An efficient method of estimating the linear expenditure system (LES) by maximum likelihood is introduced. A test for insuring that parameter estimates obtained by numerical maximization satisfy the first-order conditions is given. In some experiments the new method of estimating the LES is compared to the more conventional method of estimation and is found to use significantly less computer time than the conventional method.
- Commodity demand
- Labor supply
- Linear expenditure system
- Maximum likelihood estimation
ASJC Scopus subject areas
- Statistics and Probability
- Statistics, Probability and Uncertainty