|
Article Excerpt
THE VOLATILITY OF the U.S. economy has declined dramatically. The standard deviation of annualized changes in quarterly real seasonally adjusted gross domestic product (GDP) declined from 1.1 percentage points in 1965-84 to 0.5 percentage points in 1984-2004. The stabilization of inflation has been similar. This "Great Moderation" has been described as one of the most striking changes in the business cycle in recent decades (Stock and Watson 2003, Bernanke 2004). It is the subject of a large and growing literature, of which McConnell and Perez-Quiros (2000), Kim and Nelson (1999), and Blanchard and Simon (2001) are prominent examples.
However, what matters to most people is not volatility but uncertainty. If a change in inflation or activity is expected, then people adjust to it. That may involve a change in prices or a shift in production from one period to another, but interviews with firms (Blinder et al. 1998) and everyday observation suggest these adjustment costs are often small. In contrast, when people do not know the future, investments occur when they should not, goods and assets are mispriced, precautions are taken against events that do not occur, and so on.
Perhaps the clearest evidence of the importance of uncertainty relative to volatility is the lack of interest in seasonal economic variations. Seasonal variations are huge, accounting for about 85% of the quarterly variability of output (Beaulieu and Miron 1992, Table 1). But because they are predictable, almost no one pays attention to them at a macro-economic level. Even the studies of so-called "volatility" use seasonally adjusted data. They do not measure the total variation in the data, only the variation not accounted for by one specific influence. But there is no obvious reason for singling out seasonality. Just as predictable seasonal variations are appropriately removed from the data, so should other predictable influences.
If one is interested in unpredictability, one can measure it directly, as the difference between actual outcomes and what people were expecting. There are many available measures of expectations. I use the forecasts of the staff of the Federal Reserve Board of Governors, as published in a document called the Greenbook. Differences between these forecasts and actual outcomes are the Greenbook errors.
The Greenbook errors provide an interesting measure of uncertainty for several reasons. Previous researchers have found that the Greenbook forecasts have been at least as accurate as other forecasts (Romer and Romer 2000, 2008, Sims 2002, Bernanke and Boivin 2003, Reifschneider and Tulip 2007). So they can be viewed as a good summary of the available information and as the near the frontier of best performance in forecasting. Furthermore, the data on Greenbook forecasts are richer than for many private sector forecasts. The forecast horizon is longer and the data extend further back in time. (1)
One limitation of Greenbook forecasts is that they have sometimes been based on unusual assumptions, such as a constant funds rate. (2) However, the competitive forecasting performance of the Greenbook suggests that any such handicap was unimportant: it still provided at least as reliable a guide as alternatives. A stronger limitation is that the forecasts are confidential. So although they are interpretable as reflecting what people could or should have expected, they do not directly measure what private sector decisionmakers did expect. But then, it is not clear that any point estimate serves this purpose.
The Greenbook errors are especially interesting for analysis of monetary policy, given the document's authorship and intended readership. From a historical perspective, changes in the quality of the forecasts might help explain changes in policy performance. From a normative perspective, the accuracy of forecasts and its stability help determine the extent to which monetary policy should be "forward looking." Last, if the forecast errors are stable over time then the distribution of outcomes about previous forecasts would provide a reliable guide to the distribution of possible outcomes about the current forecast. In this context, the Federal Reserve's Federal Open Market Committee (FOMC) now regularly publishes a Summary of Economic Projections, in which it compares its projections with historical forecast errors averaged over the last two decades, an approach partly based on research reported in this paper (FOMC 2007, Reifschneider and Tulip 2007).
Although this paper is partly motivated by these monetary policy issues, its primary focus is whether uncertainty has declined. I find that there has been a large reduction in uncertainty regarding inflation across horizons and output at short horizons. However, I do not find a reduction in uncertainty about output at longer horizons. I find that the reduction in uncertainty is significantly less than the reduction in volatility at both short and long horizons, for both inflation and output. The differing changes in uncertainty and volatility arise because the predictive power of forecasts has virtually disappeared. Whereas the Fed predicted a large share of the fluctuations in output and inflation in the 1970s and 1980s, more recent fluctuations have been surprises.
1. RELATED LITERATURE
The view that unpredictability is more interesting than volatility is not new. As noted above, almost all of the studies of volatility remove predictable seasonal influences from the data. Many others remove the predictions of a vector autoregression. Several papers in this literature--for example, Stock and Watson (2003, 2007)--focus on unpredictability.
However, when measures of uncertainty are presented, they are typically derived from the errors of an econometric model. That...
|
