Understanding the Determinants of Food Security among Rural Farming Households in Rwanda

ABSTRACT

Despite Rwanda's impressive economic success over the last decade, household food insecurity persists, particularly in rural areas. Empirical evidence is therefore deemed necessary to inform policies addressing food insecurity, hunger and poverty in general. This study examines household food security and its determinants in the rural areas of Southern and Northern provinces of Rwanda. Using cross-sectional data collected from 534 farming households, we estimated household food security status by Food Consumption Score (FCS) and Food Insecurity Experience Scale (FIES) and applied ordered Probit model to identify drivers of household food security. The results indicate that a significant proportion (74.4%) of households had an acceptable FCS, but 36.52% of the farming households experienced moderate and severe food insecurity when assessed with FIES. Ownership of livestock was identified as a key determinant of food security using both metrics, which affirms the importance of combating hunger and food insecurity in Rwanda through their ‘one cow per household policy’ and recommends that this policy be strengthened.

KEYWORDS:

• Food consumption score
• Food security
• Ordered Probit
• Rwanda

Notes

1. For details on FCS calculations, readers are advised to consult documents like WFP Food Consumption Analysis (2008). Link: https://documents.wfp.org/stellent/groups/public/documents/manual_guide_proced/wfp197216.pdf.

2. Construction of FCS comes with some limitations, including the likelihood of missing out some key food items and recipes, non-classification of appropriate food groups and ingredients for mixed dishes. The study attempted to minimized the likelihood of these limitations occurring by the following means: (1) focus group discussions, (2) key informant interviews, (3) recruitment and training of local enumerators and pretesting questionnaires. Appropriate information obtained from the above-mentioned ways were used to amend our questionnaire to suit the local context.

3. For further explanation on the questions and calculations on FIES, readers are advised to consult FAO (2018). The State of Food Security and Nutrition in the World, Building Climate Resilience for Food Security and Nutrition.

4. The study also conducted preliminary analysis of whether the ordered logit or probit fits the data well using the Akaike Information Criterion (AIC) and the Bayesian Information Criterion (BIC). The ordered logit has AIC and BIC values as: 944.79 and 1043.24, respectively. The ordered Probit has AIC and BIC values as follows: 943.09 and 1041.55, respectively. Since the ordered Probit has lower AIC and BIC values than the ordered logit, the ordered Probit fits the data better.