Transparency on inflation of OECD countries? An Application of LSDVC Estimator on a dynamic Panel Model

We outline the implications of central bank transparency coupled with economic globalization on the effectiveness of monetary policy at achieving low and stable inflation, through a two-step empirical analysis on a sample of 34 OECD countries. Our results are fourfold: (i) There is a highly significant negative impact of economic globalization (measured by the composite sub-index of Dreher et al., 2008) on inflation (ii) An appropriate and efficient U shape test proposed by Lind and Mehlum (2010), indicates a robust optimal intermediate degree of transparency under the first estimation period (1998-2010), but suggests new evidence as to its level differently from van der Cruijsen et al. (2010). (iii) Although our findings are generally confirmed by a wide set of robustness tests, some caveats associated with the length of time period and the characteristics of inflation pattern observed under the second estimation period (2006-2015) should be taken into account. The estimations were run using a bias corrected Least Square Dummy variable (hereafter, LSDVC), developed by Bruno (2005) for short dynamic panels with fixed effects, and extended to accommodate unbalanced data. Alternative results using Generalized Method of Moments (hereafter GMM) estimators: (Arellano and Bond, 1991, hereafter AB; Blundell and Bond, 1998, hereafter BB) are also provided. (iv) Overall, we find that LSDVC estimator outperforms GMM estimators as it exhibits a satisfactory fit of our hypotheses.