Alex Tabarokk over at Marginal Revolution has a lucid explanation of the business cycle theories that won Kydland and Prescott this year's Nobel Prize in economics.
Recessions have almost always been thought of as a failure of market economies. Different theories point to somewhat different failures, in Keynesian theories it's a failure of aggregate demand, in Austrian theories a mismatch between investment and consumption demand, in monetarist theories a misallocation of resource due to a confusion of real and nominal price signals. In some of these theories government actions may prompt the problem but the recession itself is still conceptualized as an error, a problem and a waste. Kydland and Prescott show that a recession may be a purely optimal and in a sense desirable response to natural shocks.
The idea is not so counter-intuitive as it may seem. Consider Robinson Crusoe on a desert island (I owe this analogy to Tyler). Every day Crusoe ventures out onto the shoals of his island to fish. One day a terrible storm arises and he sits the day out in his hut - Crusoe is unemployed. Another day he wanders out onto the shoals and finds an especially large school of fish so he works especially long hours that day - Crusoe is enjoying a boom economy.Now add into Crusoe's economy some investment goods, nets for example, that take "time to build." A shock on day one will now exert an influence on the following days even if the shock itself goes away - Crusoe begins making the nets when it rains but in order to finish them he continues the next day when it shines. Thus, Crusoe's fish GDP falls for several days in a row - first because of the shock and then because of his choice to build nets, an optimal response to the shock.
An analogy is one thing but K and P showed that a model built from exactly the same microeconomic forces as in the Crusoe economy could duplicate many of the relevant statistics of the US economy over the past 50 years. This was a real shock to economists!
I'll say. And it does -- or should -- change the political discussion around recessions too. But it probably won't ....
And if you're a fan of Schumpeter, as I am, recessions are good things because they clear away the underbrush of the 'creatively destructed'.
A.L.
I can see it now, "Hey that recession started on MY watch!".
I am a Schumpeter fan (big surprise for a VC, no?). However, I've long had another metaphor for the business cycle that may be too obscure for general use, but has the advantage of an already attached body of mathematics. That's the optimization technique of simulated annealing.
Huh? Well, see here, but it's an analogy to how a material cools into a form that reflects applied stress. Take a simulation problem that can be solved by somewhat random perturbations - say optimal placement of ICs on a circuit board - and 'bounce' the placement of the chips around to a degree determined by a 'temperature'. Examine each proposed change of placement for goodness, and accept the winners. Reduce the temperature - and hence magnitude of perturbations - over an 'annealing cycle' and you'll usually end up with a decent solution to the problem even though the process was pseudo-random. YMMV in some cases, see the linked article.
OK, here's the analogy bit. Annealing works when the stress (problem constraints) remain more or less constant (stationary, in statistical parlance). But as anyone who has handworked metal knows (I have), after you work the metal for a while you have to reanneal it or it's brittle. Call the build up of stress a drift of market supply from potential demand, and the recession an annealing cycle. It takes (in the Valley at any rate) the 12-16 hour day gonzo engineers and entrepreneurs out of that mode and turns them loose at networking events and parties to jabber about new ideas and trade e-mail addresses. It dumps cheap assets from servers to office chairs to subleases onto the market. It drives those whose talents are in oversupply into training courses or out of the Valley. The people and assets are bouncing around loose, trying to find a team or idea or demand to hang onto. The Valley is never in bigger ferment than when the economy is rotten. Eventually (at least up to now) some of the assets congeal around new demands, the assets stop being so cheap and available, and we head back toward a tight labor market.
OK, Schumpeter is a lot more understandable, but I'd love to see some analytic economist take a swing at the SA mathematics.
Tim Oren - at one time I was on a mailing list from a guy who was big into SA, and he had lots of examples of its uses in economics, as well as some development libraries you could get from his site (ftp as I recall). Ring a bell at all?
As I recall, my husband has advised at least one master's level thesis on SA as well as having used it some himself. (His expertise is in ops research and decision analysis.)
Interesting comment, Tim. Your description matches my experience in high tech startups ... I think it would be promising to put actual numbers on it. Any of our readers looking for a dissertation topic in that area?
Oscar - I'm afraid that doesn't sound familiar. My encounter with SA was over a decade ago, considering it as an possibility for clustering and visualization of personal information items. We went with something else for the experiment - a sort of variant on Kohonen maps - but the idea was so elegant that it's stuck. One of those rare cases (for me) where the mathematics makes a direct connection to a tactile experience - beating on steel, then putting it into the heat.
And I'll take another moment to tie it back to Rittel's 'tame' and 'wicked' problem formulation; SA appears to be a way to effectively model the otherwise unmodelable...interesting.
A.L.