| Nonparametric Hypotheses Testing of Optimizing Behavior (2002) | |||||||||||||||
Abstract | |||||||||||||||
| This paper proposes and applies a practical method for including a stochastic component in the traditional nonparametric tests of economic hypotheses developed by Afriat and Varian. The method proposed here requires no a priori knowledge of the variance covariance structure of the stochastic components within each observation as in the test proposed by Epstein and Yarchew. In this work nonparametric regression is used to get estimates of the price response functions under the null hypothesis. The information obtained from these regressions is then used in combination with a minimum distance statistic to test the null hypothesis. | |||||||||||||||
Publication details | |||||||||||||||
| |||||||||||||||