Larry Summers weighed in on the famous Reinhart-Rogoff Excel spreadsheet error in a Washington Post column this morning. His first big lesson from the debate is:
"Anyone close to the process of economic research will recognize that data errors like the ones they made are distressingly common."
Summers immediately demonstrates the truth of this assertion as he tries to make a second point about inferring the future based on statistical regularities from the past.
"Trillions of dollars have been lost and millions of people have become unemployed because the lesson learned from 60 years of experience between 1945 and 2005 was that 'American house prices in aggregate always go up.' This was no data problem or misanalysis. It was a data regularity until it wasn’t. The extrapolation from past experience to future outlook is always deeply problematic and needs to be done with great care."
The problem with Summers story is that American house prices in aggregate did not always go up. In fact, for the century from 1896 to 1996 they just kept pace with the overall rate of inflation. Here's the story using government data from 1953. (Robert Shiller constructed a series going back to 1896 from a variety of data sources.)
Source: Bureau of Labor Statistics, Federal Housing Finance Authority, and Author's calculations.
It was easy to see that house prices, adjusted for inflation, did not have a clear upward trend until the bubble started to propel prices upward in the late 1990s (yes, coinciding with the stock bubble, just like Japan). For some reason Larry Summers chose not to notice the data right in front of his eyes both at the time and apparently even to this day. (What is it about Harvard economists and data?) So yes, the tendency for economists to misread data is distressingly common and will probably continue to be as long as economists who make such errors face no consequence in terms of their career or public standing. (At least, this is what economic theory would predict.)
While it is tempting to end on this note, it is important to take issue with Summers' lesson #3:
"Third, while Reinhart and Rogoff’s work was shown not to support the claims made by prominent right-leaning U.S. and British figures regarding the urgency of deficit-reduction efforts, much of the joy taken on the left in their embarrassment is inappropriate. It is absurd to blame them for austerity policies. The authors of those policies chose the policies first and then cast about for intellectual ballast. While there may be no threshold beyond which debt becomes catastrophic, and while the British and American experiences both suggest that fiscal contraction in a slack economy where interest rates are near zero is inimical to growth, it is a grave mistake to suppose that debt can or should be accumulated with abandon."
It's nice of Professor Summers to offer this lecture to the left. (Maybe he'll offer an expanded version in a MOOC.) I suspect that just about everyone on the left knows that the politicians pushing austerity couldn't care less about Reinhart and Rogoff's arguments. However, their work was very valuable in providing intellectual cover. It is much easier for politicians to say that they are cutting programs like Social Security and throwing people out of work because they fear an economic crisis than to say they are doing it just to ensure that corporate profits stay high and rich people don't have to pay higher taxes. For this reason, Reinhart and Rogoff's work was very important for the austerity policies that have been implemented in much of the world.
As far as his point that, "it is a grave mistake to suppose that debt can or should be accumulated with abandon," it would be interesting to know who he thinks he is arguing with. The vast majority of the left have argued that we need not worry about deficits now. There are few, if any, economists on the left or elsewhere who say that debt levels or deficits would never be a problem. So, let's give Professor Summers a big victory over his straw man opponent!
(Only one link allowed per comment)