Thesis (M.Sc.) - Cairo University - Faculty of Economics and Political Science - Department of Statistics

This study attempts to present the use of the Fractional Multinomial Logit model (FMLM) to model the household budget allocation using the data set of The Household Income, Expenditure and Consumption Survey )HIECS) for two rounds (2008/2009, and 2012/2013 ( . The HIECS is conducted by Central Agency for Public Mobilization and Statistics (CAPMAS). Household budget allocation reflects how the budget of a household is allocated to buy different commodities or services. It is presented by the monthly expenditure shares on the basic items: (Food and drink - Smoking - Education expenses - Health and medications - Transportation {u2013} Housing - Culture - Restaurants and Hotels - Communications- Cloths {u2013} Furniture - Others ) . In this study, the main goal is studying the determinants of how families allocate their budget. Two models are used: FLM and FMLM, the first one is used to study the determinants of food share only, while FMLM is used to study the determinants of all the different items in the budget,simultaneously. The results indicate that before the revolution of 25 January 2011 the main variables that affect household budget allocation are place of residence, percentage of highly educated members, percentage of worker, monthly income (on the logarithm scale) and status of the household according to the poverty line as these variables were significant on all items of monthly expenditure (household budget allocation). While after the revolution of 25 January, the main variables that affect household budget allocation are percentage of members with disability or chronic illness and monthly income (on the logarithm scale). As these variables have positive significant on all items of monthly expenditure (household budget allocation)

Issued also as CD