Nuptial Determinants Of Fertility In Western Kenya

Abstract

This research on nuptial determinants of fertility investigates the re l at ions hip between selected nuptial variables and total fertility rates in western Kenya. Its basic objectives are as follows: to investigate the level and direction of correlation between selected nuptial variables and total fertility rate; to demonstrate that socio-economic factors influence fertiity net only directly but also indirectly through their impact on nuptial variables; and to identify appropriate policy measures which could be employed to control fertility in the study region. The null hypotheses tested were: (i) nuptial variables are less important than socio-economic variables as predictor variables of current total fertility rate, (ii) nuptial variables are less important than socio-economic variables as predictor variables of lifetime total fertility rate. The research utilizing both the1969 and1979 census data reinforces its analysis with the sample survey data of 1978. The sample survey data was collected by a two-stage stratified random sampling method. Approximately 411 household units were surveyed in Western Province, and another 854 household units were surveyed in Nyanza Province , These household units yielded probability population sample sizes of p = 0.8% and p = 0.4% of the total populations in the two provinces respectively. The research findings based on a step-wise partial correlation analysis led to the rejection of the first null hypothesis. However. on the strength of 8 - coefficients, the nuptial variables which emerged as relatively good positive predictor variables of current total fertility rate (Yl) were: the proportion of total adult females aged 15-49 years, (Kl); and the proportion of currently married women aged 15-49 years, (K3). On the other hand, the nuptial variables which emerged as relatively good negative predictor variables of current total fertility rate (Yl) were also: the proportion of women widowed aged lS-49 years, (K4); and the proportion of divorced and separated women aged lS-49 years. (KS). The effect of all these nuptial variables on current total fertility rate (Yl), were found to be statistically significant from zero at 0.05 level of significance. The research findings for the lifetime total fertility rate (Y2), again led to the acceptance of the second null hypothesis, because both socio-economic and situational variables examined dominated the rank of significant predictor variables. The only few nuptial variables which emerged as good predictor variables of Y2 were: age at first marriage, (K7]; the proportion of currently married women aged lS-49 years, (K3); the proportion of widowed women aged 15-49 years, (K4); and the proportion of divorced and separated women aged 15-49 years, (K5). All these coefficients were found to be statistically significant from zero at 0.05 level of significance. When socia-economic and situational variables were incorporated in the analysis, those found to be good predictor variables of current total fertility rate, (Yl). were: Hie proportion of Temal es of lower primary education who we re aged 15 .-49 years, (X3); tile,proportion of women of upper primary education who were aged 15 - 49 year~, (X4); and, the proportion of females aged 15 - 49 years, (X8), who were using contraceptives. All these coefficients were also statistically significant from zero at 0.05 level of significance. Further reference to lifetime total fertility rate (Y2), indicates the following socio-economic and situational variables as important negative predictor variables: the proportion of illiterate women aged 15 - 49 years, (X2}; the,proportion of females of upper primary education who were aged 15 - 49 years; the proportion urbanized (X?); and, infant and childhood mortality rate, (X9). It was noted-that the proportion of females u,sing contraceptives, (X8) also, exerted negative effect on Y2. It shauld , however, be emphasized that most of these negative effects on lifetime total fertility rate, (Y2), were not statistically significant from zero at 0.05 level of significance. The main conclusion is. therefore, that in the context of the'present and future levels of socia-economic development of Western Kenya, socia-economic and situational phenomena will continue to be relatively better predictor variables of total fertility rate than nuptial variables. Therefore, it is recommended that further research should investigate the impact of accelerated rural agricultural change on family formation and dissolution with the aim of finding out an acceptable new legal age at first marriage which should be enforced for the entire nation. Furthermore the following anti-natalist policies are suggested: age at first marriage policy; population education policy; employment and income policy for female population to make women selfreliant; health care policy that focuses on reduction of infant mortality rates; land adjudication policy to reduce land fragmentation; and a new approach to the current family planning policy.