Disequilibrium, self-selection, and switching models.

*(English)*Zbl 0603.62115
Handbook of econometrics, Vol. 3, Handb. Econ. 2, 1633-1688 (1986).

[For the entire collection see Zbl 0588.00025.]

The title suggests that three different types of econometric models are reviewed. But as the author remarks the general model is the switching regression model and the two remaining ones can be derived from it by suitable assumptions and restrictions on the parameters and/or the covariance matrix. This close relationship between these models is explored in the introduction of the paper. Consequently, the first main body of the review article is concerned with switching regression models. Within this class of models the author distinguishes switching regression models with sample separation known, sample separation unknown, and the case of imperfect separation information. The first part closes with the topic of switching simultaneous systems.

Disequilibrium models constitute the second main part of the survey. Starting with some well known examples of the literature different disequilibrium formulations are discussed such as the directional methods, the price adjustment equation and the so-called min-condition. Additional problems of specification in disequilibrium models follow. Special emphasis is given to the problem of serial correlation in the error terms and testing for distributional assumptions and disequilibrium. Extensions to multimarket disequilibrium models close this section.

Models with self-selection are discussed in section nine of the paper, and the last topic discusses multiple criteria for selectivity.

The paper is an excellent survey over switching regression models and the extended bibliography enables the reader to go easily into interesting details of this branch of econometrics.

The title suggests that three different types of econometric models are reviewed. But as the author remarks the general model is the switching regression model and the two remaining ones can be derived from it by suitable assumptions and restrictions on the parameters and/or the covariance matrix. This close relationship between these models is explored in the introduction of the paper. Consequently, the first main body of the review article is concerned with switching regression models. Within this class of models the author distinguishes switching regression models with sample separation known, sample separation unknown, and the case of imperfect separation information. The first part closes with the topic of switching simultaneous systems.

Disequilibrium models constitute the second main part of the survey. Starting with some well known examples of the literature different disequilibrium formulations are discussed such as the directional methods, the price adjustment equation and the so-called min-condition. Additional problems of specification in disequilibrium models follow. Special emphasis is given to the problem of serial correlation in the error terms and testing for distributional assumptions and disequilibrium. Extensions to multimarket disequilibrium models close this section.

Models with self-selection are discussed in section nine of the paper, and the last topic discusses multiple criteria for selectivity.

The paper is an excellent survey over switching regression models and the extended bibliography enables the reader to go easily into interesting details of this branch of econometrics.

Reviewer: H.S.Buscher