ABSTRACT
The inflation rate has ambiguous effects on income inequality, implying that the effects could be affected by another variable. This paper examines the implication of institutional quality on the relationship between inflation and income inequality. The two-step system generalized method of moment is applied to the unbalanced panel dataset which consists of 4-year non-overlapping average data from 1987 to 2014 for 65 developed and developing countries. The coefficients of inflation and institutional quality indicate that an increase in inflation will worsen income inequality, while better institutional quality will improve income inequality. Meanwhile, the effect of inflation will be mitigated by better institutional quality, suggesting the existence of a mediating effect from institutional quality. On the other hand, the marginal effects suggest that inflation and institutional quality will reduce income inequality. Thus, policymakers are advised to improve the institutional quality as it has a direct as well as an indirect impact on income inequality via its interaction with inflation.
Acknowledgments
This study was supported by the Tun Ismail Ali Chair (TIAC) research grant under grant TIACRG2018.2. An earlier version of this paper was presented at the 4th TIAC-BNM Monetary and Financial Economics Workshop.
Disclosure statement
No potential conflict of interest was reported by the authors.
Notes
1 According to Colciago, Samarina, and de Haan (Citation2019), a monetary policy affects income inequality via the inflation channel and the conventional channel (changes in interest rate). This paper does not treat inflation as a proxy for monetary policy since the results are not confirmed by testing with interest rate.
2 To avoid the high correlation between interaction term and and , the interaction term is regressed on the and . Then, the residuals from the regression were used to represent the interaction term (Azman-Saini, Baharumshah, and Law Citation2010).
3 The statistical significance of the marginal effects is decided by measuring the t-statistic from standard errors; see Brambor, Clark, and Golder (Citation2006) for the formulae to measure the standard errors.
4 The Cook’s distance outlier test identifies the outliers from the dataset and are excluded from the estimations. The countries list is available upon request.
5 The sample is also separated into countries identified as high (low) inflation and developing (developed) countries. The findings from the high inflation and developing countries are generally the same as the full sample estimation. However, the results are not robustly supported in the low inflation and developed countries. It could due to the small sample size. We would like to thank two anonymous referees for these suggestions.