Post hoc ergo propter hoc
By Scott Sumner
Noah Smith recently discussed a study that suggests a new approach to business cycles:
Gennaioli and Shleifer explain these patterns by turning to their own preferred theory of human irrationality — the theory of extrapolative expectations. Basically, this theory holds that when asset prices rise — home values, stocks and so on — without a break, investors start to believe that this trend represents a new normal. They pile into the asset, pumping up the price even more, and seeming to confirm the idea that the trend will never end. But when the extrapolators’ money runs out, reality sets in and a crash ensues. Gennaioli, Shleifer, and their coauthors have been only one of several teams of researchers to investigate this idea and its implications in recent years.
When extrapolative expectations are combined with an inherently fragile financial system, a predictable cycle of booms and busts is the result. At some point during good economic times, irrational exuberance takes hold, pushing stock prices, house values, or both into the stratosphere. When they inevitably come down, banks collapse, taking the rest of the economy with them.
Except for the “predictable” part, this is pretty much the standard view of the Great Recession. Obviously I think it’s completely wrong, but today I’ll focus more on how it became the standard view.
In the past, I often point to the 27-month period from January 2006 to April 2008, when the housing market crashed and the banking system came under severe stress. The unemployment rate was pretty stable during this period, rising from 4.7% to 5.0%. So obviously a housing crash and banking stress don’t directly cause a Great Recession, at least not immediately.
One counterargument is that these problems made a Great Recession inevitable, but with a lag. That may be true, but it’s important to emphasize that the economics profession as a whole did not accept this theory in mid-2008. Not even close. For instance, here’s the consensus forecast of professional economists in the 3rd quarter of 2008, when the Great Recession was already well underway:
The profession did expect a sluggish performance over the next 12 months, but certainly no big recession. Importantly, when this forecast was made the profession was fully aware of the severe housing slump, and also the stress on the banking system. Banks had already sharply tightened lending standards due to the subprime fiasco. It’s true that Lehman had not yet failed, but that’s just one bank. And soon after the Lehman failure the other big banks were bailed out.
In September 2008, the Fed met and decided not to ease policy after Lehman failed, because they were worried that the economy might overheat if rates were cut. So even after knowing about everything that supposedly caused the Great Recession, the Fed was not expecting this economic outcome after Lehman failed. More importantly, there was little criticism of Fed policy at this time, which represented something close to the consensus view of the profession. (In contrast, I was quite critical of the Fed.)
Before explaining how the profession changed its mind, I’d like to emphasize one point very strongly. Although my current view of the Great Recession is now regarded as wildly implausible, it was close to the standard view as of mid-2008. That is, severe housing slumps combined with highly stressed banking industries do not cause big recessions. You may not agree that I am correct, but it’s important that you see that this was the standard view a decade ago.
So then what happened?
1. My view is that tight money caused NGDP growth to plunge, and this led to both the Great Recession and also worsened the housing/banking crisis. This view might be right or it might be wrong, but it’s certainly consistent with mainstream macro circa 2008. Today it’s a highly heterodox view.
2. The other view is that the housing/banking crises caused the Great Recession. This view was once heterodox, but is now standard.
Why did economists abandon the mainstream view circa 2008? You could argue that there was new information; they saw a housing/banking crisis followed by a Great Recession. But why assume any causal link? Why not just assume that is was tight money that caused NGDP to plunge? After all, history is littered with dozens of examples of financial crises associated with severe recessions, including the early 1930s. It’s not like we learned anything new from 2008-09. We didn’t learn anything that we didn’t already know before the Great Recession. Why change the standard model?
This reminds me of science in ancient times. There’d be an eclipse and then a week later a severe earthquake. The priests would claim the eclipse somehow made the earthquake inevitable. The Great Recession is now seen as being caused by housing/banking problems, because it followed housing/banking problems.
Actually, a severe recession was still not inevitable in mid-2008—it was excessively tight money that caused NGDP to plunge and unemployment to soar.
HT: Tyler Cowen