Description:
ABSTRACTA general type of Engel function which can be called the Box‐Cox Engel function is developed. This function is used to estimate the relationship between the dependent and explanatory variables for which many commonly used Engel functions become special cases. The Maximum Likelihood (ML) method is used to estimate its parameters using grouped Household Expenditure Survey (HES) data. These estimated parameters are then used to estimate Engel elasticities for total food and various non‐food items. The Box‐Cox Engel function is also used to discriminate among the commonly used Engel functions, by performing a parametric test on the power parameters. A joint test on the different power transformation has been done by the likelihood ratio test and it turns out that none of the commonly used Engel functions are appropriate for the Australian HES 1975‐76 data, although the double log Engel function is not significantly different from the Box‐Cox Engel function for five out of ten consumption items.