Party affiliation rather than former occupation: the background of central bank governors and its effect on monetary policy

Abstract

We find that tenures of central bank governors who are affiliated with a political party are characterized by a relatively dovish monetary policy stance. In general, party affiliation is more important than occupational background for the degree of hawkishness.

Keywords:

• central bank governors
• monetary policy
• occupation
• party affiliation
• Taylor rules

JEL Classification:

• E31
• E43
• E52
• E58

Acknowledgements

The authors would like to thank an anonymous referee, Marcus Drometer, Matthias Uhl, and participants of the 2014 Annual Meeting of the European Public Choice Society in Cambridge, United Kingdom, for their helpful comments on earlier versions of the article. The usual disclaimer applies.

Notes

¹ The null hypothesis of a unit root in all panels is rejected for all three variables. The Im et al. (2003) test statistics with eight lags are W_tbar = −3.4** (central bank rate), W_tbar = −2.0* (inflation rate) and W_tbar = –12.4** (output gap). The Fisher-type unit-root test statistics based on augmented Dickey–Fuller (1979) tests with eight lags are P = 73.2** (central bank rate), P = 56.9* (inflation rate) and P = 270.2** (output gap). The Fisher-type unit-root test statistics based on Phillips–Perron (1988) tests with eight lags are P = 131.8** (central bank rate), P = 84.9** (inflation rate) and P = 409.0** (output gap).

² We weight our estimates based on the assumption of contemporaneous correlation between the cross-sections in order to account for potential spillover effects, as central bank rates, inflation rates and output gaps exhibit a notable degree of contemporaneous correlation across the sample countries. Note that the inclusion of cross-sectional dependence in the weighting matrix can also be interpreted as a proxy for time fixed effects.

³ Data sources: Central bank rates – IMF; inflation rates and output gap (based on industrial production and the Hodrick–Prescott filter with λ = 1600) – OECD. Since Orphanides (2001), real-time data for inflation and output growth are often used in TR estimations. We cannot use real-time data because of limited data availability for our sample period. For a similar reason, we employ a contemporaneous TR specification, whereas other scholars have employed forward-looking TR specifications. Again, the nonavailability of inflation and output growth expectations for our sample period restricts our degrees of freedom in the econometric set-up. See also Belke and Potrafke (2012).

⁴ We define two dummy variables; the first takes the value 1 for party members, the second takes the value 1 for nonparty members. Then, we consecutively interact both dummies with the already existing interaction terms. Note that we do not distinguish between members of leftist parties and rightist parties, since preliminary estimations (not shown) indicate that there are no party ideology effects when it comes to interest rate setting. The estimates for governors with a specific occupational background on the one hand and without a party affiliation on the other hand (upper half of Table 1) are based on 285/831/729/409 observations for those who have gained working experience in academia/the central bank staff/politics/the private sector. Similarly, the estimates for governors with a clearly documented party affiliation (bottom half of Table 1) are based on 227 (academia)/256 (central bank staff)/570 (politics)/407 (private sector) observations. Therefore, we are confident that our results are not driven by a few outliers.

⁵ The p-value for the private sector coefficient on output is 0.09.

⁶ The p-value of the effect of private sector experience on the equilibrium nominal interest rate is 0.07.

⁷ However, neither coefficient is significant at the 5% level.