Determinants of the food insecurity at household level in Pakistan: A multilevel model approach

Food insecurity is a major concern for the developing world and around 37% of the population of Pakistan is food insecure. This paper utilizes the Food Insecurity Experience Scale (FIES) to assess the population prevalence of food insecurity and to identify their risk factors and determinants at the household level in Pakistan. This study employs a multi-level random coefficient model, using the Pakistan Panel Household Survey (PPHS-2010) dataset; representative data from 4,130 households. Factors like; income of the household, gender, education, household size, land ownership, and shocks of food insecurity allow the incidence of idiosyncratic shocks (injuries and/or casualties) at the community level, which affects the food insecurity situation of the community, rather differently were included. The study confirms a statistically significant inverse relationship between household income, household size, and household head education with food insecurity and a positive association of shocks and inflation with food insecurity at the household level. Specifically, with increasing per capita income of the household, food insecurity declines (coefficient: -0.083, statistically significant at 1%) and food insecurity increases with shocks (coefficient: 0.058, statistically significant at 1% significance level). The study also reveals a significant heterogeneity at a one percent significance level in the determinants of food insecurity at the district, community, and household levels. The income of the household, household head gender and education level, household size, household assets, shocks, injuries, and inflationary pressure are important determinants of food insecurity in Pakistan.