Finance and Growth: A Bayesian Model Averaging Evidence

The question whether financial development is conducive to economic growth has entered the debate with new intensity following the financial crisis of 2007-2008. We use standardized dataset on economic growth established by the literature and Financial Development Database by World Bank to inspect the relationship. Unlike other studies, we employ Bayesian Model Averaging (BMA) to address model uncertainty inherent to modelling of economic growth. Apart from dealing with omitted variable bias it also allows us to compare relative importance of banking sector and financial markets along with their varying characteristics. Examining real economic growth rates 1960-2011 in 68 countries, we find little evidence in favour of traditional financial development proxy - financial depth - to affect economic growth. Our initial results point to the importance of banking sector efficiency, approximated by net interest margin, as essential growth determinant. Moreover, we use financial indicators to construct overall measure of financial development and find it highly relevant to economic growth. The results are robust to different parameter and model priors in BMA, but not to specifications dealing with potentially endogenous nature of the finance-growth correlation