We find that in the vast majority of cases output never returns to previous trends. Indeed there appear to be more cases where recessions reduce the subsequent growth of output than where output returns to trend. In other words “super hysteresis” to use Larry Ball’s term is more frequent than “no hysteresis.”
Pointer from Mark Thoma.
In the AS-AD paradigm, AS and AD are independent equations. The long-run aggregate supply curve is what the economy departs from when it goes into recession and what it returns to when there is a recovery. What Summers and co-authors are finding is that the economy is more like Charlie and the MTA. It never returns to the previous long-run aggregate supply curve, and instead the aggregate supply curve appears to shift adversely in response to recessions.
Summers’ explanation for the anomaly might be that when people lose their jobs they lose some of their skills. However, another explanation that fits the data is that people who lose their jobs lose them because their skills were valuable only in patterns of specialization and trade that are no longer sustainable.
Summers points to this paper, where he and Antonio Fatas find a correlation between fiscal consolidations and downward revisions to potential GDP. I worry that both fiscal consolidation (trying to reduce unsustainable budget deficits) and downward revisions to potential GDP would be lagged responses to adverse “supply shocks” (or, in my preferred framework, an unusually large number of specialization and trade patterns becoming unsustainable).
It is always amazing to watch economists debate changes in “Real” GDP without stepping back and thinking about how the number is calculated. As I try to make clear in my own blogpost — GDP and CPI: Broken beyond repair — the GDP calculation emerges from a massive mashup of different numbers and adjustments that are all highly subjective and dubious. Why does GDP growth after a recession? Perhaps the bureaucrats are more pessimistic about growth, so when they do the adjustments they just adjust it lower than normal. Or conversely they were overly optimistic during the boom phase, and now are compensating the other way. Or perhaps more likely, the “real” GDP still substantially driven by changes in nominal GDP, which can never be perfectly adjusted for. So the lower “real” GDP growth trajectory is simply reflecting lower nominal GDP growth, which reflects that it takes time for balance sheets to get repaired after a debt deflation crisis.
This whole exercise consists of 1) observing changes in an insane number that is calculated in a way that no single human likely understands 2) trying to figure out what drives those changes. If economists are trying to figure out cause-and-effect in the economy, they should start from numbers that are actually comprehensible. To do otherwise is just silly.
Of course one of the problems is that AS and AD are not independent of each other. AD comes from the expenditure side of the circular flow model while AS comes from the factor market (income) side.
If there were a dearth of innovative investment opportunities during the Great Depression, I would be more inclined believe this but it takes more of an emphasis on the innovations not having reached a level of maturity for investment, something that seems likely to respond to inducement.