Association for Public Policy Analysis & Management

Countries of the Arab region consistently face significant inequalities in wealth and education, as well as low female and youth participation in the labour force. Susceptibility to food insecurity likely follows similar patterns with differences in vulnerability to food insecurity within and across countries, and by gender and age.  This study aimed to explore differentials in food insecurity experience and its socio-demographic determinants in the Arab region.

Methods

We used data from the 16 Arab countries surveyed by the Gallup World Poll in 2014 (n=17,085), and categorised countries as Low (LIC), Lower-Middle (LMIC), Upper-Middle (UMIC), and High income (HIC) according to World Bank definitions. Dichotomous variables of food insecurity and severe food insecurity - as measured by the 8-item Food Insecurity Experience Scale (FIES) - were generated using country-specific thresholds for FIES score. Prevalence of food insecurity and severe food insecurity was estimated on national and pooled datasets. Multivariate logistic regression was used to model the effects of age, gender, marital status, educational attainment, employment, household income, household size, household composition and assistance on food insecurity and severe food insecurity.

Results

The prevalence of food insecurity ranged from 14.5% in the HIC, to 75.0% in the LIC of the Arab region. Across all GDP groups, low household income, low respondent educational attainment, and unemployment were consistently significantly associated with higher odds of food insecurity (FI) and severe food insecurity (SFI) except in LIC where income was not associated with FI or SFI, and in HIC where unemployment was not associated with FI or SFI. Across low and middle income countries, the presence of under-5y old children in the household was significantly associated with FI, whereas larger household size predicted SFI in LMIC and was protective of FI in HIC and SFI in UMIC and HIC. Marital status was associated with higher FI and/or SFI in MICs and HICs, with higher odds in widowed or divorced respondents. Receiving assistance was associated with higher odds of SFI in LIC, and lower odds of FI and SFI in UMIC. Female respondents were more likely to experience SFI in LIC only, otherwise, no significant associations were found between gender or age and food insecurity in pooled models, although there was some variation across countries.

Conclusions

Large inequalities in food insecurity experience exist across countries of the Arab region. The determinants of food insecurity also vary by national income level. In addition to household income, and respondent’s education and employment, household composition is an important predictor of food insecurity in middle income countries of the Arab world, whereas marital status plays a stronger role in the middle to high income countries; likely a reflection of cultural factors in the latter.  Age and gender did not predict food insecurity in models that grouped countries by GDP categories, except in LIC, where women were at a disadvantage, but across country variations were observed hence justifying further research on gender and age differentials within those countries.