NONLINEAR ERRORS-IN-VARIABLES ESTIMATION OF SOME ENGEL CURVES

The most common solution to the errors in variables problem for the li near regression model is the use of instrumental variable estimation. However, this methodology cannot be applied in the nonlinear regressio n framework. In this paper we develop consistent estimators for nonlin ear regression specifications when errors in variables are present. We apply our methodology to estimation of Engel curves on household data . First, we find that the 'Lesser-Working' specification of budget sha res regressed on the log of income or expenditure should be generalize d to higher-order terms in log income. Also, we find that errors in va riables in either reported income or expenditure should be accounted f or. Lastly, and perhaps most interesting, we find rather strong suppor t for the Gorman rank restriction on the matrix of coefficients for th e polynomial terms in income.