Fertility in Rural Sudan: The Effect of Landholding and Child Mortality

Abstract

Two stage least squares and descriptive statistics are used to examine the impact of land holdings on fertility and child mortality in rural Sudan. Data are based on surveys of 523 households in 26 villages which involve tenant farming according to public Elsuki schemes. Independent variables are womans age husbands years of schooling immigrant status and presence of county health facilities. The results indicate that womans age and the immigration dummy are significant and positively associated with child mortality. The regional health dummy is slightly significant and negatively associated with child mortality. About 50% of the variation in fertility is explained by maternal age cultivated land uncultivated land immigration dummy child deaths and the mortality rate. Husbands years of education does not appear to have an effect on fertility. The strongest effects on fertility are maternal age and child mortality. Cultivated land is highly significantly and positively associated with fertility. Uncultivated land has a small effect. Immigrant households have lower fertility than nonimmigrant households which is attributed to the possibility of higher child mortality among immigrant households. Olsens correction method is used to correct the bias in ordinary least square estimates of replacement response of fertility to child mortality which reveals a replacement rate of 0.56-0.73. Reduction of child mortality from 1.54 to 0.77 would reduce fertility from 7.16 to 6.59 and surviving children would increase from 5.62 to 5.82. An increase in cultivated land by 50% would increase fertility from 7.16 to 7.7. The quantitative impact of womens education on fertility is not examined due to data deficiencies. Models are based on the household production model and a demand function.