Wednesday, September 10, 2014

Economics beyond shocks??

Place a lit cigarette in an ashtray in a closed room where the air is perfectly still. As everyone knows, the smoke will rise, but not in a simple regular flow; the rising flow is unstable, begins to wobble, and then breaks out into a tangled mess -- turbulence. You don't need any outside cause or shock to the rising smoke to make it happen.

Economies do highly irregular things too, as a rule, going through repeated booms and busts, and yet economists seem quite hesitant to see such fluctuations as the result of similar natural instability. In recent decades, at least, they seem to have greatly preferred the idea that fluctuations around average growth must be caused by "shocks" to the economy of some kind. Noah Smith recently gave a nice summary of  the Real Business Cycle theory of Kydland and Prescott, which Prescott and colleagues are still pushing today:

Here’s how recessions work. Sometimes, scientists and engineers invent less new stuff than normal. Fewer new inventions this year mean fewer new inventions next year, too. Anticipating this, companies invest less, and they also cut workers' wages. When wages go down, workers decide to take a vacation instead of work. Voila -- a recession!

Actually, I’m kidding. I don’t think this is how recessions work at all. But the theory I just described is a real macroeconomic theory. It came out in 1982, and its name is the Real Business Cycle model. In 2004, its creators, Edward Prescott and Finn Kydland, won a Nobel Prize for their work. The theory inspired a generation of researchers, and became the dominant theory in certain places, such as the University of Minnesota.

You might be forgiven for thinking that Real Business Cycle theory, or RBC for short, doesn’t deserve its moniker. Just as the Holy Roman Empire was none of the above, RBC theory doesn’t seem to have much to do with business or cycles, and for that matter doesn’t sound particularly real. Most people think that recessions are caused by asset-price crashes, by disturbances in the financial sector, or by Federal Reserve tightening of the money supply.

RBC says we’re all wrong about that. The financial sector, RBC adherents claim, isn't important. Asset prices crash because people see a recession coming ahead of time and act accordingly. And the Fed, according to Prescott in a recent interview, has no more effect on the economy than a rain dance has on rain. In fact, RBC is really sort of a giant null hypothesis -- a claim that the phenomenon known as the business cycle is just an illusion, and that recessions are the normal, smooth functioning of an efficient economy.

In Bloomberg, I've written about some new work that puts this RBC theory into a very new light. It suggests, in fact, that theories of this basic class, if examined more closely, actually predict the existence of inherent instabilities in economies. Specifically, if you relax even slightly some of the heroic assumptions usually employed in such theories -- regarding agents' perfect rational foresight, or the ability of firms to adjust their output instantaneously to changing economic conditions -- then these models become unstable, so that large recessions will happen even without any external shocks, simply because of coordination failures within the economy.

More people should know about this work, which is the result of a serious collaboration between some physicists AND economists. Full text of the Bloomberg piece below:

Economists still don't know what makes it happen. An economy thrives for years, and then suddenly, without warning, falls into a hole. Unemployment soars until somehow, sooner or later, growth resumes. Every economy on the planet has experienced these painful, mysterious and apparently unavoidable slumps.

Among academics, the most popular theory is the Big Shock, which has many variations. In this view, you get a recession when some big thing like an oil crisis whacks the economy, causing a corresponding reaction. Conservative economists assume individuals and businesses will react in the best and most rational possible way, creating an optimal economic response, so the government shouldn't get involved. Others take the less extreme view that governments and central banks, acting wisely, can intervene to help an economy recover.

A few economists instead prefer what you might call the amplification theory. They suggest that interactions between different parts of an economy might make it possible for even tiny shocks to have big consequences, much as a spark in a parched forest can trigger a vast fire. A small downturn for an auto manufacturer might hurt its suppliers, undermining their ability to supply other auto makers and creating a growing cascade of distress. The cause is less the shock and more the links that amplify it.

For most economists, that's the end of the discussion: Recessions are either the result of big shocks, or of small shocks with amplification. They ignore a third possibility: that an economy might sometimes get seriously out of shape with no shock at all. The omission is odd, because this way of thinking was quite common in economics some 50 years ago.

Fortunately, a group of economists and physicists is reviving the old “no shocks” idea. Interestingly, they start with a mathematical model of the economy built by Big Shock theorists -- specifically, the so-called Real Business Cycle model, which still garners lots of attention from economists. Like many mainstream economic models, it assumes that individuals and businesses make perfectly rational, optimal decisions, which lead the economy to a stable economic equilibrium. The new research then makes some adjustments to this picture: It assumes that individuals, rather than having perfect foresight when predicting future prices, sometimes make small errors. The result is radically different. The interactions of firms and individuals now create an ongoing turbulence with sporadic recessions arising from a natural lack of coordination, without any shocks at all.

The researchers go on to show that if you make the model more realistic along any of a number of dimensions -- firms taking a little time to adjust their production to new levels, for example, rather than doing so instantaneously -- you always end up with an inherently unstable economy. The conclusion is pretty much the opposite of what the Real Business Cycle theory's creators originally intended. They wanted to defend the notion that markets work perfectly, not to entertain the possibility that recessions might reflect an inability of markets to coordinate supply and demand. Their own model actually destroys that hope.

It's an amusing and ironic outcome, with implications beyond recessions. For years, many economists have argued that their assumptions of perfect rationality, self-interest and equilibrium are merely convenient elements in valuable thought experiments; they learn about the real world despite the manifestly false assumptions. That position now looks completely indefensible. It looks more as if Big Shock theorists are worried that relaxing their assumptions will lead to some very different and very inconvenient conclusions.

This isn't to say that we know for sure what really causes big recessions. Big shocks, little shocks and inherent instability may all play a role. It will take some honest science to figure that out.  

No comments:

Post a Comment