Fertility Differentials by Ethnicity in Rural Tripura, India

Prasanta K. Majumdar, Indian Statistical Institute

Fertility differentials by ethnicity and other cultural factors have been observed in both developed and developing countries. Several studies around the world have found notable ethnic differences in fertility behaviour as well as in fertility performance. It has been confirmed from two National Family Health surveys of India that the tribal fertility has always been higher than non-tribal fertility. The population of Tripur is heterogeneous. Two major ethnic group lived in Tripura: Bengalees and the Tribal population. During 1874-1991, the total population of Tripura has increased by about 37 times : tribal population by about 18 times and non-tribal population by more than 70 times. The influx of Bengali Hindus from East Pakistan may be responsible for the differential variation in growth rate among tribals and non-tribals. The international migration used to play an inportant role upto 1891 in the growth of Tripura population. However, since 1961 this international migration is negligible. Therefore, the future growth and composition of population in Tripura will depend mainly on fertility. Thus differential in ethnic fertility may be examined in the light of four main hypotheses : characteristics, particularised theology, minority status and interaction. Therefore, the objective of this paper is to assess the trends in fertility among two ethnic groups : Tribals and Non-tribals by examining their cumulative marital fertility. About 1350 currently married females within reproductive ages were interviewed from a sample of 1500 rural households selected through a multistage sampling procedure. This paper examined cumulative marital fertility by ethnicity. First, ethnic differentials in mean children ever born by age of women and by marital duration of women are examined. In order to ascertain whether the fertility differentials are attributable to characteristics or to ethnicity, there is a need to control socio-economic variables. To estimate net differences by ethnicity, the technique of multiple classification analysis (MCA) has been adopted. Dependent variable for this analysis is children ever form (cumulative fertility) and the independent variables are : education of husband or education of eligible women, ownership of agricultural land, family Income per annum, occupation of eligible women, experience in child mortality, etc. Since children everborn (cumulative fertility) is affected by age at marriage (cohabitation) and age at survey date, these two variables are used as co-variate in the MCA analysis. In the study area, the currently married women within ages 15-49 had on an average 2.9 children ever born (CEB) and increases from 0.48 CEB at age 15-19 to 4.90 CEB at ages 45-49. However, the average varies by ethnicity being higher for tribes (3.04) than non-tribes (2.82) after standardisation of age of eligible women, the mean number of CEB for tribal women is 3.34 and for non-tribal women is 2.83. similarly after standardising marital duration of womne. CEB for tribal women is 3.36 as against non-tribal women whose CEB is 2.81. It has also been found that deviations in mean children everborn by ethnicity are large. Tribals showing fertility well above the average the remaining variables education of eligible women, ownership of agricultural land, occupation of women show significant effects. The multiple classification analysis results clearly show that differentials in marital fertility by ethnicity persists even when controlling for socio-economic and demographic variables, clearly the characteristic hypotheses does not gain much support atleast in Tripura. Instead, specific factors associated with ethnicity appear to be crucial in fertility differentials.