by F. Owen Irvine and Scott Schuh
No. 3, January 2005 - June 2005

Motivation for the Research
As has been widely observed, the volatility of GDP has declined since the mid 1980s compared with prior years. One leading explanation for this decline is that monetary policy improved significantly in the later period. Most of the initial studies investigating volatility reduction examined aggregate data. More recent studies have used disaggregated data, which offer the opportunity to exploit cross-sectional variation across industries for better identification of hypothesized explanations.

In this paper, the authors investigate the better-monetary-policy explanation of reduced volatility of GDP, analyzing a cross-section of 2-digit manufacturing and trade industries.

Research Approach
Since a major channel through which monetary policy operates is variation in the federal funds rate, the authors hypothesized that industries that are more interest sensitive than others should have experienced larger declines in the variance of their outputs in the post-1983 period. Using quarterly data from the Bureau of Economic Analysis for the manufacturing and trade sector and a variety of vector autoregression (VAR) models, the authors estimate for each industry three interest-sensitivity measures—the standard deviation of the impulse-response function (IRF) of sales to a shock to the federal funds rate, the cumulative IRF of sales to a change in the federal funds rate, and the sum of the lagged coefficients on the federal funds rate in the sales equation of the VAR. They then run cross-sectional regressions explaining industry output volatility reductions as a function of these interest-sensitivity measures.

Key Finding

• Although the estimated coefficients in the cross-section regressions are generally of the expected sign, there is little evidence of a statistically significant relationship between industry output volatility reductions and the authors’ measures of interest sensitivity.

Implications
Although the findings do not rule out conclusively the better-monetary-policy hypothesis, they pose challenges for the hypothesis that improved monetary policy is a major factor in the decline in GDP volatility. In particular, monetary policy must have improved in such a way as to reduce the output variances of all industries without influencing interest-sensitive sectors relatively more.

These results using manufacturing and trade data are based on studying only a subset of the goods sector of GDP. It is possible that the major channel through which improved monetary policy lowered the volatility of GDP was by reducing the variance of the structures sector and the covariance of the structures sector with the goods sector. However, from the authors’ decomposition of variance in a previous paper, this channel accounts for at most about 28 percent of the reduction in GDP volatility.

Full text of Working Paper 05-4