Abstract
Using a dynamic panel data approach, we estimate the impact of the political and institutional factors on inflation. Estimation results show that a lower degree of political instability generates lower inflation only for developed and low-inflation countries. However, when political freedom is taken into account, political instability appears to be influential on inflation also for developing countries and turns out to be significant only for high-inflation countries. Such findings emphasize the inflation-reducing effects of political stability depending on democratic political structure.
Notes
1 See Brunetti (Citation1997) for a detailed survey of this topic.
2 See, for example, Acemoglu et al. (Citation2003).
3 See Appendix A for the other details of the data and its sources.
4 For a detailed discussion of the system GMM (GMM-SYS) estimation, see Arellano and Bover (Citation1995), Blundell and Bond (Citation1998).
5 See Podrecca and Carmeci (Citation2001), Kottaridi (Citation2005) and Yao (Citation2006) as examples of recent studies employing GMM approach in dynamic panel-data-model estimations.
6 The lags of the variables are selected with respect to their statistical significance in the regression.
7 For the details of the diagnostic test, see Arellano and Bond (Citation1991).
8 The SEs of the estimators are computed by a small-sample variance correction suggested by Windmeijer (Citation2000) to eliminate the downward bias in the SEs of the two-step estimators.
9 Because the regression coefficient of was not found significant either with lags when estimated without the distinction, the relevant regression estimate is not reported here for brevity but it is available upon request.
10 DEV = 1 for high-income countries and zero otherwise, whereas UNDEV = 1 for low- and middle-income countries and zero otherwise. Income classification for the countries in the sample is shown in Appendix B.
11 Linnemann (Citation2005) gives another interesting explanation of this issue in the context of a sticky-price model. He argues that raising the nominal interest rate reduces demand, but if the government balances the budget through income taxes it also reduces supply and the net effect on inflation can be positive.
12 This is in contrast with the textbook principle that the currency of any country should be allowed to depreciate when faced with adverse import price shocks to accommodate this unfavourable shift in their terms of trade, to give resources the incentive to shift out of the production of importables (Frankel, Citation2005).
13 See, for example, Taylor (Citation2000) and Devereux and Engel (Citation2001).
14 LOW = 1 for countries with one-digit rates of inflation and 0 otherwise, whereas HIGH = 1 for countries with two-digit or higher rates of inflation and 0 otherwise (see Appendix B).
15 EU = 1 for EU countries and 0 otherwise, whereas NONEU = 1 for non-EU countries and 0 otherwise (see Appendix B).
16 Considering the possible importance of geographical proximity in the cross-border transmission of political and economic shocks, we also made use of distinctions for Latin American and Asian countries. However, estimates of those regressions, which are not reported here for the sake of brevity, did not provide any statistically significant evidence for the inflationary effect of political instability in these groups of countries.
17 FREE = 1 for politically free countries and 0 otherwise, whereas NOTFREE = 1 for politically not-free countries and 0 otherwise. Political freedom statuses of the countries in the sample are obtained from the Freedom House database (see Appendix B).
18 Note that higher values of the Gastil Index mean lower degrees of political freedom.