|
Abstract:
|
National Sample Survey Organization (NSSO), India collects data on household monthly expenditure at a national level, throughout the year in "quinquennial rounds." We use the data on rural and urban households of West Bengal from 61st round NSSO data (July 2004- June 2005). Since income data are not taken in any of the NSSO rounds we focus our attention in analyzing how the total expenditure data fit the various income distribution forms. We also show how the Nelder Mead simplex algorithm can be used to get maximum likelihood estimator of complicated 4 or 5 parameter distributions. We use the five parameter Generalized Beta distribution and 11 of its special cases to fit to the given data and also calculate the expenditure inequality by the Gini index. It is concluded that GB distribution family gives quite good fit to modelling expenditure and urban households have more inequality in household expenditure as compared to rural households.
|
