Cross-country evidence on the demand for money

This paper examines the demand for real money balances using both cross-country and panel data for 48 countries over the 1980-95 period. We utilize the cross-country data to investigate conventional money demand functions and the role that institutions, financial structure and financial development may have in the demand for money. On the basis of possible heterogeneity within the cross-country data set, we use Bayesian classification and finite mixture models to partition the data and re-examine our initial regression results. We then continue by examining the conventional money demand function through a panel data context. Our empirical modeling not only utilizes traditional panel methodology, but exploits recent state-of-the art advances in the panel unit root and panel cointegration literature. Such procedures allow us to take advantage of desirable statistical properties and obtain consistent estimates in order to test long-run hypotheses.