A form of conventional wisdom has developed in the years since the beginning of the Great Recession about financial crises: When economies go through major systemic failures of their financial systems the ensuing economic recoveries will invariably be tepid and prolonged. This view is based primarily on research by economists Carmen Reinhart and Kenneth Rogoff, particularly their book, “This Time Is Different,” which looked at hundreds of years of data on financial crises.
A new working paper indirectly challenges that conventional wisdom by digging deeper in the question of how financial crises affect economic output. The new paper, by Christina D. Romer and David H. Romer of the University of California-Berkeley, isn’t a direct rebuttal to the work of Harvard University’s Reinhart and Rogoff and others who have done similar work. Rather, the new paper is a refinement of those analyses that may be more useful to policy makers in high-income countries today.
The biggest difference between the work by Romer and Romer and other research is that the two Berkeley economists focused specifically on financial crises in rich economies in the latter half of the 20th century, specifically from 1967 to 2007. In contrast, Reinhart and Rogoff looked at the long historical relationship between financial crises and the pace of economic recoveries, meaning their data set include crises that happened centuries ago and/or in developing countries. This kind of analysis is important and necessary, but its utility as guidance for policymakers in the United States and other advanced countries today is quite low.
So what do Romer and Romer find? According to their analysis, the impact of financial crises, or “financial distress” in their terms, doesn’t, on average, lead to large recessions. When a moderate financial crisis hits a country, Romer and Romer find that the decline in economic output is about 3 to 4 percent. When it comes to subsequent growth in gross domestic product, the hit leads to a long-lasting decline in output growth as the recovery take a while. But that result is merely the average. Japan, with its well-known “lost decade,” is driving the entirety of the result.
Then the two economists take a look at what happens to economic output during very large financial crises, such as those experienced by Scandinavian countries and Japan during the 1990s. Instead of finding a hard and fast relationship, Romer and Romer find a large variation in the response of economic output. Sometimes, as in the case of Japan, the crisis leads to significant recessions and slow recoveries. And sometimes, such as with Norway in the early 1990s, the effect was pretty much non-existent.
The reason for this variation is an open question. Romer and Romer wonder if policy responses aren’t the main driver.
So it seems that on the whole, the Reinhart-and-Rogoff result doesn’t seem to hold up for advanced countries in the recent history. A large financial crisis doesn’t necessarily mean a large economic downturn has to happen. Economic events aren’t forces that sweep over us, but things that we can react to and very possibly control.