Population change and food availability in Tanzania; a multilevel modeling approach: the case of Ukerewe Islands

Thumbnail Image



Journal Title

Journal ISSN

Volume Title


University of Dar es salaam


Primary data from a survey of 150- rural households in Ukcrewe district, one of the densely populated rural districts in Tanzania mainland, were used to examine the relationship between socio-demographic variables and household food availability. The study employed the Principal Component Analysis (PCA) in the construction of households’ possession indices which was then used as the proxy determinant of food availability. The 2-level random intercept model with 3-ordered categorical response variable was used in the estimation of coefficient of effects. The findings show that household seasonal migrants, education attainment by household members, farm land resources and life cycle of household at different stage of development that was measured by the age of household head, were significant. It was further noted that seasonal migrants and farm land were likely to reduce the odds ratio of a household of being extremely poor or poor. An increase of a person with secondary or post-secondary education in a household reduced the odds ratio of a household of being extremely poor by 52.8 percent as compared to 19 percent odds ratio reduction by increase of non­ educated household member. The findings show that the young aged head of household was 3 and 5 times less likely to be extremely poor as compared to old and middle age headed household respectively. The data could not allow any viable conclusion to be drawn from variables such as; household sex ratio, sex of household head, size of the household and proportion of household permanent migrants because they were statistically insignificant


The Degree of Master of Arts (Statistics)


Population growth, Population change, Food availability, Multilevel modeling approach, Ukerewe Islands