Disaggregation in Econometric Modelling

Aggregation is a significant problem in much economic theory and in most applied work in economic modelling. A proper understanding of the problem and the methods chosen to resolve it are crucial to the design and evaluation of applied research in economics. In this book, leading theorists and applied economists address themselves to the key questions of aggregation: what level of aggregation should be chosen in applied research?; when is an aggregate approach justified?; is there an optimal level of disaggregation?; does disaggregation improve predictive performance? and why?; how can aggregation bias be detected?; how does aggregation obscure casual relationships?; can macro-equations be derived from non-linear micro-equations?; what methods are available for linking macro and micro models? and how do they work? These issues are covered both theoretically and in wide-ranging applications.

Table of Contents:
Part 1: aggregation of time-series variables - a survey, C.C.W.Granger; on the dynamic specification of aggregated models, M.Lippi and M.Forni; aggregation versus disaggregation in forecasting construction activity, P.Ilmakunnas; aggregation problems in a model of wage formation and employment demand, E.Deutsch and K.Rodler; aggregation bias in labour demand equations for the UK economy, K.Lee, et al; optimal aggregation of linear net export systems, E.E.Leamer; the choice of aggregate production functions in Mexican industries, T.Sterner; disaggregation of the demand for hospital care, R.C.J.A.Van Vliet and E.K.A.Van Doorslaer; aggregation in discrete choice models - an illustration of non-linear aggregation, A.C.Cameron. Part 2: macroeconomic and microeconomic modelling - some issues, C.Higson and S.Holly; integrating a micro labour-supply model and a macro model of the Dutch economy, P.T.Hacken and A.Kapteyn; macro to micro linkage - experiments with the UK commodity flow accounts, T.Barker; what determines the bank borrowing and liquid lending of UK companies - explanation based on aggregated and disaggregated data, C.Green, et al; threshold problems in modelling the company sector, G.Meeks.