Emerging economies, turnover rates and inflation variability: A comparison of generalized maximum likelihood and SUR models

Theoretical exposition and empirical evidence in central bank independence (CBI) literature confirm an inverse relationship between inflation and measures of CBI mostly in developed economies. Based on this ex ante information on CBI-inflation tradeoff, this paper proposes two functional forms for the diagonal and off diagonal elements in the residual covariance matrix in the estimation process. The proposed functional forms are used in a generalized maximum likelihood and then in a generalized least squares (GLS) (with the restricted covariance matrix) framework for the empirical test. The results are compared to the outcome of an SUR model (unrestricted). The tests involve 14 emerging economies and covers the period 1960-1990. Compared to SUR, majority of results of GLS model in samples with and without outliers provide stronger and more significant evidence confirming the CBI-inflation tradeoff. Notably, the standard errors of the GLS estimates are lower than that of the SUR estimates. Without outliers, the GLS estimates show even lower standard errors as compared to the outcome of the SUR model. Low standard errors provide baseline indication of more accurate estimates.