Structural relations among variable coefficients: with applications to demand systems and labor supply
Describes a regression model in which the variable (random) coefficients are structurally related. The model utilizes repeated measurement of economic units. In some cases, as in demand systems, the model may be used to test theoretical constraints at the level of the individual economic unit by testing the constraints on the joint distribution of parameters among units. In other cases, as in labor supply, the variable coefficients may themselves be interpreted as economic variables which follow a simultaneous equations scheme. Issues of identification and estimation are addressed briefly. Both maximum likelihood and generalized least squares procedures are developed.