Effects of crop farming income on food insecurity as measured by per capita food expenditure are consistent across the models. Increase in crop farming and income will lead to increased per capita food expenditure hence will reduce vulnerability of population to food insecurity. Covariates Demographic characteristics Across all climate shocks, age of the head of the family is negatively correlated with income from crop farming and gardening. Female headed family is also negatively associated with income. Educational attainment of the head of the family generally has negative effect on income which indicates that lower educational attainment is associated with higher reduction in income. Age, sex and educational attainment of the head of the family with regard to food expenditure is positive and significant. A year added on age of the head of the family increases food expenditure by .006. In addition, as level of educational attainment increases, food expenditure also increases.. Assets Proxy variables of non-agricutural assets such as strong materials of walls and roof and access to safe water have different effect on income. Dwelling unit with strong walls and roof has positive effect on income, however families with access to safe water supply have negative effect. With regard to food expenditure, both have significant positive effect. In particular, there is about 5 percent increase in food expenditure among those families with strong materials of walls and roof as compared to those made from light/makeshift materials. Similarly, expenditure on food is higher by 6 percent among families with access to safe water. Family’s access to information as approximated by ownership of radio and TV significantly influences income from crop farming and gardening as well as food expenditure. Income of families with radio is 1315 percent higher than the income of families without radio. Families who own a TV have higher income of about 6-9 percent as compared to those families who do not own a TV. Both the ownership of radio and TV increase food expenditure by 2 percent and 9 percent, respectively. Ownership of car and motorcycle as proxies to family’s access to transportation is significantly associated with both income and food expenditure. Ownership of car tends to increase income by almost 51-54 percent and thus increases spending on food by 15-17 percent. Onwership of motorcycle also gives higher income by 16-20 percent therefore food spending increases by 6-7 percent. Infrastructure Included in the models are several variables that are proxies of communication infrastructure (ownership of telephone), financial infrastructure (cash loan payments) and social welfare infrastructure (expenditure on insurance and premiums). These variables, however, tend to have different effect on income. For instance, both ownership of telephone and payments to loan affect income positively while expenditure on social welfare affects income negatively. With respect to spending on food, both ownership of telephone and expenditure on social welfare are associated with higher spending on food while expenditure on loan decreases spending on food but not significant. CBMS-FAO | 33