The observed shift in statistical demand-for-money relationships during the mid-1970s was once thought to reflect an unexplainable change in behavior. More recently, economists have recognized that the conventional regressions inadequately represented the demand for money. Specifically, the standard models overpredicted money demand during the 1970s since they failed to capture the effects of sophisticated cash management techniques.
In "An Investigation of Cash Management Practices and Their Effects on the Demand for Money," Michael Dotsey examines ways of augmenting the conventional models to overcome this problem. By looking at the causes of changes in cash management practices, Dotsey finds four variables related to cash management, which he tests for ability to explain the mid-1970s shift in a standard regression explaining the demand for money. Each of the proxies reduces the instability of the equation. Indeed, one such proxy, the number of electronic funds transfers over the Federal Reserve's wire system, captures the entire shift in the conventional model in the 1970s.
Amanda L. Kramer