Women's wellbeing and the sex ratio at birth: some suggestive evidence from India.(phenomenon of sex selective foeticide)

I. INTRODUCTION

The sex ratio of a population is a quantification of the weight of females in a population, and is employed in this article to denote either the ratio of females to males or the proportion of females in the population. It has been fairly widely recognised that the sex ratio at birth (SRB) is, in principle, a factor that contributes to both the level and trend of the overall sex ratio (SR). It is, however, only relatively recently--and specifically as an offshoot of observations in the Indian context--that serious attention would appear to have been paid to the possibility that a declining SRB might have contributed to the continuing decline in the overall SR of the Indian population. The favoured hypothesis--on which there seems to be some consensus (see Krishnaji [2000] for a review)--is that recent declines in the SRB are to be attributed to the phenomenon of sex-selective foeticide. In this article, we look at other possible explanations, and advance the view that insufficient weight has been attached to the possibility that a sex-neutral reduction in foetal wastage is a plausible candidate for explaining a decline in the SRB. Foetal wastage, in turn, depends on the wellbeing of the mother. We explore this hypothesis by looking at suggestive evidence thrown up by the Indian experience. The division of emphasis among alternative explanations for a declining SRB has important implications for assessing the role of developmental efforts in promoting women's wellbeing and for how the latter could impact on the SR, and therefore for how to interpret secular trends in the SR of a population.

This article is organised as follows. In section II we briefly review such evidence for India as there is on temporal variations in the SRB. In section III we present three alternative logical hypotheses for a declining SRB, the plausibility of each of which is briefly discussed. Section IV undertakes a more detailed exploration, in the Indian context, of the hypothesis that seeks to explain a declining SRB in terms of a genderunbiased reduction in foetal wastage, with specific reference to those factors which might be expected to affect women's wellbeing and, through that route, the extent of pregnancy losses. Section V concludes, with a general discussion on development, women's wellbeing, and the sex ratio of a population.

II. THE SRB IN INDIA: A BRIEF REVIEW OF ORDERS OF MAGNITUDE

It must be stated at the outset that it is hard to find reliable long-term published data on the sex ratio at birth in India. An account of trends in the SRB over the first half of the twentieth century is to be found in Visaria [1971], who employed data from the Civil Registration System (CRS). The quality of these data, as Visaria has pointed out, is affected by under-registration of (most particularly, female) births, though there is little evidence to suggest any accelerated rate of under-registration over time. Visaria's computation of the average SRB for each of the five decades between 1891-1900 and 1931-39, supplemented by the average for the period 1940-46, yielded a trend of decline over the first half of the century, with the number of female births per 1000 male births dropping from 936 in 1901-10 to 912 in 1940-46 (Table 1). This tallies with Bhat's [2002: 525960] assessment: '... the vital registration data for British India had shown a steady downward trend in the female-male ratio at birth in the first half of the last century' (though he goes on to add: 'But it is not clear whether this suggested a real trend, or deterioration in registration completeness'). A decline in the SRB from toward the end of the last third of the century, based on Sample Registration System (or SRS) data, is also noted by Bhat (op. cit.): 'The SRS too has shown a declining trend in the FMR [female to male ratio] at birth since its inception in the late 1960s'. More recent data, covering roughly the last two decades of the twentieth century, are available on the Indian Census website: Figure 1 is a reproduction of a Census graphic, based on SRS estimates, and reveals a renewed decline in the SRB from the early 1980s to the mid-1990s, with an upturn toward the end of the 1990s, but to levels which are still lower than those obtaining in the 1980s.

The trend in the national SRB discussed above can be supplemented by some estimates available at specific regional levels. Bhat [2002], employing estimates provided by Dyson, presents a time series on the SRB for the territory of Berar (which includes four districts of present-day Maharshtra in western India) over the period 1881-1940: he concludes that the SRB '... began to decline after the first decade of the twentieth century'(p.5255). A time series on the SRB is also available for the southern State of Tamil Nadu. The Census of India 1961, employing CRS estimates, provides information on the numbers of female and male births for Madras (the erstwhile Tamil Nadu), from 1921 to 1961, and this series can be updated from CRS-based data in the Annual Statistical Abstract of Tamil Nadu: the resulting profile is summarised in Table 2, which reveals a fairly systematic decline in the SRB (interpreted as the number of female births per 1000 male births), from 957 in 1921-30 to 916 in 1981-90.

The general trend of decline noted above for India--at both national and selected regional levels--provokes the question: what are some of the factors that could account for variations over time in the SRB of a population? This question is addressed in the following section.