Socio-demographic and economic determinants of household expenditure on eating out in India - Evidence from India Human Development Survey (IHDS) II.

'Eating out (EO)' has witnessed a significant increase across nations and in India. Understanding food behaviour from an economic standpoint is essential to recognize food systems and nutrition transition. Our work compared the macro-data on per capita income and gross state domestic product from India with household expenditure on eating out. Using the India Human Development Survey II micro-data we studied the socio-demographic and economic determinants of household expenditure on eating out. The study design is a secondary data analysis of India Human Development Survey II from a sample of  = 42,152 households. It included socio-demographic and economic predictors, such as religion, caste, residence, migration for work, education, number of household members, income source, total annual income, consumption expenditure and poverty status. No consistent trends emerged in the macroeconomic comparisons using gross state domestic product and per capita income. Results of logistic regression showed that the reserved caste categories OBC [COR 1.2 (1.1-1.2)] and SC/ST/Others [COR 1.1 (1-1.1)], household heads education up to secondary [COR 0.4 (0.4-0.5)] and higher secondary and above [COR 0.7 (0.6-0.8)], urban place of residence [COR 0.55 (0.53-0.58)], organised sector [COR 0.7 (0.6-0.7)] as the income source, higher annual income [COR 0.52 (0.50-0.54)], higher annual consumption expenditure [COR 0.44 (0.42-0.46)] and above poverty line [COR 2.1 (1.9-2.2)] were significantly positively associated with expenditure on eating out ( = <0.05). All these variables stayed significant in the multinomial regression model as well. This analysis confirms the increase in household income along with other demographics, increases the frequency of eating out. However, such trends remained inconsistent in macroeconomic assessments using gross state domestic product and per capita income.