Social and Behavioral Sciences Commons^™

The Problem Of Aggregation In Spatial Equilibrium Analysis, Brian Schmiesing, Steven Blank

Economics Staff Paper Series

Aggregation across time has long been recognized as a potential source of specification error within time series modeling. This source of specification error has major implications for the use of causality tests in spatial equilibrium analyses. This study reviews the theoretical econometric literature to generate hypotheses which are empirically tested to illustrate the impacts of aggregation on causality tests. Granger causality tests for three major corn markets reveal one-way causality to be "fragile" with respect to changes in the level of time aggregation in data. Dynamic multipliers are discussed as being one method of checking the specification of the model.