Imagine a group of producers preparing for the Christmas season. Some years they will produce bracelets instead of Star Wars figurines, and turkeys instead of ducks. Business cycles are then the result of forecasting noise. Every now and then, we simply have bad luck.
In contrast, Keynesian theories see weak aggregate demand as the problem. "Real" business cycle theories cite negative real shocks; one example might be oil price hikes. Black’s theory focuses on mismatched demands.
The dot.com bust would seem to be a partial example of Black’s thesis. Entrepreneurs believed that consumers would pay for Web-ordered, home-delivered groceries, when in fact they wanted to buy more homes.
Black also believed that adjustment processes are slow. Human capital is highly specific to particular endeavors. So it takes an economy a good bit of time to recover from its bad guesses.
OK, but why do so many entrepreneurs err in forecasting consumer demand at the same time? The Law of Large Numbers would seem to imply a more even distribution of errors. You might expect that about ‘n’ percent of entrepreneurs guess wrong this year, next year, and so on. I can think of three possible (and possibly wrong) answers to this criticism:
1. The American economy does not, in reality, have so many independent sectors. Bad luck in a few of the larger sectors spreads to all or most sectors.
2. Many sectors may rely on common and sometimes erroneous forecasting techniques, or may be trying to forecast common variables. (But what are those common variables? Does Black’s theory then collapse into the Austrian or Lucasian argument that people are fooled by monetary policy? What else could be fooling them?)
3. The Law of Large Numbers does imply that an economy, of a given size, is less risky when it has more independent sectors. This does not mean that business cycles are impossible. The frequency of cycles will depend on how many independent sectors are operating, how risky is each sector, and how frequently we are "rolling the dice" with demand forecasts. So we can still get a business cycle more than once every 2 million years.
Black’s theory has one feature which is simultaneously appearling and infuriating. It explains why economists find business cycles so hard to predict. But at the same time this drains Black’s theory of any obvious testable predictions. Black did not mind. He once said to me: "The easiest theory to falsify is a theory which is false."
We might make other criticisms of Black’s account as well. Must we not look to labor markets for a closer account of the true action? What was the taste mismatch in the Great Depression? Why does everyone pay so much attention to the Fed?
The final verdict: There are very few serious contenders in the area of business cycle theory, so I am not ready to dismiss this one, warts and all.