Specification and estimation of a simultaneous-equation model with limited dependent variables
- Creators
- Nelson, Forrest D.
- Olson, Lawrence
Abstract
Single equation models with limited dependent variables have received considerable attention in the econometrics literature, but little is known about the specification and estimation of simultaneous-equation models in which some or all endogenous variables are limited. The one relevant published paper by Amemiya [1974] considers a model in which all endogenous variables are truncated at zero, reveals certain necessary restrictions on the model, and suggests a method of estimation. But his model is not applicable to all situations and the merits of his estimation procedure are confined to consistency and tractability. Here we propose an alternative specification which may be more appropriate than Amemiya's for many problems and discuss a relatively simple limited-information estimation procedure.
Additional Information
© 1978 Economics Department of the University of Pennsylvania. Manuscript received February 7, 1977; revised July 18, 1977. We would like to thank Halbert White, James Heckman, Lung-fei Lee, Kenneth Gaver, David Grether, and Donald O'Hara for helpful comments. A mini grant from the NBER Computer Research Center covered a portion of the computation costs. Olson's participation in this research was supported by the U.S. Department of Labor Dissertation Grant #91-36-75-04. Opinion and conclusions do not represent policy of the labor department. Formerly SSWP 149Additional details
- Eprint ID
- 83672
- DOI
- 10.2307/2526334
- Resolver ID
- CaltechAUTHORS:20171204-152231284
- National Bureau of Economic Research
- 91-36-75-04
- Department of Labor
- Created
-
2018-01-02Created from EPrint's datestamp field
- Updated
-
2021-11-15Created from EPrint's last_modified field