australian policy adviser Fred Argy has pointed out, such economic fundamentalism based on simplistic models commonly lead to dogmatic policy extremism. “We saw in the 1960s and 1970s how the work of John Maynard Keynes was vulgarised by many of his followers, and used to justify the most extreme forms of government intervention. And in the 1980s and 1990s we saw how monetarism, public choice theory and neo-classical economics have been misused by some to justify simplistic small government policies.”

An example of this misdirected thinking can be seen in the way several columnists have announced smugly that it is wrong to describe the current financial panic as ‘irrational’: it is perfectly ‘rational’, they say, for people to be offloading stock before it becomes valueless. That’s true, but it fails to acknowledge that this is not the kind of rationality described in conventional economic models. Rational herding behaviour is called irrational because it is not what the models predict. In other words, there’s something badly wrong with those models.

Scientists are used to the need for approximations and simplifications in teaching. But this doesn’t mean that they regard Lamarkism as a useful approximation to the Darwinism that graduate students will learn, or terracentrism as the best system for Astronomy 101.

Sadly, this often becomes an argument about how backward and unscientific economics is. That is not only unhelpful, but untrue: it is quite correct to say that a glance at the Nobels reveals (with a few exceptions) the dramatic leaps in understanding and realism that the discipline has made since its origins in misguided analogies from microscopic physics. Knowledgeable economists and critics of traditional economics are on the same side; they need to unite against the use of vulgarized, introductory or plain incorrect models as instruments of policy. After all, as the British economist Joan Robinson, a pioneer in the understanding of imperfect competition, put it, “the purpose of studying economics is to learn how to avoid being deceived by other economists.”