The work generalizes switching regression models and limited dependent variable models in order to apply them in a dynamic context. Using the Panel Survey on Income Dynamics, the investigators will estimate the parameters of a lifecycle model involving Euler equations. Incorporated in this model will be the assumptions of a lifetime budget constraint as well as a liquidity constraint. One of the important developments in this research is the use of the method of simulated scores to estimate the model. Households do appear to switch their consumption and labor supply patterns over long periods. They are likely to be constrained early on in their lifetime, at the time of major purchases or unforseen events (such as unemployment, catastrophic illness, etc). The incorporation of the endogeneity and imperfect observability of liquidity constraints in the type of model that is being estimated will make it easier to confirm the importance of these constraints on household behavior. Because of the methods used, the techniques can be generalized to the constraints that firms face, as well.
