POPULATION, GENDER RATIO OF

POPULATION, GENDER RATIO OF India's gender ratio (females per 1,000 males, or FMR)—933 in 2001—is strikingly low compared to both developed and developing countries (about 1,050 in Europe, 960 in North Africa, and 940 in China). Moreover, the FMR has declined monotonically since 1901 (from 972), reaching its lowest level in 1991 (927) with small improvements in 1981 (934) and 2001. This secular decline has prompted estimates of "missing women," based on standard census benchmarks. P. N. Mari Bhat (2002) estimates that the number of women "missing" in 1951, relative to 1901 age-specific FMRs, was 5 million; based on 1951 benchmarks, the number "missing" in 1991 was 9 million.

Though FMR declines in Indian states from 1901 to 1991 were diverse, they traced regional patterns that cut across state and district boundaries. Since antifemale bias in northern India is well documented, demographers initially focused on a north-south divide with a notional boundary at the Narmada River. In 2001, however, the Narmada boundary was breached. FMRs were low (below 925) not only in the north, but also in the west and center, and high (above 975) in the south and east.

The real causes of India's low and declining FMRs are hard to determine. Explanations have centered on spatial and demographic patterns, cultural phenomena, socioeconomic determinants, and the impact of poverty and growth.

Spatial and Demographic Patterns

State and regional patterns

Jean Drèze and Amartya Sen (1995) identified declines in state-specific FMRs as proximate causes of India's falling gender ratio, noting that six states (Bihar, Madhya Pradesh, Maharashtra, Orissa, Tamil Nadu, and Uttar Pradesh) were largely responsible. Bhat's regional analysis (2002) revealed that though FMRs declined across all regions during 1901–1951, the east (Assam, West Bengal) and center (Madhya Pradesh, Bihar, Orissa) led the fall; however, during 1951–1991, the central region clearly spearheaded the decline. The north (Haryana, Himachal Pradesh, Punjab, Rajasthan, Uttar Pradesh), with the lowest FMR in 1901, was still at the bottom of the ladder in 2001. Though FMRs in the south (Andhra Pradesh, Karnataka, Kerala, Tamil Nadu) declined throughout the last century, the region ended up in 2001 with the highest gender ratio. By 2001, FMRs for the center, east, and west (Maharashtra, Gujarat), which had started out at very different levels, had converged to a range of 920–935.

Three fallacies

An early explanation of low Indian FMRs, the underenumeration of females, buckled under scrutiny. Though occasional census undercounts (for censuses before 1931 and for 1971 and 1991) cannot be ruled out, a sustained and increasing underenumeration is implausible. Another explanation, the migration of males in search of economic opportunities, clearly cannot explain the secular decline in national FMRs. A third rationale, exceptionally high masculinity at birth in India, is spurious as Asok Mitra (1979) pointed out; India's sex ratio at birth has varied in the normal range of 104–107 male births per 100 female births (FMRs of 935–960) during 1901–1981.

Fertility

M. Das Gupta and Bhat (1997) demonstrated that in India where son-preference is strong, fertility decline itself could worsen child sex ratios. Even if son-preference falls with fertility decline, access to sex-abortion technology could exacerbate gender bias. However, fertility decline in India clearly reduced adult female mortality because childbearing declined.

Mortality

Pravin M. Visaria (1969) firmly established that India's low FMRs were due to sex differentials in mortality. Others suggested that a small sex differential in infant and childhood mortality persisting over a period might have caused India's high masculinity, but the decomposition of India's FMRs by Bhat and Drèze and Sen suggested greater complexity.

Adult FMRs

Drèze and Sen's analysis revealed that the overall FMR was driven by a sustained decline in the FMR of the thirty plus age group during 1901–1971. Bhat corroborated their conclusion for adults aged fifteen plus, surmising that the substantial survival advantage that women aged fifteen plus enjoyed over men in 1901 quickly eroded as mortality levels fell during the last century. Bhat also found that of the 5 million women estimated missing in 1951, all were aged fifteen and over. The central region had the sharpest decline in adult FMRs during 1951–1991, and contributed nearly half the women estimated missing in 1991 (using 1901 benchmarks).

Juvenile FMRs

Bhat found that the age structure of missing women changed dramatically after 1951: almost half the 9 million women missing in 1991 were children aged fourteen and under. This dramatic change justifies an in-depth focus on juvenile FMRs, but the definition of "juvenile" has varied.

Bhat argued that age misreporting could distort juvenile FMRs, because in societies with son preference, censuses tend to overestimate the age of males and underestimate female ages beginning from early childhood. Such misreporting exaggerates early childhood FMRs, like FMR aged 0–4, and lowers FMRs in later juvenile age groups; however, as age reporting improves with increased literacy and birth registration, FMR 0–4 falls, and older juvenile FMRs rise. Bhat claimed that more accurate age reporting was therefore partly responsible for the observed fall in FMR 0–4 and the rise in FMRs for children aged 5–9, 10–14, and 0–14 during 1901–1951. On the other hand, the subsequent fall in FMR 0–4 (38 points) during 1951–1991 and the significant associated declines in FMR 5–9, FMR 10–14, and FMR 0–14 (27 points) indicated that these declines in childhood FMRs were largely real.

Satish B. Agnihotri's classic analysis (2003) of recent juvenile gender ratios was built on his insights that gender gaps in mortality affect FMRs more strongly than mortality levels; gender differentials in infant mortality rates affect FMR 0–4 most acutely, while gender differentials in mortality in the 1–4 age group affect FMR 5–9 most severely. Thus, since FMR 0–4 and FMR 5–9 levels and gaps are good indicators of gender differentials in mortality, Agnihotri focused on these FMRs. In the 1981 Census, low values of FMR 5–9s and large gaps between FMR 0–4 and FMR 5–9 in numerous districts indicated significant excess female child mortality and confirmed a north-south divide; low FMR 0–4s (>910) in Punjab, Haryana, and western Uttar Pradesh suggested excess female infant mortality. The picture darkened in 1991. Districts with FMR 0–4 of less than 910 increased, pointing to the prevalence of female infanticide and (possibly) sex-selective abortions, while the startling rise in districts with FMR 0–4 of less than 960 suggested that increased girl mortality in the post neonatal and 1–4 age groups was "pushing further and deeper into the south."

The rural-urban divide and sex-selective abortions

Census and survey data confirmed that the rural FMR 0–14 fell continuously during 1951–1991, while the urban ratio fell precipitately after 1981. In 2001 the 18-point drop in the overall FMR 0–6 and the shrp plunge in the mean urban FMR 0–6 relative to the rural confirmed that the "northernization" of gender ratios was taking an urban route, thanks possibly to greater urban access to technologies for sex-selective abortion. Micro-research and S. Sudha and S. Irudaya Rajan's (2003) estimates of sex ratios at birth supported the hypothesis that the increased masculinity at birth over the last two decades was manmade, and confirmed that the pattern had extended beyond the north-northwest and had penetrated deeply into the four southern states.

Cultural Phenomena

Ethnicity

In 1961 FMRs of scheduled tribes (STs) and scheduled castes (SCs), which now constitute about 8 and 16 percent of India's population respectively, were higher than for the "rest of the population" (987, 957, and 934 respectively). Though FMRs of all groups declined during 1961–1991, the SC decline was so dramatic that the FMRs of the SCs and the "rest of the population" converged at about 922. Bhat found that almost two-fifths of the women missing between 1961–1991 were from SCs.

Proximity to Muslim influence is alleged to have lowered northwestern FMRs, but the notion is spurious: in both 1981 and 1991, Muslims were associated with less risk of excess female child mortality.

Kinship

Conventionally, the north-south divide in FMRs has been associated with different kinship systems. Northern women generally have low status associated with relatively high fertility and mortality, strong son preference, limited female property rights, low female labor participation, female seclusion, neglect of female children, and endogamous marriage; southern women enjoy stronger female agency. Agnihotri matched linguistic and kinship systems to arrive at a new binary classification of 1981 Census districts: a male-centered, Indo-Aryan "core" of 164 districts (primarily in the north and west, but including pockets in the south) with significantly low FMRs; and a "periphery" of the remaining 202 districts (including some female-friendly northern districts), combining Indo-Aryan, Dravidian, and Munda kinship and language systems, with high FMRs. Agnihotri found that in 1981 the "core" was associated with both excess female infant mortality and excess female mortality under the age of five.

Socioeconomic Determinants

Female labor participation (FLP)

Demographers have long recognized that work affects FMRs by enhancing women's worth and their access to basic household resources. In 1981 FLP was highest among STs, lower for SCs, and lowest for the rest of the population. Agnihotri found that the relationship between FLPs and FMR 5–9s was strong in the "core" in 1961 and 1981, both for SCs and the "rest of the population." However, Agnihotri found that kinship explained a larger part of the variance in FMR 5–9s than FLPs.

Multivariate analysis of other socioeconomic correlates of female disadvantage in child mortality revealed that the association between male literacy and female disadvantage was either positive or not significant, and that between female literacy and female disadvantage was negative but not consistently significant between 1981 and 1991.

Poverty and Growth

Both the literature and cross-section data indicate that poverty and gender inequality are negatively associated. If so, are economic growth and poverty reduction likely to induce greater gender inequality and lower the FMR over time, or is a U-shaped Kuznets curve likely (whereby the FMR initially decreases, bottoms out, and then rises as income increases)? Analyzing National Sample Survey Organization data (for 1987–1988, 1992–1993, and 1999–2000), Agnihotri found a negative association between FMRs and per capita expenditures for the total, adult, and juvenile populations, and urban and rural areas. The results therefore appear to negate the prosperity optimism of proponents of the Kuznets curve. Thus, despite five decades of growth, declining FMRs persist as a symptom of gender inequality and female deprivation, which Drèze and Sen characterize as "among India's most serious social failures."

Jayati Datta-Mitra

See alsoEconomic Development, Importance of Institutions in and Social Aspects of ; Poverty and Inequality

BIBLIOGRAPHY

Agnihotri, Satish B. Sex Ratio Patterns in the Indian Population: A Fresh Exploration. New Delhi: Sage Publications India, 2000. A classic work on the importance and determinants of disaggregated juvenile sex ratios using post-1961 census and survey data.

——. "Survival of the Girl Child: Tunnelling Out of the Chakravyuha." Economic and Political Weekly 38, no. 41 (11 October 2003): 4351–4360. An excellent exploration of the urban spread of the declining sex ratio for children aged 0–6 years during 1991–2001, and the possible role of sex-selective abortion.

Bhat, P. N. Mari. "On the Trail of Missing Indian Females." Economic and Political Weekly 37, no. 51, 37, no. 52 (December 2002): 5105–5118, 5244–5263. A solid exploration of the conundrum of India's falling gender ratio during 1901–1991, from a demographic, regional, and social perspective.

Das Gupta, M., and P. N. Mari Bhat. "Fertility Decline and Increased Manifestation of Sex Bias in India." Population Studies 51, no. 3 (1997): 307–315. A signal contribution to the literature on the impact of fertility decline on sex bias in India in the context of other Asian societies with strong son-preference, estimates of "missing" women between 1981 and 1991, and the emerging evidence on sex-selective abortion.

Drèze, Jean, and Amartya Sen. India: Economic Development and Social Opportunity. Delhi: Oxford University Press, 1995. Chapter 7 provides one of the best introductions to gender ratio issues in the broader context of gender inequality, socioeconomic indicators, and women's agency.

Mazumdar, Veena, and N. Krishnaji, eds. Enduring Conundrum: India's Sex Ratio: Essays in Honour of Asok Mitra. Delhi: Rainbow Publishers, 2001. Contains valuable reviews and analysis of the issues as well as a glimpse into the role played by the Committee on the Status of Women in India in insinuating gender concerns into public policy.

Mitra, Asok. Implications of the Sex Ratio in India's Population. New Delhi: Allied Publishers, 1979. Mitra, a pioneer in promoting women's studies in India, was one of the first to direct public attention to the country's declining gender ratio.

Sudha, S., and S. Irudaya Rajan. "Persistent Daughter Disadvantage: What Do Estimated Sex Ratios at Birth and Sex Ratios of Child Mortality Risk Reveal?" Economic and Political Weekly 38, no. 41 (11 October 2003): 4361–4369. The authors present an insightful perspective on increasing (manmade) masculinity at birth during 1981–1991, based on estimated sex-ratios at birth and multivariate analysis of socioeconomic and cultural variables.

Visaria, Pravin M. The Sex Ratio of the Population of India. Census of India 1961, vol. 1, monograph no. 10. New Delhi: Office of the Registrar General, 1969. Recognized by demographers as one of the first landmark studies in the field.