The econometric consequences of nonstationary data have wide-ranging implications for empirical research in economics. Specifically, these issues have implications for the study of empirical relations such as a money demand function that links macroeconomic aggregates: real money balances, real income and a nominal interest rate. Traditional monetary theory predicts that these nonstationary series form a co-integrating relation and, accordingly, that the dynamics of a vector process comprising these variables generates distinct patterns. Recent econometric developments designed to cope with nonstationarities have changed the course of empirical research in the area, but many fundamental challenges, for example the issue of identification, remain. This book is an effort to determine the consequences that nonstationarity has for the study of aggregate money demand relations. The object of this study is to utilize the tools of modern time series analysis to determine the role of an aggregate demand for real balances in the generation of macroeconomic time series. A significant characteristic of this research is the identification and estimation of this demand function in a multivariate framework, in contrast to most existing studies that concentrate on a single equation framework.