Since time immemorial, in various cultures, men have been considered superior to women. Sons carry on the family lineage and bear the responsibility of caring for their parents in old age. Daughters, on the other hand, move to their husband’s home after marriage, and are often financially dependent on them. Practices such as dowry, and laws that favor sons in inheritance, worsen the economic and social calculus of having female children. In the 1980s, modern technologies and declining fertility combined with this deep-rooted male preference, to make sex selection at birth widespread in the Asia Pacific region. The result was a skewing of birth rates, with the number of male births per 100 female births rising above the naturally expected level. For example, in India, in 1990, there were 110.89 newborn males for 100 newborn females1, as against a figure of 104.76 in the United States (UN World Population Prospects).
“Missing women”
The naturally expected sex ratio at birth is biased towards males by about 5% everywhere, that is, for 100 female births there are 105 male births. However, women are genetically and biologically equipped to survive better at all ages – as long as they receive similar care as men2. Assuming that developed countries have conditions that ensure similar care for women and men, the naturally expected sex ratio at birth, together with females being “hardier”, implies that these countries have more women than men at the population level.
In developing countries, the higher-than-usual sex ratio at birth begins redressing itself soon after, and the net outcome is expected to be a roughly equal proportion of women and men in the population (Coale 1991). Yet, as noted by Amartya Sen (1990, 1992), the female-to-male ratio in these countries is typically lower; the figure is in the range of 0.90-0.95.
Sen highlights the case of sub-Saharan Africa (SSA), which, in contrast, has a significant excess of women – attributable to their high participation in the labor force. For the lagging countries in Asia and north Africa, he demonstrates that if the female-to-male ratios of SSA were adopted – a region with comparable life expectancy and fertility – there would be an estimated 44 million “missing women” in China, 37 million in India and a total of over 100 million worldwide. To put it another way, over 100 million extra women would have been alive, had their circumstances been different.
Besides maternal mortality in the childbearing years, he pins this down to social inequality: “the comparative neglect of female health and nutrition, especially – but not exclusively – during childhood.” Sen contends that while gender attitudes are shaped by traditional cultures and values, it may be possible to influence these via public policies pertaining to women’s economic status, such as female education, the ability to earn, and inheritance rights. This is illustrated by the Kerala exception, a south Indian state with a female-to-male ratio similar to that in Europe and America, where school education and health systems are strong, women engage in the workforce, and many of them inherit property.
Not a monolith
In 2010, Anderson and Ray extended Sen’s work by looking at the distribution of missing women across all age groups, and by age-disease groups. For each category, they calculate a reference death rate for females, defined as the death rate of males in that country, rescaled by the relative death rates for males and females in the same category in developed countries. Subtracting this reference rate from the actual death rate of females and then multiplying the figure by the population of females in that category, gives the “missing women” for the category. A key difference between Sen’s early 1990s work and this study is that while the former focuses on the ‘stock’ of missing women at a point of time the latter examines annual ‘flows’ of missing women.
This analysis presents several notable results. First, although India and China have similar overall sex ratios, the age distribution of missing women varies. The majority of missing women in India and a significant chunk of missing women in China are of adult age. Second, as observed earlier, SSA had an excess of women in stock terms. However, when looking at annual flows of missing women as a proportion of the total female population, the figures for SSA in fact surpass India and China. Further, in younger ages, missing women are primarily an outcome of infectious, preventable diseases, with non-communicable diseases (for example, cardiovascular) being a major cause for older ages.
Up until this point, the explanation for missing women was being located in gender discrimination at birth and the mistreatment of young girls. Anderson and Ray’s research is crucial to the literature on gender and health in developing countries because it suggests that while these are certainly important factors, there is a lot more to the story.
Devil in the detail
As noted by Anderson and Ray (2010), the factors besides direct gender discrimination that can contribute to excess female mortality include biological, social, environmental, behavioral or economic. By unpacking missing women by age, disease and region, they put these diverse sources into one unified and comparative framework. “Injuries” as a cause of excess female deaths may indicate gender-based violence, but are cardiovascular deaths among women driven by genetics, lifestyle, or poorer medical care? They acknowledge that the accounting exercise cannot disentangle the role of the different drivers, but it is a first step towards such an exploration and can inform future studies on the issue. Indeed, effective policy would require an understanding of the pathways – a much broader research agenda.
Figure 1. Missing women across states in India, 2003
Source: Based on data from Table 4, Anderson and Ray (2012).
Notes: (i) The shaded areas represent the number of missing women (in ‘000s) across Indian states by age group, based on estimates from 2003. (ii) States marked with a checkered pattern indicate no available data.
So, through the following decade, Anderson and Ray continued to delve deeper into the data – focusing first on India and then Africa. In their 2012 article, they present a decomposition of India’s missing women by state (Figure 1). Based on the huge variations, the authors contend that no simple conclusions can be drawn to explain the country’s missing women phenomenon. Two north-western states of Punjab and Haryana have the highest missing women at birth, while southern states have lower excess female mortality at all ages. However, these regions add up to less than one-fourth of the total missing women; the bulk are spread across the other states and die in adulthood.
Since Anderson and Ray’s missing women formula involves a developed country reference, which may not be considered suitable for a low-income country such as India, they also repeat their computations with Kerala as the benchmark instead – the state with the lowest excess female mortality within the country. By doing so, the estimated numbers of missing women for the different states do reduce, but the key takeaways of their analysis still hold.
The readers are left with the following thought: “…the plight of adult women…is as serious a problem as that of young girls who were either never born or die prematurely in childhood.”
Moving on to Africa in their 2017 work, Anderson and Ray examine how the continent’s missing women are distributed across different regions and age brackets (0-14 year olds and 15-59 year olds), and which diseases are responsible. Similar to the India conclusion, the significant regional variation (Figure 2) makes it challenging to offer a one-dimensional explanation for missing women. Overall, there are 1.7 million excess female deaths in Africa annually – about three-fourths in the older category and the balance in the younger category. Spatially, West Africa is the worst-performing, followed by the East and the North. It turns out that almost all of Southern Africa’s excess female mortality comes from the older ages. Primary causes of deaths in younger ages are malaria, respiratory infections, and diarrhoeal diseases. For the older age category, the lead killer is HIV/AIDS, followed by maternal mortality.
Figure 2. Excess female mortality in Africa, 2011
Source: Based on data from Tables 6-10, Anderson and Ray (2017).
Notes: (i) The shaded areas represent the level of excess female mortality (in ‘000s) in each African country for the selected age group. (ii) Countries with a checkered pattern indicate no available data.
Partly, the regional distribution stems from what diseases are prominent in a given region. The excess female deaths caused by HIV/AIDS are concentrated in Southern and East Africa. Adult females in Central Africa are more impacted by tuberculosis relative to other regions. Nevertheless, the question remains: why are females more likely to die of these diseases than males? Is it biological gender bias in these diseases or that males are more likely to receive medical attention?
Until this point, research had focused mostly on excess female mortality in Asia, but the numbers produced by this study made it abundantly clear that the problem in Africa warrants attention as well.
Marriage and mortality
Starting from the mid-1800s until the beginning of the 21st century, studies in the developed world have consistently established that married individuals – both women and men – experience lower mortality rates than those that are unmarried. Similar is true for developing countries, although the data are sparser. This link between marriage and mortality is the starting point of Anderson and Ray’s 2019 paper.
They further argue that marriage at young ages is more or less universal in developing countries, and unmarried adults are typically widowed2. Hence, the excess mortality in these contexts may be attributed to widow(er)hood. Given that husbands are typically older than wives (by an average of about six years) and there is greater remarriage incidence among widowers as compared to widows, the number of widows is quite significant (Jensen 2005). Moreover, it is well-known that women pay a higher ‘price’ of losing their spouse in the form of reduced access to economic resources and social marginalisation.
On the basis of these extreme vulnerabilities of widows in developing countries, the researchers hypothesise that being unmarried can be a cause of excess mortality among adult women. How much of this excess mortality is due to non-marriage alone? There are 1.5 million missing adult women (30-60 year olds) each year, of which 35% can be attributed to non-marriage. Additionally, the phenomenon seems to be driven by young unmarried women of reproductive age.
Across regions, India is home to the largest proportion of missing adult women who are sans husband, followed by East Africa, while China has almost no unmarried missing women (Figure 3).
Figure 3. Excess mortality among women in Asia due to non-marriage, 2005
Source: Based on data from Table 1, Anderson and Ray (2019).
Notes: The graph presents the numbers of missing unmarried women in each age subcategory, in the year 2005.
Living, but with disease or disability
So far, the focus has been only on conditions that lead to death and the gender differences therein. However, those who are alive may also be suffering on account of disease(s) that they are living with. The Global Burden of Disease study has been measuring Disability Adjusted Life Years (DALY) since 1990, a metric that captures both mortality and morbidity. Conceptually, one DALY represents one lost year of healthy life because of either premature death, or disease or disability. It is calculated as the sum of Years of Life Lost (YLL) due to premature mortality and years of Life Lived with Disability (YLD).
A 2024 study in The Lancet by Patwardhan et al. seeks to compare DALY rates for females and males for the 20 leading causes of disease burden for individuals above 10 years of age, between 1990 and 2021. Globally, the researchers find that women bear a greater burden of morbidity-driven conditions, with the largest gap in DALYs being for low back pain, depressive disorders, and headache disorders (Figure 4). On the other hand, there are mortality-driven conditions where men exhibit higher DALYS, such as COVID-19, road injuries, and ischaemic heart disease. Essentially, for the diseases where women have higher DALY, the YLD component is dominant, and for diseases impinging on male DALYs more, the YLL constitutes the bulk of the burden. Further, the age-based analysis reveals that gendered health disparities begin early in the life cycle and widen over the years for several conditions.
Figure 4. Global absolute difference in DALY rates (per 100,000 population) between females and males, 2021
Source: Patwardhan et al. (2024).
Notes: (i) The causes of disease burden represented on the vertical axis are the top 20 causes of DALYs observed across females and males for the age group of 10 years and older globally in 2021. (ii) Health conditions are ranked based on the difference in age-standardised DALY rates (per 100,000 population) between females and males observed at the global level. (iii) The absolute differences between females and males were calculated as the DALY rate for females minus the rate for males for each specific cause and geography. A positive value indicates a higher rate for females than for males. The direction of the differences is also represented through colour coding.
Along the lines of the research discussed in earlier sections, this work too takes cognisance of the potential roles of both ‘sex’ (biological factors associated with physical and physiological traits) and ‘gender’ (socially constructed roles, behaviours and identities) in explaining the findings. Accordingly, there is a call for applying a sex and gender lens to data, and adoption of targeted and gender-sensitive approaches in health policy.
Throughout the literature summarized in this blog, one question has been consistently raised by the scholars over time: Do women access or receive medical care less than men? In the next blog, at the end of June, we will discuss a clutch of relatively recent studies that explore gender disparities in healthcare-seeking behaviors, as observed at the medical facility level.
FOOTNOTES
1 Prenatal sex determination was banned in India in 1994. However, the sex ratio at birth in the following years continuing to be above the naturally expected level, is indicative of the fact that the practice remained common – spurred by easily available ultrasound technology and the desire to have sons. Anecdotally as well, it is known that legal prohibition has not been able to put an end to sex-selective abortions.
2 Accordingly, in almost all settings across the world, developed or developing, women have higher life expectancy than men but the gap is greater in case of developed countries. In 1990, women’s life expectancy exceeded that of males by 7 years in the United States, 5 years in China, and 2.8 years in Nigeria.
3 While rising divorce rates and delayed marriages are increasingly visible in urban India, nationwide, widowhood remains the predominant reason for being unmarried among adult women. According to the Census of India 2011, which is the latest available, there were over 43.3 million widowed women, compared to just 2.37 million separated and 0.91 million divorced women. The number of never-married women aged 35-and-above stood at 4.17 million, suggesting that lifelong singlehood continues to be uncommon.
REFERENCES
Anderson, Siwan, and Debraj Ray. “Missing Women: Age and Disease.” The Review of Economic Studies 77, no. 4 (2010): 1262–1300.
Anderson, Siwan, and Debraj Ray. “The Age Distribution of Missing Women in India.” Economic and Political Weekly 47, no. 48 (2012): 87–95.
Anderson, Siwan, and Debraj Ray. Missing Women: Age and Disease. WIDER Working Paper 2017/116. Helsinki: UNU-WIDER, 2017. https://www.wider.unu.edu/sites/default/files/wp2017-116.pdf.
Anderson, Siwan, and Debraj Ray. “Missing Unmarried Women.” Journal of the European Economic Association 17, no. 5 (2019): 1585–1616.
Chen, Lincoln C., Emdadul Huq, and Stan D’Souza. “Sex Bias in the Family Allocation of Food and Health Care in Rural Bangladesh.” Population and Development Review 7, no. 1 (1981): 55–70. https://doi.org/10.2307/1972764.
Coale, Ansley J. “Excess Female Mortality and the Balance of the Sexes in the Population: An Estimate of the Number of ‘Missing Females.’” Population and Development Review 17, no. 3 (1991): 517–23. https://doi.org/10.2307/1971953.
Das Gupta, Monica. “Selective Discrimination Against Female Children in Rural Punjab, India.” Population and Development Review 13, no. 1 (1987): 77–100. https://pubmed.ncbi.nlm.nih.gov/6344225/.
GBD 2021 Disease and Injury Incidence and Prevalence Collaborators. “Differences across the lifespan between females and males in the top 20 causes of disease burden: a systematic analysis of the Global Burden of Disease Study 2021.” The Lancet Public Health 9, no. 7 (July 2024): e532–e545. https://doi.org/10.1016/S2468-2667(24)00053-7.
Jensen, Robert T. “Caste, Culture, and the Status and Well-Being of Widows in India.” In Analyses in the Economics of Aging, edited by David A. Wise, 357–374. Chicago: University of Chicago Press, 2005. https://www.nber.org/system/files/chapters/c10366/c10366.pdf.
Li, D. “Preference for Sons: Past and Present.” China Population Today 14, no. 5 (October 1997): 15–16.
Ritchie, Hannah. “Sex Ratio – The Male-to-Female Ratio at Birth.” Our World in Data, May 18, 2023. https://ourworldindata.org/gender-ratio.
Roser, Max, Hannah Ritchie, and Fiona Spooner. “Burden of Disease.” Our World in Data. Global Change Data Lab, February 2024. https://ourworldindata.org/burden-of-disease.
Sen, Amartya. “More Than 100 Million Women Are Missing.” The New York Review of Books, December 20, 1990. https://www.nybooks.com/articles/1990/12/20/more-than-100-million-women-are-missing/.
Sen, Amartya. “Missing Women: Social Inequality Outweighs Women’s Survival Advantage in Asia and North Africa.” BMJ: British Medical Journal 304, no. 6827 (March 7, 1992): 587–588. https://doi.org/10.1136/bmj.304.6827.587.
United Nations Population Fund (UNFPA) Asia and the Pacific Regional Office. Sex Imbalances at Birth: Current Trends, Consequences and Policy Implications. Bangkok: UNFPA APRO, 2020. https://asiapacific.unfpa.org/sites/default/files/pub-pdf/UF_APRO_GBSS_09-online.pdf.