Central bank independence and low inflation: who leads the dance?

ABSTRACT

One strand of empirical literature finds that central bank independence (CBI) lowers inflation. Another strand of literature finds that low inflation is a key determinant of reform towards increased CBI. This paper investigates whether either variable can be identified as a first instigator. Using the largest CBI dataset to date, this paper applies rolling balanced-panel Granger causality tests between CBI reform and changes in inflation. For advanced economies, CBI reform is found to significantly lead disinflation, while there is no Granger causality in the opposite direction. Instead, among emerging and developing economies, CBI reforms tend to follow quickly upon disinflation episodes, while the lags from CBI reform to disinflation are long. An interpretation is that in emerging and developing economies CBI reform often followed on crises that involved high inflation, whereas in various advanced economies a shift in thinking about central banking first triggered CBI reform.

KEYWORDS:

• Central bank independence
• monetary reform
• price stability
• granger causality

JEL CLASSIFICATION:

• E58
• O11

Acknowledgments

Hala Moussawi provided excellent research assistance.

Disclaimer

The views expressed herein are those of the author and should not be attributed to the IMF, its Executive Board, or its management.

Disclosure statement

No potential conflict of interest was reported by the author.

Notes

¹ See, e.g. De Haan and Eijffinger (2019), Goodhart and Lastra (2018), Laurens et al. (2015), and Masciandaro and Romelli (2015).

² See the overviews of the literature in Berger, de Haan, and Eijffinger (2001), Garriga and Rodriguez (2020), and Klomp and de Haan (2010).

³ See, e.g. Agur (2018; Agur 2019), Dincer and Eichengreen (2014), and Garriga (2016).

⁴ Garriga (2016) applies the most common CBI index methodology, developed by Cukierman, Webb, and Neyapti (1992). This index measures de jure independence. While this does not always translate into de facto independence, de jure independence of a central bank is often a prerequisite for de facto independence and successful monetary reform more broadly (Cukierman 2008).

⁵ See also Lopez and Weber (2017).

⁶ Advanced economies are defined as countries that are members of the OECD by the end of the sample period.

⁷ The inflation data (based on annual, end-of-period CPI indices) is sourced from IMF-WEO and IMF-IFS.

⁸ We also check outcomes with Hannan-Quinn (HQC) and Bayesian (BIC) information criteria. The HQC gives the same results as the AIC for all our (sub)samples. Instead, the BIC always selects a single lag, which is not economically meaningful here, leaving too little time for the interaction between CBI and inflation to play out.

⁹ See, for instance, Goodhart and Lastra (2018) who discuss the relation between changing socio-political attitudes in advanced economies and the degree of independence granted to the central bank. Seminal early contributions that affected the discourse on politics and central banking, and thereby the attitude towards CBI, include Barro and Gordon (1983) and Rogoff (1985).

¹⁰ For example, Abiad and Mody (2005) find that in developing countries, financial sector reform often followed on crises, and that reform is path dependent, with an initial impetus triggering further steps.

¹¹ Identified using the IMF AREAER database, and applied to the most recent sub-sample, as inflation targeting only gained prominence in the 1990s.

¹² While most emerging and developing countries are not inflation targeters, enough of them are that they constitute a slight majority of the inflation targeting group too.

¹³ An updated version of their dataset is available at kof.ethz.ch/en/data/data-on-central-bank-governors.html. Also note that in Table 3 there are fewer advanced economies than in Table 1, because there is no governor turnover data available for Euro Area countries after the formation of EMU.