Tuesday, 9 June 2009

No blame in an equilibrium

Who's to blame for the financial crisis and the recession?

This idea of 'who's to blame' is an omnipresent theme of the last six months. A quick search for "to blame for the crisis" or "to blame for the recession" makes hilarious reading. Here are the first few candidates that show up:
What is the appropriate response to this orgy of vilification? Is it to join in and give your own opinion of who's really to blame? Or is it to look at how the system works and try to draw some more reasoned conclusions?

Let me point to a few specific examples:

Rob Killick of UK After The Recession wrote an article last week explaining that "individual behaviour is not to blame for the recession".

It's a tricky statement to evaluate, because it's either trivially false or obviously true - depending on how you interpret it.

Literally speaking, if the behaviour of individuals is not to blame, whose behaviour was it? The economy is not a machine run from outer space, it's the agglomeration of trillions of actions by billions of people. And everything that happens in it is, ultimately, traceable back to individual decisions by human beings.

But I guess what Rob means is that the structure of the economic system is one cause of the situation, and that political leaders shouldn't duck the need for changes by blaming a faceless mass of citizens. It's hard to argue with that.

It's a false dichotomy to claim that the only two options are "blame the individuals and carry on as before" or "reform the system". If the system needs reforming, it can only be done with an accurate understanding of human behaviour - avoiding the neoclassical fantasy of rational utility maximisation.

Another: Scott Sumner is convinced that the Federal Reserve is to blame for this recession (and, essentially, every other one). It's their fault for making money too tight (or allowing it to be).

And there are plenty more, as the above examples show.

But something feels a bit wrong about these attempts to pin blame on somebody. And not just the general moral queasiness of scapegoating.

One of the great things about markets is that they are - in theory - self-correcting. A market in equilibrium is a complex system where every part interacts with many others. Thus, every action involves more than one party. Subprime borrowers and lenders both enter into a transaction. Politicians and voters influence each other. Broadcasters and viewers reinforce each other's prejudices. Companies and employees set wages and conditions in an ongoing negotiation. The US government and Chinese central bank enter into a bond contract which suits both of them at the time it is struck.

So any action you might have taken which contributed to the crisis - taking out a subprime loan or lending money to the US government - was enabled by the party on the other side of the transaction. And implicitly influenced by every other person in the economy who has bought or sold anything else and thus affected the price you paid. The blame, if there is such, is diffused from you to everyone else in the economy.

And any overvaluing or undervaluing of a product, every innovation that is crowded out by overinvestment in housing stock, every person laid off by their employer will - in theory - be corrected immediately by arbitrage or repricing.

In this theory, a recession never happens. Any potential crisis is fixed by the system before it takes place, and nobody even notices that it was imminent.

Of course, this is not what happens in reality. Clearly the market does not immediately reach this kind of equilibrium - sticky prices and wages, or contracts, or adjustment costs, or high gross debt which cannot easily be unwound, slow the process and can cause major pain and a permanent loss of welfare. But what's "to blame" for the recession is not the action of any one individual, or any group. It's the overall behaviour of the system - which is nothing more than the aggregate behaviour of every person in the system.

Rob Killick characterises behavioural economics as "blaming individuals" and absolving everyone else (who else would that be, exactly?) But this is not at all the point. What behavioural economics does - like most other branches of economics - is to provide explanations of behaviour, and explanations of the consequences of behaviour, which can then be used to determine the appropriate solutions.

Maybe part of the solution is to do with US-China capital flows. Maybe it is to do with greater innovation in the private sector. Maybe it's to do with bankrupting or recapitalising the banks. But whatever it is, it was the choices of individuals that led to the problem and no solution can work without an understanding of how people make those choices.

And government probably does have a role in smoothing adjustments, combating market failure, putting a proper price on externalities and providing large-scale risk pooling and other coordination goods. Does this mean it's the government's fault if it doesn't get it all right first time?

We may as well stop blaming people: if economics is a science, its goal is to understand things as they are and provide tools to make them better. Don't try to shut down the process of understanding because the resulting tools aren't the ones you would wish them to be.


PunditusMaximus said...

It's not just sticky prices, etc. It can also be that one or the other party is outright wrong about the value of the asset. That appears to have been the case in the mortgage and real estate speculation markets; you can't tell me the folks who bought houses two or three years ago expected these kinds of drops in their asset values. Or that the banks which underwrote those loans did.

In addition, the presence of occasionally and unpredictably correlated risks means that it may be impossible to insure against certain risks, and it may be impossible to predict which ones.

Markets are only completely self-correcting if parties are omniscient.

Leigh Caldwell said...

Yes, absolutely - I agree almost completely with your comment. A key insight within your example is that neither the banks nor the homebuyer thought they were exploiting the other nor had any reason to believe they were contributing to systemic risk - neither is therefore "to blame" in the conventional sense.

There are a number of "standard" departures from rational efficient market theory: imperfect information, cognitive biases, principal-agency problems, inability to form complete and enforceable contracts, insufficient market size and so on. In essence they are all systemic problems - flaws inherent in the makeup of the market system which prevent it from reaching a self-correcting equilibrium.

One could argue that some of these flaws do allow room for individual blame - for instance if there is a principal-agent problem, you might blame the agent for acting in their own interest instead of that of the principal. But even if you can get some agents to act more 'honestly', that does not solve the real problem, which is that there's a flaw in the nature of the system that allows principal-agent conflicts to arise.

Where I slightly disagree is that omniscience is not one of the required conditions for markets to self-correct. If certain other conditions are met (a market for every real and contingent good, published prices in all markets, no limits on information processing and so on), no party needs to know everything. However if some of those other conditions are not met, omniscience could hypothetically substitute for them.

PunditusMaximus said...

I don't like that way of looking at it -- "flaw" in the nature of the system. The system is. We have ideals about how a system might be self-correcting, but it may not be reasonable or possible to bring a given system into line with those ideals. Indeed, the default is always transactions not entirely at arms' length with parties not entirely certain of everything.

I'm reminded of that story about the fine for being late at the day care in Predictably Irrational. Some problems are just better solved with messy social norms than by commoditizing them. Indeed, sometimes it feels like we're better off not even knowing they're problems. People are going to be late to pick up their kids. It's not really so much a problem as an inevitability. The problem, if it exists, is if the tardiness gets so extreme and unpredictable that it interferes with the day care's ability to amortize it easily.

You're right that I should have said infinite information processing, rather than omniscience, but both of them have an infinity involved, which is a real issue.