Thursday, April 17, 2014

What Happened to the Great Moderation?

In the 1990s and into the early years of the 2000s, it was common to hear economists speak of a "Great Moderation" in the U.S. economy. After the economic convulsions of the 1970s and early 1980s, in particular, the path of the U.S. economy seemed to have smoothed. To be sure, there was an 8-month recession in 1990-91, and another 8-month recession in 2001. But both recessions were fairly mild: unemployment topped out at 7.8% in the aftermath of the 1990-81 recession, and reached only 6.3% in the aftermath of the 2001 recession. And the recessions seemed more scarce: the average length of an economic upswing since World War II has been 58 months, but the upswing before the 1990-91 recession was 92 months, and the upswing before the 2001 recession was 120 months.

Of course, after 2007 when the Great Recession had crashed the party, talk of a Great Moderation seemed disconnected from reality. Jason Furman, chair of President Obama's Council of Economic Advisers, has taken on the question of "Whatever Happened to the Great Moderation?" in an April 10 speech.

Furman makes the interesting point that even now, including the Great Recession and its aftereffects in the data, the level of short-term volatility in economic statistics like quarterly GDP or monthly job growth seems to be lower than it was from the 1950s to the 1970s, not only in the United States but also in other high-income countries. (Of course, "less volatile" doesn't mean "healthy growth rate.")

Peering into the inner workings of the US economy, Steven J. Davis and James A. Kahn provided an overview of the evidence in the Fall 2008 issue of the Journal of Economic Perspectives in "Interpreting the Great Moderation: Changes in the Volatility of Economic Activity at the Macro and Micro Levels."  (The article, like all articles in  JEP, is freely available on-line courtesy of the American Economic Association. Full disclosure: I've been Managing Editor of the journal since its inception in 1987.) They find that the drop in short-term volatility of GDP can largely be traced to a drop in the volatility of production of durable goods. The volatility of production of nondurable goods falls only a little, and production of service was never that volatile to begin with. Volatility of production inventories declined substantially, too.

Furman points out an intriguing pattern here: "From 1960 to 1984, inventories were quite volatile, and were also procyclical, meaning that when sales increased, inventories also increased, further contributing to the volatility of production. During the post-1984 Great Moderation period, inventory investment itself became much less volatile, and the previous relationship between inventories and sales reversed, so that the two became negatively correlated. Focusing specifically on durable goods, the change in the covariance between inventories and sales accounts for nearly half of the decline in the variance in durable goods output. However, including the Great Recession, it appears that the relationship between output, sales and inventories partially reverted to the pre-Great Moderation pattern. The covariance of inventories and sales turned positive again, suggesting that improved inventory management was not enough to cushion the massive blow of the Great Recession, and in fact exacerbated it." Furman is careful to note that the argument that inventories have become procyclical is based on only a few years of data.  But if the pattern continues, it will need exploring and explaining.

Another pattern here is that consumption patterns have continued to show less short-term volatility, even through the Great Recession. Furman writes: "Disaggregating the GDP data, the reduced volatility of consumption is one of the major sources of the Great Moderation—and this reduced volatility has continued to hold up during and after the Great Recession, especially in consumer durables. The continued stability in consumption stands in contrast to other components of GDP like business fixed investment, which became less volatile during the initial Great Moderation but has since at least partially reverted to its earlier volatility."

Improvements in macroeconomic policy offer another potential explanation for the Great Moderation: that is, monetary policy was less disruptive after the mid-1980s than it had been in, say, the 1970s. The use of fiscal policy to stimulate the economy during downturns arguably became more purposeful and effective. Indeed, as Furman points out, one can make a case that monetary and fiscal policies helped to prevent the Great Recession from being even greater (citations omitted here, and  throughtout):

"Improvements in monetary and fiscal policy have likely contributed to the patterns in the high-frequency data originally identified as the Great Moderation, although one could debate the share of the credit they deserve. I believe policy steps have also played a critical role at lower frequencies as well, with the best example being the Great Recession itself, which in many ways started off looking like it could be as bad or worse than the Great Depression. To appreciate this point, consider that the plunge in stock prices in late 2008 proved similar to what occurred in late 1929, but was compounded by sharper home price declines, ultimately leading to a drop in overall household wealth that was substantially greater than the loss in wealth at the outset of the Great Recession. . . .Moreover, Alan Greenspan (2013) has argued that short-term credit markets froze more severely in 2008 than in 1929, and to find a comparable episode in this regard one has to go back to the panic of 1907. However, in large part because of an aggressive policy response, the unemployment rate increased 5 percentage points, compared to a more than 20 percentage point increase in the Great Depression from 1929 to 1934. And real GDP per working age population returned to its pre-recession peak more quickly in the United States than in other countries that also experienced systemic crises in 2007-08."
The pattern that emerges from Furman's discussion is that the Great Moderation was quite real as measured by smaller short-term fluctuations in GDP, employment, consumption, production of durable goods, and inventories. Even more surprisingly, many of these factors (although not inventories) have continued to show lower short-term volatility in the aftermath of the Great Recession. But of course, this lower level of short-term quarter-to-quarter or month-to-month economic fluctuations did not protect the economy from the enormous economic blow of the Great Recession, which lasted 18 months, spiked the unemployment rate from under 5% in mid-2007 to 10% in October 2009,m and then has been followed by years of frustrating sluggish (and without a lot of short-term volatility) recovery.

One possible interpretation here is that the Great Moderation is real, and the Great Recession was a sort of perfect storm, best understood as a one-off divergence from the long-run trend. Another possible interpretation is when short-term volatility is lower and when recessions become milder and less common, firms and households become less wary of risk, and more willing to take chances--which in turn leads to the kind of risky conditions that can create the underlying conditions for a deeper recession.  And yet another interpretation is that while the old vulnerabilities that led to the economic volatility of smokestack industries back in the 1950s and 1960s have declined, the U.S. and world economy how face some new vulnerabilities due to changes in technology, globalization, and the financial sector. In this view, the Great Recession was only a first foretaste of the kinds of disruptive interactions that can occur in this new economic configuration.