Assessment of trends and determinants of household food insecurity in Marsabit county-Kenya

Abstract

Food insecurity is a growing concern worldwide despite the government strategies to implement measures. However, a lot of studies conducted so far in the field gave more emphasis to the high potential counties of Kenya. Better understanding of major determinants and trends of household food insecurity is important to design appropriate interventions. In Marsabit County, food insecurity has been exacerbated by multiple hazards such drought, COVID-19 pandemic, conflict and desert locust invasion. A lot of studies have been conducted in the field of food insecurity but such assessments do not verify situations in Marsabit and hide the true determinants, trends and prevalence of food insecurity. The purpose of the study was to assess trends and determinants of household food insecurity in Marsabit County. The study was guided by the following objectives: to assess temporal variation of food insecurity in the period 2017-2020 in Marsabit County, to estimate the prevalence of food insecurity in Marsabit County and to identify the key determinants household food insecurity in Marsabit County. The study adopted cross sectional descriptive research design where primary data was gathered at the household level and triangulated with secondary data from National Drought Management Authority. In this research, monthly sentinel data collected by NDMA in the period 2017-2020 was used to understand the temporal variation/trends of food insecurity in the area. Furthermore, primary data collection was implemented to understand the current determinants and prevalence of household food insecurity. Purposive sampling was used to select the pre sampled wards and thereafter simple random sampling applied to select households from the pre sampled wards. A total of 322 households were interviewed across the pre sampled wards from a target sample size 399 households which represented 81% response rate. Primary data was collected through structured questionnaires. Factors of household food insecurity were livestock owned, livestock prices, educational level, milk production, milk consumption, main sources of income, water sources, and household and livestock water distances. Five food security outcome indicators were assessed: Reduced Coping Strategies Index; Household Hunger Scale; Food Consumption Score; the Household Dietary Diversity Scale and Food Insecurity Experience Scale. For objective 1; Time series plots were applied on food security outcomes to determine temporal variation of food insecurity in the period 2017-2020 in Marsabit County. For objective 2; proportion of the food insecurity outcomes was computed so to estimate the prevalence of food insecurity and in objective 3, ordinal logistic regression models adopted to identify determinants of household food insecurity. Overall prevalence of household food insecurity was moderate. Spikes of food insecurity were witnessed in a good year of 2020. Although the mean food consumption score fell outside the normal ranges, households had minimal food consumption gaps and adopted coping mechanisms were less frequent and severe. Household hunger was at moderate levels during the study period. Tertiary training, university degree, sale of livestock, sale of crops and water sources such as natural ponds and shallow wells were significant and thus showed a relationship with food consumption score. University educational level of household head, sale of livestock, trade and sources of water for households such as water pans, piped water, traditional river wells and seasonal rivers were significant thus showed an association with the predicted variable coping strategy index.