Factors leading to inflation targeting – the impact of adoption

Abstract

This article examines how the analysis of inflation targeting (IT) adoption is affected by allowing for a structural change after adoption, using panel probit models for 60 countries over the period 1985 to 2008. Our findings suggest that there is a structural change after IT adoption. Including the post-adoption period when estimating the factors of IT adoption leads to biased results when interested in the question as of what drives countries’ decision to adopt IT.

Keywords:

• inflation targeting
• panel probit
• factors of adoption
• structural change

JEL Classification:

• E42
• E52

Acknowledgements

We thank Jakob de Haan and Jan Jacobs for their helpful comments.

Notes

¹ Note that three EU members (Finland, Spain and Slovakia) abandoned IT when they joined the euro area. However, this decision was caused by the institutional commitment to adopt the euro and to unify their monetary policy conduct with the European Central Bank (ECB). Although these European Economic and Monetary Union (EMU) countries gave up explicit IT, their new monetary strategy under the ECB framework resembles implicit version of IT.

² In this article, endogeneity of IT is understood in a broader economic sense – endogenous means ‘having an internal cause or origin’. Thus, when we say that IT is endogenous, we infer that its continuation is internally affected by the institutions and economic conditions that are shaped under the IT regime.

³ Pooled probit specifications do not lead to qualitatively different results (see Table A3). A fixed effects model would drop all countries that did not adopt IT and has therefore not been used in previous studies of IT adoption either.

⁴ As a robustness check, we use a different specification of D(τ), namely , where , larger ρ means a slower transition. The results with this specification are qualitatively similar to the ones with D(τ) specified in Equation 2. These results are available in Samarina and Sturm (2013).

⁵ Alternative dates for soft IT and full-fledged IT adoption are used in a robustness check and available on request.

⁶ The data set contains six additional variables that have also been used in previous studies (e.g. Hu, 2006; Mukherjee and Singer, 2008): output growth, fiscal balance, trade openness, external debt, financial structure and actual central bank instrument independence. Including these variables as well does not change the conclusions. Results are available in Samarina and Sturm (2013).