The topic of food insecurity has been more widely studied in the developed nations58. This study documents the prevalence and correlates of food insecurity in a developing country, Nepal, using recent, nationally representative data. In this section, some of the key findings will be discussed.
First, ethnicity was an important determinant of food insecurity. Food insecurity was common among almost all ethnic groups. Strikingly, however, 76% of Dalit women were in food insecure households. The odds of a Dalit woman of childbearing age experiencing HFI were significantly higher than for women of nearly all other ethnic groups even after accounting for other relevant factors. These results align with other studies that have documented that HFI is substantially higher among Dalits in Nepal 38,39. They also suggest that social exclusion plays a role in food insecurity. Due to generations of caste-based discrimination, Dalits in Nepal have very low access to economic opportunities--education, employment, property ownership, and economic institutions. They are often concentrated in rural areas serving as landless agricultural laborers with high malnutrition among women and children 39. Studies from India also suggest that food insecurity and malnutrition are particularly acute among Dalit women in that country59,60. Some have suggested reviving the agriculture of indigenous food crops—sorghum, pulses, vegetables and animal sources of food—and increasing the consumption of these products among Indian Dalit mothers to improve their nutritional status 59,60. In Nepal, social policy has been directed toward reducing disparities between Dalits and other groups. Since 1997, the government has funded programs and activities aimed at improving the quality of life of Dalits. These initiatives include scholarship programs for secondary and higher education of Dalit children, income generation activities of Dalit men and women, and mass communication programs to raise public awareness on caste discrimination; these programs, however, are often poorly funded and implemented 39.
One option for social policy intervention may be to expand Nepal’s income transfer policies to specifically benefit the most food insecure populations. Low-income countries around the world, including Nepal, have been developing and expanding income transfer policies. For example, Nepal has been slowly building its Social Security Program since 1994/95 and has now instituted a universal old age (70+), disabled, and widows (60+) pension plan that transfers a set amount of monthly income to eligible elderly, disabled and widowed persons 61. As Nepal prepares to tackle food insecurity among women of reproductive age, children, and minorities, perhaps benefit policies could specifically target Dalit women and their children. Such programs could be piloted in a district with a high concentration of food insecurity and proportionally large Dalit population. For example, according to the 2016 NDHS, the population of Baitadi District of the Far-Western Development Region was 41% Dalit, while nationally Dalits are about 12% to 14% of the population. Also, nearly 90% of the women from Baitadi had experienced food insecurity in the past 12 months.
Second, consistent with previous studies 13,27, education is a protective factor for food security for women of reproductive age in Nepal. Only 35% of all women with education beyond 10th grade were food insecure. Among those with no formal education, 68% were food insecure. One possible path to increasing food security among women, then, is to increase the enrollment of girls in school and retain them at least until they complete high school or 12th grade of education. Early investment in the enrollment of girls, retention of these girls in school, and their advancement into the next grade level will likely improve women’s education and subsequently reduce HFI. In recent years, Nepal has made impressive efforts to increase girls’ enrollment in school. The “Girl Summit” of 2016 was committed to supporting the education of girls and boys by improving the school and community environments 62. Nepal’s neighboring countries—India and Bangladesh—have launched financial incentive programs to increase demand for the enrollment and retention of girls in schools 63,64. Similar programs may increase the enrollment of female children in Nepal as well.
It is possible that educational attainment is a proxy for some other factor, such as household economic resources not captured in the current list of variables or strength of the social network. This analysis does not identify intermediate factors and their contributions to the reduction of food insecurity. Even if education is only indirectly related to food security, increasing educational achievement is still an important intervention strategy, as it should improve these intermediate outcomes. Future research in the context of countries such as Nepal might aim to clarify causal pathways.
Third, as expected, household wealth was a protective factor for food security. Policies could be designed to bolster the economic security of households with no or limited wealth. For example, in recent years, several developing countries in Africa have tested Unconditional Cash Transfer (UCT) programs. These initiatives make a targeted transfer without any behavioral requirements to reduce poverty and hunger immediately65,66. Across sub-Saharan Africa alone, there are now over 123 UCT programs 67. Studies assessing the impact of UCT find these programs improve dietary diversity and food security 68. A study from Zambia comparing the impact of two government-run poverty alleviation programs using cluster randomized controlled trials found that UCT increased household per capita consumption expenditures by 20% and reduced food insecurity significantly 69. In Burkina Faso, an evaluation of a UCT program, again using a cluster-randomized controlled trial, found a significant increase in the dietary intake of high-nutritional-value foods in young children between 14 and 27 months of age 70.
Nearly all evaluation studies have concluded that UCTs hold promise in reducing poverty and food insecurity. The primary arguments against such programs focus on their fiscal viability71. Nepal could test the concept of UCT in one of the high food insecure districts of the Far- or Mid-Western region (discussed in more detail subsequently) with particular attention to fiscal feasibility and cost-effectiveness. Also, there are many local non-governmental organizations and external developmental partners (EDPs)-- international governmental and non-governmental organizations-- working in Nepal to improve health, education, and agriculture sectors 72-74. Perhaps, some of the EDPs could be specifically directed to work in food insecure districts and test new ideas such as the UCT.
Fourth, geography is a predictor of food insecurity in Nepal. In our study, food insecurity is most pronounced in the Mid-Western development region compared to the Eastern Development region, a finding consistent with previous research using 2011 NDHS data 40. Out of Nepal’s 75 districts, six of the ten highest food insecure districts are in the Mid-Western Development Region. In these districts, food insecurity ranged from 83% to 100% of women of child-bearing age. These districts include: Kalikot (83%), Rolpa (89%), Dailekh (86%), Dolpa (94%), Jumla (91%) and Humla (100%). The remaining four districts are distributed in the Far-Western Development region (Baitadi, 90%), Central Development region (Rasuwa, 87% and Ramechhap, 85%) and Eastern Development region (Khotang (90%). As mentioned previously, these districts could serve as test cases for a UCT program.
Fifth, one surprising result in the current study is the lack of a statistically significant relationship between the gender of the household head and food insecurity in the multivariable model. One possibility is that women’s vital contribution as food producers may have buffered this relationship 75. Our interest in women of childbearing age is partly motivated by previous findings that female-headed households are more susceptible to food insecurity41. In contrast, there was no statistically significant difference in food insecurity by the sex of household head. About 31% of women lived in a female-headed household. These household heads could be grandmothers, widows, divorced women, or married women whose husbands were not at home. One or more of these subtypes of households may be more susceptible to food insecurity. In turn, omitted factors could explain the differences between our study and previous research. Future studies using qualitative data might be able to better describe the food insecurity experiences of women who head households, further explaining the discrepancy.
Lastly, this study has strengths and limitations. A strength of this study is that it uses nationally representative data with very few missing cases, so the results are generalizable to the population. It also incorporates a more extensive food security measure than earlier studies of food security in Nepal. For the first time, the 2016 NDHS employed the full nine-item Household Food Insecurity Access Scale. The 2011 NDHS used only seven of the nine items. This has implications for monitoring progress toward attaining the Sustainable Development Goals (SDGs). In September 2015, the United Nations and its 193 member countries adopted 2030 Agenda for Sustainable Development, which includes 17 SDGs to be attained by 2030 76. Goal 2 aims to eliminate hunger globally. Specifically, SDG 2.1 seeks to end hunger and ensure access to safe, nutritious, and sufficient food all year round for all people. SDG 2.2 intends to end all forms of malnutrition, stunting and wasting in children under five years of age, and ensure the nutritional needs of adolescent girls, pregnant and lactating mothers and older adults. This study focuses on a subpopulation—women of childbearing age—for whom food insecurity, because of the subsequent consequences for children, has broader implications. This analysis showed that ethnicity is associated with HFI even after accounting for some economic, social, and geographic factors among these women. As we move toward the 2030 Agenda, these findings provide baseline data to monitor progress toward the elimination of food insecurity among women of reproductive age in Nepal and offer potential vectors for intervention.
While using a strong data source, a limitation of this analysis is that it is cross-sectional and reflects correlational relationships only. Additional research, whether qualitative or using more advanced quantitative methods, is needed to make persuasive causal claims. The nine-item HFIAS employed in the 2016 NDHS to assess household food insecurity has received mixed evaluations. A review of nine studies from India has questioned the reliability of four items corresponding to anxiety (e.g., ‘worried’) about food and quality of food (e.g., ‘preferred food’, ‘limited variety’) 77. Sethi and colleagues (2017) imply that the response to these items varies by culture, threatening validity. The current study used the entire scale to define food insecurity (see table 1). If the criticisms of the scale are accurate, at least two issues arise. First, this study may have overestimated the actual level of food insecurity in the overall population of women of child-bearing age. If the differential interpretation of items is culturally patterned, the current estimates of disparities by ethnic group may be systematically biased. Second, the reference period of “past 12 months” for the nine items assessing food insecurity is a concern. Such a long timeframe increases the risk of recall bias, and also restricts us from examining the known seasonality of food insecurity in Nepal.