How robust is the Fed reaction function to changes in the output-gap specification?

Abstract

Using alternative measures of the output gap, we estimate a forward-looking monetary policy reaction function, similar to those estimated by Clarida et al. (1998, 2000), for Greenspan's period as a chairman of the Board of Governors of the US Federal Reserve System. We are able to show that the reaction function that considers an output gap based on the industrial production index instead of the unemployment rate better captures the behaviour of the Federal Reserve during Greenspan's period.

Notes

¹ More precisely Greenspan's period as a chairman of the Federal Reserve Board spans from August 1987 to January 2006.

² We are referring to the finding that the Fed has been less accommodating to inflation since 79. Clarida et al. (2000) and Mehra (1999) provide the same result with historical data.

³ At least since October 1982 the Fed has been openly targeting the Fed Funds rate.

⁴ This sections follows Favero (2001).

⁵Bernanke and Mihov (1998). We also estimate the main equation assuming r_t follows an AR2 process. The main results are very similar to the benchmark case.

⁶ We are aware of the criticism of Newey and West (1987) regarding the use of the Hansen test for overidentification. However, given the sample size and the level of confidence, we may still be able to infer parameter specification problems from the J-test in this case.

⁷ Given that we are dealing with monthly data, the penalty parameter is set to λ = 14 400.

⁸ To calculate the Fed Funds rate intended, we calculated the monthly average of the series federal funds target rate (DFEDTAR), which is available from the Federal Reserve Bank of St. Louis data set website.

⁹ For other tests see Elliot and Stock (1996), Jondeau et al. (2004), Kwiatkowski et al. (1992).

¹⁰ Most tests perform on unemployment has used much larger sample. A seminal article on the topic Nelson and Plosser (1982).

¹¹ CGG looked at data until 1994:12.

¹² For further discussion see Greenspan (2004), McCallum (1999) and Woodford (2001).