“Just so” stories of modern economics
Economists like to tell stories with their theories. They look like the wrong stories…
Since the financial crisis of 2007-2008, there's been lots of criticism of how economists build theories of whole economies, i.e. their theories of so-called “macroeconomics”. These theories were useless in explaining the onset of the crisis, or in giving any forewarning of how it would quickly spread to the wider economy (many economists predicted that it wouldn’t). The models in question go by the name of Dynamic Stochastic General Equilibrium (DSGE) models. Many economists insist they are hugely useful for understanding how the economy works, even though it seems that no one on Wall St. — for whom such understanding would mean profit opportunities—uses these models. Practically, they don’t seem to be very useful (which probably won’t come as a surprise to anyone who looks inside these models). So — how are they useful, then? “Useful” in what sense?
Well, I’ve recently come across an elaboration of this argument, used by DSGE modellers to defending their models and their intention to keep doing economics just as they have in the past. These models, they say, are useful in “telling stories” about how an economy works and in thinking about how an economy might respond to various possible policy steps. I first came across this idea from macroeconomist Steve Williamson several years ago, and then heard it again last week at a conference on diversity in macroeconomics in England. An economist there asserted that no one takes these models too seriously — after all, they find little support on empirical grounds — but that they are extremely useful for telling stories to understand economic events.
I’ve been trying to understand how this can possibly make sense. And I’m having trouble. I don’t think it makes sense. I think it’s an excuse, and nothing else, used to defend an indefensible way of building theories. Most people find economics utterly boring; certainly reading economic theory is highly tedious. But taxpayer money is being spent to fund this stuff, so people should know what they are supporting. The more they knew, I suspect, the more economists themselves, and the people who direct research funds in economics, would make some long overdue changes in priorities.
Another version of the “story telling” argument says that “all models are wrong, but some are useful.” Here’s Steve Williamson making this point a few months ago:
The models are all wrong… That's what makes economics different. These things are all invalid, but we work with wrong models because they are simple, and - of course - because they are useful.
That sounds superficially convincing, but only superficially. The logic really doesn’t make sense. Here’s why.
Let’s suppose that DSGE models, as many economists say, are useful for telling stories about how economies work. There are, then, two possibilities depending on what “useful” is supposed to mean. It might be that those stories are useful because they give real insight into the mechanisms by which economies work and change; people might be able to use them to understand tricky economic cause-effect relationships in a way that goes beyond any other theory. Or, alternatively, they might be “useful” for some other reason not having anything to do with real insight. Which is it?
In the former case, if it’s real insight, then it should be possible to show that DSGE models are in fact superior to other models on scientific grounds, i.e. because they make better predictions in some setting. This doesn’t seem to be the case, as economists readily admit that DSGE models don’t predict very well at all. The fact that they aren’t used by anyone on Wall St. suggests much the same thing. Hence, I believe, this is not why they are “useful” for telling stories. It has nothing to do with real insight. The usefulness comes in something else.
So why are they useful to economists? I think DSGE models are useful in letting economists tell certain kinds of stories about how economies work. They’re not particularly plausible stories, but they are the kind that economists have been trained to tell and that they have vast experience in telling. These stories always see economic outcomes as reflecting the rational decisions and optimizing behavior of individuals and firms (typically, in an economy with one consumer and one firm). You can browse through these lecture notes of an intro macroeconomics class given by Steve Willamson to see the kind of thing that’s always involved. You will look in vain for anything even remotely suggestive of the complexity of a real economy. Economists like these models not because they have any reason to think they give a true, approximate picture of an economy, but because they’ve been telling similar stories for decades. They don’t know how to tell any other kind of story. (Sadly, this resistance to other kinds of stories often gets in the way of useful ideas being given a fair hearing in policy circles; this idea about an algorithmic approach to reducing systemic risk is a good example.)
You can wonder how it is that economists got themselves into such a strange position, whereby they prefer to go on telling stories they have no reason to believe. It’s as if they climbed into the heights of a great foggy mountainous wilderness, only to find themselves suddenly stuck on a low, ugly peak, separated by deep and threatening chasms from far more promising peaks further away. The thought of descending all the way back down to climb back up again is too much; better to stay where they are and pretend the peak is higher than it is. Economist Alan Kirman used this metaphor in a recent history exploring how and why macroeconomics hasn’t made more progress and why it suffers from the “pretense of knowledge syndrome”:
The theory that we have developed is self‐contained and logically consistent but has little to do with the actual functioning of the economy and its markets. The search for empirically refutable propositions has not been a high priority. We have reached an isolated peak far from the other challenging hills that we might profitably try to climb. Sometimes, in moments of weakness, I conjure up the vision of one of those sects that gather on a mountaintop, absolutely sure of their destiny, waiting for the end of the world.
Telling stories is useful. But not “just so” stories. We deserve something better.
Follow The Physics of Finance on Twitter: @Mark_Buchanan
Read related articles:
High-frequency trading: algorithms now rule the world
Why the controversy over HFT misses the most important point
How Google's algorithm could fix the financial system
Economics: how often to we "lose" decades?
What's the use of "econo-physics"?
