US Treasury primacy in monetary policy signalling: a public choice perspective

Abstract

This note reviews the signalling models presented in the monetary economics literature, and offers a supplementary interpretation regarding the observed US Treasury primacy in signalling. It is argued here that the legal authority given to the US Treasury, under the Gold Reserve Act of 1934, for managing the exchange value of the dollar in international markets provides an avenue for the Treasury, and thus the Administration (i.e., the Executive Branch), to use foreign exchange intervention policy to reduce the credibility of the Federal Reserve's monetary policy. This legal relationship is likely the source of much tension between the two institutions, especially during periods for which the Administration and the Federal Reserve are at odds regarding the proper direction for monetary policy. Given that the Federal Reserve is aware of this implicit power of the Treasury, it should not be surprising that monetary policy signals from that quarter have been found to have a dominant influence on monetary policy.

Acknowledgements

The authors thank Mark Taylor and Troy Gibson for helpful comments. Any remaining errors are their own.

Notes

¹ Havrilesky's coding of SAFER was very labour-intensive. Research assistants retrieved articles from The Wall Street Journal – using its ‘What's News?’ digest – that mentioned the monetary policy views of the Administration. Each mention was coded as to the particular source within the Administration. This categorization included mentions from the President/Oval Office, the Council of Economic Advisors, the US Treasury, and unidentified/other sources.

² Repeated coverage of the same news item was not counted (Havrilesky, 1995: 42). See Havrilesky (1995) for more on the construction of the SAFER index.

³ The data are truncated for two reasons. First, very little signalling occurred during the 1950s. Second, omission of data from 1961–1963 removes the impact of the ill-fated ‘Operation Twist’ strategy by the Administration to invert the yield curve. Inclusion of this latter period would result in ambiguous implications for the SAFER index (Havrilesky, 1995: 121).

⁴ The first-difference form corrects for non-stationarity regarding the dependent variable (see Enders, 1995).

⁵ As Havrilesky (1995: 119–20) points out, two time lags are important when estimating the response of the federal funds rate to signalling. First, and of greatest concern, is the lag from the date of a communication from the Administration until the date that the central bank initiates a change in the federal funds rate. The initial response lag envelopes a reporting lag from the date of communication to the date the signal is captured by the news media. The three week sum for SAFER seemed to perform best in this regard. See Havrilesky (1995) for more on these points.

⁶ For further details see Broaddus and Goodfriend (1995).

⁷ See Carlson and Lo (2004) for a discussion of the relationships between monetary policy, the exchange rate, and the degree of sterilization. Also see Darrat et al. (2003) for empirical evidence on the relationship between the exchange rate and inflation in the United States.

⁸ For details on this period see Pauls (1990).

⁹ From the beginning this relationship between the Treasury and the Federal Reserve was controversial in the latter institution. An extensive description of how this relationship developed is given in Hetzel (1996).

¹⁰ For a review of this debate see Edison (1993).

¹¹ This would tend to explain the mixed results obtained when attempting to use intervention activities as a ‘signalling’ channel for monetary policy changes in the USA. It would also tend to explain the recently reported result that intervention does not signal changes in the federal funds future rate, but does increase the conditional variance of that variable. For details on this see Fatum and Hutchison (1999).

¹² For more exact details of the nature of this consultation see Humpage (1994).

¹³ On this point see Humpage (1994) or Broaddus and Goodfriend (1995).

¹⁴ A clear discussion of this is contained in Broaddus and Goodfriend (1995). Other Federal Reserve publications discussing this issue include Humpage (1994) and Hetzel (1996).