Gender Ratio

Draft version.

Gender ratio – the balance between the number of males and females – is a measure which can have significant impact on social structure, demography and how this changes over time. It both shapes, and is shaped by, cultural and economic context.

Here we provide an overview of gender ratio across the world, how it has changed through time; how it changes from birth to late life; and how gender ratios have been influenced by biology, cultural values, and technology.

In this article we use the terms 'gender' and 'sex' – often in similar context. There is significant discussion about differences in definition between these terms, such that many agree they are not interchangeable.1 Here we use both terms, 'gender ratio' because it's typically the most well-known and searched term on the topic, and 'sex ratio' because this is arguably more accurate and used within the academic literature.

Gender Ratio across the world

How much of the population is male and female?

The sex ratio – how much of the population is male and female – varies across the world.

There are several reasons why the sex ratio of populations vary and are rarely equal:

  • differences in mortality rates and life expectancy for women and men. Women, on average, live longer than men. This means that all else being equal, we would expect females to account for slightly more than half of the total population.
  • sex ratios at birth are not equal. In all countries, there are more male than female births (the extent of which varies by country). This means that all else being equal, we would expect males to account for slightly more than half of the total population.
  • for countries or regions, migration patterns can affect the sex ratio of the population. If in some countries there is a significant amount of imported male-dominant labour, all else being equal, we would expect males to account for more than half of the total population.

The magnitude and balance of these factors determines the sex ratio of the total population.

In the chart below we see the sex ratio of populations: this is shown as the percentage of the total population which is female. Countries over 50 percent (shown in blue) have more females than males; those below 50 percent (shown in red) have fewer females.

Click to open interactive version

Here we see a varied sex ratio across the world. Most countries have a female share of the population between 49 and 51 percent (within one percentage point of parity).

There are however a few notable outliers:

  • across several countries in South and East Asia – most notably India and China – there are significantly fewer females than males. These are countries where there are large differences in sex ratio at birth.
  • in several countries across the Middle East there are many more males than females: in Oman, women are outnumbered 3-to-1; in the United Arab Emirates (UAE) it's almost 4-to-1. If we look at how this has changed over time, we see the share of the population that is female has declined from a ratio close to parity decades ago. The primary reason for this is a large male migrant stock: in 2017, 45 percent of the Oman and 88 percent of the UAE population were from immigration.2 The UN reports that only 16 and 25 percent of international migrants to Oman and UAE, respectively, were female.
  • there are significantly more females than males in Eastern Europe. Populations in Eastern Europe have some of the largest gaps in life expectancy between men and women: in Russia, for example, the average life expectancy at birth was only 65 years compared to 76 years for women.

Gender Ratio at birth and through the life course

Sex ratio at birth

Across the world there are differences in the sex ratio at different life stages. This imbalance in the male and female population can in some cases be traced back to birth: in some countries the number of boys and girls born each year is significantly skewed.

In the chart below we see the differences in sex ratio at birth across the world. Here the sex ratio is measured as the number of males births for every 100 female births; a value greater than 100 indicates there are more boys than girls born each year. A figure of 110 would indicate that there are 110 male births for every 100 female births.

The first striking point is that in every single country of the world there are more boys born than girls. This has been true as far back as 1962, as shown in the timeline below.

Does this mean every country selects for boys prior to birth; for example, through induced abortion practices which preferentially select for boys? Not necessarily.

For most countries, there are around 105 males per 100 female births. This is what the World Health Organization (WHO) define as the natural 'sex ratio at birth'. As we see in the next section, some male-bias in births is what we would expect with no deliberate gender selection. There are, however, some key outliers: in countries including China, India, Vietnam, Pakistan and Azerbaijan this ratio is too skewed to be explained without deliberate selection practices.

Click to open interactive version

Why are births in all countries male-biased?

In the absence of selective abortion practices, births of a given population are typically male-biased – the chances of having a boy are very slightly higher than having a girl.

To explain this we have to understand:

  • if there are differences in the sex ratio of conception;
  • if there are sex differences in the probability of miscarriage.

In the most comprehensive study of its kind, Orzack et al. (2015) monitored the trajectory of sex ratio from conception through to birth across five different methods.3 This produced the largest dataset available on the sex ratio throughout the stages of pregnancy.

A key result from this study was that the sex ratio at conception is equal: there is no difference in the number of males and females conceived. For births to be consistently male-biased, there must be gender differences in the probability of miscarriage through pregnancy.

Results found that the probability of miscarriage varies between genders depending on the stage of pregnancy:

  • there is a higher probability that an 'abnormal' embryo is male – in the first week of pregnancy, excess male mortality therefore means pregnancy is female-biased;
  • in the next 10-15 weeks of pregnancy female mortality is higher, which increases the ratio in favour of males;
  • male and female mortality is approximately equal around week 20;
  • between weeks 28-35 of pregnancy, there is higher male mortality.

Although the probability of miscarriage varies between genders across the course of a pregnancy, female mortality is slightly higher than male mortality over the full period. This would explain why the birth ratio in all countries is male-biased.

How does sex ratio vary with birth order?

Most countries have a sex ratio at birth which is around the expected range of 105 boys born for every 100 girls. There are exceptions to this: there are countries – most notably in Asia – with highly skewed sex ratios in favour of males.

The preference in some countries for a son is seen in overall sex ratio at birth (SRB) figures. But this comes through even more strongly when we look at how this ratio is affected by the birth order of children.

In the visualization below we look at the case of India and how sex ratios change from the 1st child in a family through to 2nd, 3rd, 4th and 5th born children. This is adapted from its Economic Survey 2018 and based on data from the DHS and National Family Health Surveys in India.4 It presents this data in two cases:

  • when the child is not the last child born (i.e. the parents have additional children afterwards) – this is shown in the charts on the left;
  • when the child is the last child born in the family (i.e. the parents stop having children afterwards) – this is shown on the right.

Let's first focus on the top row, which presents the data for India as a whole. On the left-hand side we have the sex ratio at birth when the child is not the last. For the 1st children, the sex ratio is very close to what we would expect 'naturally': a ratio of around 105 boys per 100 girls. But we see that for the 2nd, 3rd, 4th and 5th born children, this ratio is skewed towards girls. Remember again that this is the case when it's not the last child i.e. the parents go on to have another. In other words, when a girl is born, parents are more likely to have another child. There is a clear bias towards continuing to have children until a family gets a son.

Compare this to the right-hand side where we see the sex ratio when the child is the last. These ratios are much more skewed towards boys. Parents whose 1st child is a son are much more likely to stop having children (182 boys per 100 girls). This is consistent across the birth order: whether it's the 2nd, 3rd, 4th or 5th child, a family is much more likely to stop having children when they have a boy.

Combined we see a strong preference for a son in India: parents are more likely to continue having more children when the child is girl, then are more likely to stop having children once they have a boy. For some, the goal of having children is to gain a son but not a daughter.

Now compare these ratios to that of Indonesia (in the second row below) which does not show a strong preference for a son or daughter. Sex ratios here do not diverge strongly from the expected ratio of 105, and there is no clear difference when the child is or isn't the last to be born.

In the third row we also see that there are large differences in son preference within India. It shows that there are states with 'weaker' and 'stronger' preference for a son which affects the extent to which the sex ratio at birth is skewed.

[Clicking on the visualization below will open it in higher-resolution].Sex ratio at birth by birth order

Sex selection practices also become more prominent for later births

From the study presented above we see that, using India as an example, the sex of children in some countries can be an important deciding factor for when parents stop having children.

But birth order also influences the likelihood of prenatal sex selection (PSS) i.e. sex-selective abortion. We see evidence of this across several countries.

Researchers looked at Indian national survey data from 1976 to 2005 to see how the sex of children and birth order affects the use of prenatal selection.5 They found a significant and increasing skew in the sex ratio since the 1980s, and mainly as a result of male selection for 2nd or 3rd order births within a family. There was a large increase in male births in 2nd or 3rd children when the firstborn or the first and second-born siblings were female. This skewed ratio can only be explained from prenatal sex selection in favour of boys.

At a local level, a study of a large Delhi hospital known for maternal care showed very similar results.6 The overall sex ratio was male-biased with only 806 girls per 1000 boys. But this got significantly worse when the family already had a daughter: 720 girls per 1000 boys if there was one previous girl and only 178 per 1000 for two previous daughters. Even for women who had not practiced sex-selection abortion, more mothers who had previously had a girl reported taking traditional medicines (which were ineffective) for sex selection purposes.

These examples are of course not restricted to India. Many countries across Asia in particular have similar findings. In the chart below we see how the sex ratio in South Korea was affected by birth order. Here we see a very steep rise in the sex ratio of third-, fourth- and later children through the 1980s. By the early 1990s there were more than 200 boys per 100 girls for third-born children; for fourth-born or higher, the ratio was close to 250:100. This has since fallen back to close to natural levels over the past few decades.

You can compare this data for China using the "change country" button in the bottom-left of the chart. Clearly sex ratios in China have also been affected by birth order.7 Here we see that since 1980 there has been some rise in the male bias for first-born children (rising from the expected 105 boys to 114 boys per 100 girls). But we see a much more significant skew in the ratio for second or third-born children. For third-born children, the ratio was 158 boys per 100 girls, suggesting a high prevalence of sex selection abortions.

South Korea provides an important example where the male bias can be successfully addressed. But for several countries across Asia this looks like a major challenge: the data shows many parents are still desperate for a son.

Click to open interactive version

Sex ratio in childhood

Sex ratios – the ratio of males and females – at birth is male-biased across every country in the world; in our section on this, we explain why biologically we'd expect this to be the case. The so-called 'expected' or 'ideal' sex ratio at birth was around 105 males per 100 females. But how does this ratio look in childhood? Does it change from birth to five years old?

In the charts below we see two perspectives: firstly a global map of the sex ratio at five years old. Just as with the sex ratio at birth, we see clear outliers across Asia where the ratio is much more male-biased than we would expect. In China, there is close to 117 boys per 100 girls at age five; in India, more than 111 boys per 100 girls.

Secondly we see a scatterplot comparison of the sex ratio at birth (on the y-axis) versus the ratio at five years old (on the x-axis). The grey line here represents parity: a country which lies along this line has the same ratio at five years old as it does for birth. As we see, most countries lie above this line: this means the sex ratio at birth is higher than in childhood.

In other words, the male-bias tends to weaken through the first years of childhood. Why is this the case? As we explore in the next section of this entry: across most countries infant and child mortality rates are higher for boys than for girls. This means fewer boys make it through the first few years of life. For most countries this results in a decline in the sex ratio from birth.

Overall we see that despite higher child mortality in boys, the sex ratio at age five in the majority of countries is over 100: this means boys still outnumber girls in childhood.

Click to open interactive version
Click to open interactive version

Why do boys die more often than girls?

From life expectancy to mental health; substance use to cancer rates; there are important differences in health outcomes between the sexes. In a related post, Esteban looks at the evidence to answer the question "Why do women live longer than men?"

Here we focus on the youngest, asking, why do young boys die more often than girls?

Child and infant mortality is higher for boys across nearly all countries

In the chart below we see the comparison of child mortality by sex: child mortality measures the share of newborns who die before reaching their 5th birthday.

Here, the mortality rate for boys is shown on the y-axis, and the mortality rate for girls on the x-axis. The grey line running diagonally across the chart marks where the mortality rate for both sexes is equal. In countries which lie above the grey line, the death rate for boys is higher than for girls.

What's striking is that with exception of two – India and Tonga – child mortality is more common for boys in all countries of the world. This is also true for infant mortality, which is the share of newborns who die before reaching their 1st birthday. We look in more detail at why India and Tonga are outliers here.

It has been known for a long time that the mortality of boys is higher. As early as 1786 the physician, Dr Joseph Clarke, read a paper to the Royal Society of London on his observations that "mortality of males exceeds that of females in almost all stages of life, and particularly the earliest stages".8 You can compare child mortality rates for girls and boys over the past few decades here.

Why is this the case? First of all, it's important to understand what young children die from.

Click to open interactive version

Boys are more vulnerable to birth complications and infectious diseases

Let's focus for now on infants in the first year of their life. What do they die from?

In the chart below we see global death rates in infants across different causes in 2017. This data comes from the IHME's Global Burden of Disease study, which provides estimates by sex – on the y-axis I have plotted death rates in boys, and on the x-axis for girls. Just like the charts above, causes which lie above the grey line are more common in boys.

The chart shows that for all major causes of death, mortality is higher in boys. This includes early neonatal complications such as preterm births, asphyxia, birth defects and heart anomalies. But it also includes infectious diseases such as syphilis, malaria, respiratory infections, tetanus and diarrheal diseases. This is true for a broad range of infections, spanning person-to-person, vector-borne, blood-borne, and food and water borne diseases.9 The sex differences in the causes of infant deaths were already documented almost a century ago: in an impressive paper published in 1929, Bawkin explores the mortality sex ratio of specific diseases from countries across the world.10

There are some causes, HIV/AIDS, nutritional deficiencies, whooping cough, among others, for which the death rates are higher in girls. But overall infant boys are more vulnerable to infections and death than girls.11,12,13

Click to open interactive version

Boys tend to be born more prematurely, and have weaker immune systems

There are a number of factors that contribute to the higher mortality of boys:

  • Boys are more likely to be born premature: the rate of births before the end of the full pregnancy term is higher in boys.14 And, since boys tend to have a higher birthweight than girls, in some cases the rate of induced preterm births is higher (due to perceived risk of waiting to term).15 The chart above showed that more boys die from preterm births. The fact that preterm births are more common for boys contributes to this.
  • Boys have relative immaturity at birth: on average, boys are 'less mature' at birth; this means they are at higher risk of having delayed physiological function (such as lung function) and adverse neurological outcomes.16,17 For example, poorer lung function in newborn boys has been shown for both term and preterm infants.18,19 Studies have suggested that even in the case of normal pregnancies, on average lung maturation in boys lags that of girls by approximately one week.20 Differences in maturity at birth are not restricted to lung function.21 The reason for this difference has been an important question for decades – the answer is still not clear. There are however some leading hypotheses: surfactant production for lung function has been observed earlier in female fetuses, leading to improved airway flow in the lungs; estrogen has been shown to affect lung development positively in females; males, on average, have a higher birthweight meaning they may trade-off increased size for functional development; and the uterus may be less hospitable to male fetuses with the introduction of a Y chromosome in females creating and immunoreactive response to the central nervous system.22,23 This, combined with a higher risk of premature birth may explain why boys have higher rates of asphyxia, respiratory infections and birth defects.
  • Y-chromosome in boys increases vulnerability: biologically, males and females are differentiated by chromosomes: females have two X chromosomes (XX) and males one X and one Y chromosome (XY). Having two X chromosomes means that the newborn has a stronger immune system because X chromosomes contain a larger number of immune-related genes.24,25 This means that, since females have two X chromosomes, they have a stronger immune response; boys on the other hand, are more vulnerable to X-linked infections.26 Males are therefore more vulnerable to many infectious diseases. The stronger immune response of females comes with a cost however. It is the reason why women are more susceptible to autoimmune disorders such as HIV/AIDS.27 Boys are also more susceptible to specific genetic diseases where the defective genes are carried on either the X or Y chromosomes; this is because boys have only one X chromosome so a single recessive gene on that X chromosome results in the disease.
  • Sex hormones affect immune response: the different role of sex hormones is another reason that the immune response is stronger in females. Males have much higher amounts of testosterone which seem to inhibit two major parts of the immune system – B and T-lymphocytes.28 Estrogen, on the other hand, acts as an effective regulator of this. Overall, male hormones weaken the immune system relative to females.

The leading explanations for the "male disadvantage" lie in the biological differences between the sexes. More specifically, differences in maturity, sex chromosomes and hormones.

In circumstances where both sexes are treated fairly, we would therefore expect infant and child mortality rates to be slightly higher for boys. Despite these differences, it's true that over the past half-century in particular, child mortality has been falling rapidly across the world. This has been true for boys and girls alike.

Sex ratio through adulthood

The sex ratio – the number of males relative to females – at birth and in childhood are male-biased in almost every country.

In previous discussion we covered why we'd expected this to be the case biologically.

But how does this change throughout adulthood?

In the chart below we see the sex ratio at different ages through adolescence and adulthood starting at age 15, age 20 then each decade afterwards up to 100 years old. The chart shows the sex ratio which is measured as the number of males per 100 females.

In adolescence and young adulthood we see both the persistence of male-bias in birth ratios, and the large impact populous countries such as China and India – with very skewed sex ratios – have on the global average. At age 15 and 20, males outnumber females by 106 to 100. But as we move through adulthood we see this ratio steadily decline. By age 50, the ratio is 1-to-1; by 70 years old, there are only 89 males per 100 females; and in the very oldest (100 years old) there are only 25 males per 100 females.

You can explore this data for any country of region using the "change country" button in the bottom-left of the chart. You will see that for some countries this decline in the sex ratio with age is even more extreme: in Russia, for example, by age 50 there are only 88 males per 100 females; by 70 years old there are almost twice as many women as men.

In every country in the world women tend to live longer than men. Whilst this is true today, it hasn't always been the case. In his text 'Why do women live longer than men?' Esteban takes a look at the evidence and explanations for why this is the case.

Click to open interactive version

Why do women live longer than men?

On average, women live longer than men – this is true for every country in the world. This fact plays an important role in how the sex ratio changes with age through adulthood.

A detailed discussion of why this is the case is available in our explainer on 'Why do women live longer than men?'.

Missing girls and women

Biology or discrimination: which countries have skewed sex ratios at birth?

Today, and at several points historically, the sex ratio at birth in some countries is too skewed to be explained from biological explanation alone. The expected (sometimes called the 'ideal') sex ratio at birth is around 105 males per 100 females.

In a recent study by Chao et. al (2019) re-modelled sex ratios at birth (SRB) across the world based on a range of population sources, including census and household survey data.29 The results are shown in the chart below since 1950. The authors identified 12 countries with strong statistical evidence of a skewed sex ratio: Albania, Armenia, Azerbaijan, China, Georgia, Hong Kong, India, Montenegro, South Korea, Taiwan, Tunisia, and Vietnam.30

Most of these countries are in Asia. Why is this the case? Is there a biological or environmental difference, or is it the result of social discrimination?

Click to open interactive version

Hepatitis B was proposed then later debunked

Amartya Sen was the first scholars to publicly bring attention to the concept of "missing women" as a result of practices (sex selective abortion, unequal treatment and infanticide of girls) in favour of males.31

The reason for this skew in sex ratio has been previously challenged. One of the leading hypotheses was put forward by economist Emily Oster. In a 2005 paper she argued that a large proportion – approximately 45%, around 75% in China, 20-50 percent in Egypt and western Asia, and under 20% in India, Bangladesh, Pakistan, and Nepal – of the overrepresentation of men could be explained by the high rates of hepatitis B carriers in Asia.32 The rationale was that (1) hepatitis B carriers have an offspring sex ratio much more in favour of boys (1.5 boys per girl), and (2) that hepatitis B carrier rates were notably higher in Asian cultures than in the West. The combination of these two observations Oster calculated could account for a large proportion of the "missing women".

A few years later, Oster debunked her own hypothesis on the role of hepatitis B in a paper titled "Hepatitis B does not explain male-biased sex ratios in China". In a study of 67,000 people in China – 15% of which were heptatis B carriers – Oster and colleagues found no link between hepatitis B status and offspring sex ratio.33 The hepatitis carrier status of neither the mother or the father had an effect on the offspring sex: carriers were not more likely to have a boy than non-carriers. The authors concluded that hepatitis B rates could not explain the skewed sex ratio in China.

Other studies – such as that by Lin and Luoh (2008) in Taiwan – have also found minimal to no effect of hepatitis B on the sex ratio.34,35

Sex-selective abortions and discrimination against girls

After the hepatitis B hypothesis was debunked, no clear evidence of a biological factor in such skewed sex ratios has emerged. There is some variability in the "natural" sex ratio which may result from biological or environmental factors – a figure of 105 boys per 100 girls is usually adopted, but often this can vary from 103 to 107 boys per 100 girls. But skewed ratios to the extent we see in some countries cannot be explained through this alone.

The evidence for sex-selective abortion and discrimination against girls is now strong across some countries. Not only does the increase in sex ratios coincide with the availability of prenatal sex determination technologies, there is also clear evidence from studies investigating the use and promotion of such methods.

In India, for example, prenatal diagnosis (PD) became available in the 1970s, shortly after legalization of medical abortion in 1971.36 Whilst its use was intended for detection of abnormalities, it was soon used and promoted by the Indian medical profession for sex determination.37 Even after its use for sex determination was prohibited, studies suggest many gynaecologists did not consider sex-selection abortion as unethical: that it was an important intervention to balance population control with the desire for sons.38

Results from some of the earliest studies on abortions following the availability of prenatal sex determination are striking. Over the year from 1976 to 1977, at an urban hospital in India, 96% of the girls who were tested were aborted; in contrast, all of the 250 boys tested – including those with an identified risk of genetic defect – were born.39 At a clinic in Mumbai, all of the 15,914 abortions after sex determination in 1984/85 were girls. Results from another six hospitals in the city found 7,999 of the 8,000 aborted fetuses in 1988 were girls.40

The evidence that highly skewed sex ratios at birth have been largely the result of gender discrimination and selective abortions has been well-established across several countries. We discuss the reasons for this discrimination here.

Infanticide and excess female mortality

Sex discrimination can occur prenatally (in the form of sex-selective abortions, as we discuss here) or postnatally in the form of excess mortality. Excess mortality can come in several forms and at different life stages – such as the deliberate killing of infants (infanticide) or premature mortality from neglect, poor or unequal treatment.

Over time, prenatal discrimination has increased as both abortions and sex determination technologies have become more readily available. Nonetheless, postnatal discrimination still occurs and has a long history.

Infanticide has a long history

The prevalence of infanticide – the deliberate killing of newborns and infants – has been acknowledged in recent decades. But the history of the practice across the world dates back much further.41,42 Historically, infanticide was prevalent for both boys and girls.43

Over the past century – the prevalence of infanticide – more specifically female infanticide – in countries with a skewed sex ratio has received much more attention.44 Even today, cases on infanticide still exist, despite being outlawed in most countries.45

Infanticide is the most obvious case of postnatal sex selection. More often overlooked is the excess mortality which results from neglect and unequal treatment through childhood, but also later in life. Son preference has had notable impacts on the sex ratio across several countries.

Poor treatment of girls results in increased mortality in childhood

In almost every country young boys are more likely to die in childhood than girls – as we explore here, there are several biological reasons for is. But this is not true in a few counties – India is one notable example today. There, girls die more often than boys.

When we compare infant (under one year old) and child (under-5s) mortality rates between boys and girls in India we see that the difference is bigger for the older age group. While infant mortality rates are approximately the same, the child mortality rate for girls is higher.

Let's then focus on children rather than infants. In the chart here we see mortality rates for boys (on the y-axis) and girls (on the x-axis) for various causes in India. This data is shown for children aged 1-4 years old in 2017. Here we see that for many, death rates are significantly higher for girls. Some of these – hepatitis, measles or tuberculosis, for example – we expect to be higher for girls. But not for infections, respiratory and diarrheal diseases. Note also the much higher mortality rates for nutritional deficiencies and protein-energy malnutrition for girls.

Click to open interactive version

Poorer health outcomes for girls across some countries – often in Asia and not restricted to India – has been well-documented.46,47 Even for some countries where the child mortality rate remains higher in boys than for the girls, death rates for girls are still higher than would be expected.

Social preference for a boy has resulted in unequal treatment of young girls in a number of ways. Studies have shown in some countries:

  • poorer nutrition for girls and unequal food distribution;48
  • less breastfeeding from mothers for daughters than for sons;49
  • lower healthcare utilization for girls;50,51
  • during pregnancy there is evidence of preferential treatment for boys with more antenatal visits and increased tetanus vaccinations.52

This combination of poorer nutrition and healthcare investment can result in higher mortality rates for girls than would be expected in childhood, but also to excess mortality for women in later stages of life.

How many missing women are there?

The term 'missing women' was first coined by Indian economist Amartya Sen in a 1990 article; he estimated "more than 100 million women are missing".53

'Missing women' refers to the shortfall of women from the number we would expect in the absence of sex discrimination. It is the sum of women who are missing at birth (as a result of sex-selective prenatal practices such as abortion) and excess female mortality (either through infanticide, child neglect or maltreatment later in life).

How many 'missing women' are there?

How many women are missing?

Many researchers have tried to calculate the number of missing women. Using sex ratios at birth, and at different ages we can compare the observed and expected values: the difference is then defined as girls and women who are missing.54

In the table below we provide a summary of a range of estimates – note here that the year of the estimate is different for each. Here we see that although there is significant uncertainty in these estimates, all all within the range of over 50 million by 1990, and upwards likely upwards of 100 million today.

SourceYear estimate is forNumber of missing women
Sen (1990, 1992)1990More than 100 million
Coale (1991)199060 million
Klasen (1994)199092 million
Klasen and Wink (2002)1990s102 million
Hudson & Den Boer (2004)2000s90 million
Guilmoto (2012)2010117 million
Bongaarts & Guilmoto (2015)2015136 million

In the chart below we see some of the most recent estimates of missing women from 1970 through to 2015 from Bongaarts & Guilmoto (2015).55 In 1970 the researchers estimated there were 61 million missing women; by 2015, 136 million.In 1970 the researchers estimated there were 61 million missing women; by 2015, 136 million. This is more than the population of Mexico. We see that more than 80% are missing from China and India – 68 and 45 million, respectively.

The authors also provided projections of the number of missing women through demographic changes to 2050 which you find here.

Click to open interactive version

How many girls and women are missing every year?

Above are estimates of the total number of missing women in the world. But how many are 'missing' each year?

In the chart below we see estimates of (1) the number of 'missing' female births from prenatal sex discrimination; and (2) excess female deaths as a result of postnatal discrimination.56

In 2015 it's estimated there was an additional 3.3 million missing – approximately half of which were missing births and half from excess female mortality. Looking at trends since the 1970s, the rise of sex-selective abortions (shown as missing births) becomes clear. Until 1980, nearly all missing women were the result of excess female mortality; by 2015 as many were from missing births each year.

Click to open interactive version

Reasons and consequences of a skewed gender ratio

Why is there preference for a son?

In countries where there is a clear imbalance in sex ratio, there's nearly always preference for a boy. But why does this preference exist?

Son preference is most common in countries across East and South Asia, in addition to some in the Middle East and North Africa.57 Although there are significant cultural, economic and societal differences between these countries there are important parallels which explain strong preference for a boy.

What these countries tend to have in common is a strong logic of 'patrilineality': the logic that productive assets move through the male line within the family.58 Although many other countries also have patrilineality to some degree, in countries with a strong preference for a son, it's much more rigid. The social order of families resides with the males: lineage is passed from father to son. Men within the social order are the fixed points, and women the moving points: when a daughter marries, she leaves the current family to join a new one.

This can produce perceived economic and social benefits to having a son rather than a daughter, including:

  • Family name: the lineage within a family stays within the male line. Having a son therefore continues the family name.
  • Old age support: sons within a family are often responsible for supporting parents in illness and old age.59 Elderly parents often live with married children and overwhelming with sons. This provides a strong economic incentive for a son for old age support. It's a key differentiating factor between countries with and without strong son preference: in Taiwan, it's rare for parents to live with a married daughter, but in the Philippines (without strong son preference) parents are equally likely to live with the son or daughter.60
  • Dowries: a dowry – the transfer of property or money from the bride's to the groom's family – is an important economic concern for having a daughter. Studies in India have suggested that a dowry is the most common reason for not wanting a girl.61 Even when parents themselves didn't have a dowry in their own marriage, many carry the expectation of having to provide one for a daughter, or asking for one from a future daughter-in-law.
  • Labor force opportunities: males are often seen to bring better economic opportunities for a family. This can be the result of real sex differences in economic opportunities, but also results from undervaluing the work of women. In many countries, unpaid or informal work accounts for a majority share of female employment. In this sense, informal employment is often seen simply as an extension of a women's domestic work rather than being valued in its own right.62
  • Family and societal pressure: it's not just economic concerns that parents worry about: surveys also reveal persistent pressures from family members and communities. In interviews, women often note pressures from female in-laws and husbands, verbal and physical abuse when they didn't produce a son or were pregnant with a girl.63
  • Religion: in some cultures it's believed that only a son can light the funeral pyre and carry out death-related rites and rituals needed for salvation.64

What are the consequences of a skewed sex ratio?

Across several countries in Asia, and North Africa we see highly skewed sex ratios in favour of males. Strong preference for sons has led to an increasing number of "missing women" and therefore an "excess of males".

There are a number of adverse consequences which could result from highly imbalanced numbers of men and women in society. These present a risk to men, women, family structures and society as a whole.

The obvious consequence of gender imbalance is a large number of unmarriageable men.65 A lack of women of marriageable age could have several impacts on marriage dynamics for men66:

  • many men will have to delay marriage;
  • this will have a knock-on impact on younger men: not only will they be impacted by a skewed sex ratio for their generation, but will also have a 'backlog' of men from the previous generation in the 'marriage market';
  • a significant number of men will have to forego marriage altogether;
  • the poorest men will likely be most highly impacted by this – it's hypothesised that women will "marry up" in society, leaving the men with least resources disproportionately affected;
  • one of the most common reasons for son preference is for family lineage (the passing down of the family name through males in the family) – this will likely not occur in families where sons do not marry.

Whilst we might assume that this dynamic would favour women, they could also suffer negative consequences:

  • a women's value may be increasingly lie in her role as a wife, daughter-in-law or mother. This may deny her alternative pathways, such as remaining single or becoming more career-oriented;
  • women may feel increased pressure or incentives to marry or have children earlier in life (possibly closing off different economic or labour opportunities);
  • women may be at increased risk of violence (emotional, sexual or physical) or trafficking.67

Imbalanced gender ratios could have negative consequences for both men and women. For society more broadly, there are several hypotheses that it will also result in more crime, violence and disorder in communities. The rationale being that not only will there be an excess of men who do not marry and have their own family, but that the most affected will be those of lowest socioeconomic status, the most uneducated and with fewer opportunities.

Hudson and De Boer (2004) propose that this situation will result in significant social stability and security concerns.68 Although studies have found correlation between the prevalence of violence and more imbalanced gender ratios, it's difficult to disentangle a causal effect.69 Other researchers suggest that the evidence does not support this concern: that rather than showing through aggression and violence, excluded men will be more marginalised, lonely and at risk of psychological problems.70

What's clear is that the persistence of imbalanced sex ratios means there will be demographic consequences for many decades to come.

What does and doesn't work in addressing gender bias?

Do richer and more educated parents have less gender bias?

If our aim is to address issue of imbalanced sex ratio and female discrimination, an obvious question to ask is: will development fix the problem? Is it the case that son preference is restricted to those at lower incomes, therefore the problem disappear if poverty falls and societies develop?

The evidence to date suggests, no: development on its own does not address it.

Studies which have looked in detail at the extent of sex ratio imbalances by demographic (comparing income level, education, literacy, employment factors amongst others) often conclude the opposite: that richer, urban families tend to discriminate more than the poor.71 For example, in North India higher castes (who tend to be richer) had more skewed sex ratios than lower castes.72 Narrowing in on specific districts and household in India and South Korea suggested that it is richer families who show greater discrimination.73

Despite rising incomes and development across many countries in Asia, North Africa and Eastern Europe, the sex ratio got worse, not better. An important driver which may link the increase in sex ratio with development and rising incomes is fertility rates. Falling fertility rates could play a key role in exacerbating gender preference. This is often called the 'fertility squeeze'.

In smaller families, you're less likely to have a son by chance

Imagine you have five or six children: there's a very high chance that at least one of them will be a boy. Now imagine you have only one or two children: the likelihood you have a son is much smaller.

In many countries across the world, this transition happened: in India, average fertility rates fell from almost 6 to 2.3 children per woman; in China from over 6 down to 1.6 children; in South Korea from over 6 to close to only one child per woman. As countries got richer; education levels increased; women gained more rights to education labour participation and healthcare decisions; fertility rates fell rapidly.

But the result is that if you really want a son, your chances of having one naturally are now much lower. For parents of smaller family sizes, this could make sex-selective abortion much more likely.

Seema Jayachandran (2017) tried to quantify how much of the skew in sex ratio in India might be attributed to its decline in fertility rates.74 In the visualization below we summarize results from this research, through the use of three charts:

  • Top chart: we see the percentage of respondents of a survey in India who said they would prefer a son when given a hypothetical family size (number of children) from one to five i.e. having only one child; two children etc.75 Here we see that if they were having only one child, more than 80% of both mothers and fathers would prefer a son. This preference rapidly dropped off with more children; in fact, after having a few sons parents then preferred to have daughters. These results suggest that parents have a very strong preference to have one son or two sons (preferably as the eldest) but then prefer additional children are girls.76 Overall the trend is clear: in smaller families, there is stronger son preference.
  • Middle chart: in red we see that the desired fertility rate (children per woman) has declined in India in recent decades. At the same time, the child sex ratio (the number of males per 100 females) increased. The dashed line here shows the estimated change in sex ratio attributed the effect of this declining fertility (the effect we see in the top chart of smaller families showing greater son preference).
  • Bottom chart: the actual child sex ratio has increased over time (shown in blue). Shown in grey is the estimated increase in sex ratio which can be attributed to the decline in fertility rates.

Overall, Seema Jayachandran estimates that between one-third and one-half of the change in sex ratio in India since 1981 can be attributed to the decline in fertility rates. If there had been no change in fertility rates (family size) over time, there would still be a skew in the sex ratio. But fertility declines have had a significant impact on exacerbating son preference.

[Clicking on the visualization below will open it in higher-resolution].Impact fertility rate on son preference

Development drives factors which can both increase and decrease gender preference

Disentangling a clear relationship between development and sex ratio is difficult because education and rising incomes can impact aspects of gender selection in opposing ways.

That's because a skewed sex ratio not only results from gender preference but also the ability to act upon it (by having a sex-selective abortion, for example). In fact, sex-selection practices result from three conditions being met77:

  1. willingness to consider sex selection because you have a gender preference;
  2. ability to act upon it through access to prenatal sex-determination screening and abortion services;
  3. readiness to act upon it to get the desired sex preferences within the family.

How these three conditions change with development, education and rising incomes can be in conflict. For example, we might expect that son preference (or conditions 1; willingness) falls with increased education.78 But, increased education is also associated with falling fertility rates; as we explored above, this decline in fertility (the 'fertility squeeze') could increase son preference. Here we have conflict: education and progress decreases the willingness for gender preference, but increases readiness for influencing sex preferences within the family.

Development also affects condition (2): the ability to act upon gender preference. It is often the case that richer, urban families have greater access to technologies which would allow for sex identification during pregnancy, and a safer abortion. Poorer families in rural areas may not have this opportunity, despite having strong son preference. As they get richer, these opportunities may become available to them.

Kashyap and Villavicencio (2017) tried to model how the changes in willingnessability to act, and readiness all affect the change in sex ratio in specific countries.79 They found that even very low levels of son preference can have a significant impact on the sex ratio if sex-selection technology diffuses steadily and fertility declines quickly.

The relationship between development and sex ratio is complex and non-linear: what's clear is that the issue does not quickly disappear through economic development alone.

Does banning prenatal practices reduce sex-selective abortion?

One of the three conditions for sex-selection is the ability to act upon gender practices through access to technology.

Prenatal sex selection (PSS) relies on two technologies: prenatal sex determination (being able to determine the sex of a foetus during pregnancy) followed by selective abortion.

This raises the question of whether policies related to prenatal sex determination and abortion have an impact on the prevalence of sex selection.

With it becoming clear that sex selection was an ongoing and increasing issue, several governments tried to limit prenatal sex selection through regulation. South Korea, China and India all implemented sex-selective abortion bans. Were they successful?

Much of the literature on the topic suggests the bans were futile – that there are other policy options which would have been more effective.80,81.

Looking at how the sex ratio at birth changed pre- and post-ban in each country would also suggest that they were ineffective. South Korea: enacted a ban on prenatal sex identification in 1987/88. At this point the sex ratio at birth was around 110 male per 100 female births – as we see in the chart below. But following the introduction of the ban, the sex ratio continued to increase – reaching over 115 males per 100 female births in 1990 and maintaining a high ratio through the early 1990s.

China: in 1994, as part of its Law on Maternal and Infant Health Care, China introduced a ban of prenatal sex determination. As we see in the chart below, the sex ratio at birth continued to increase after the introduction of the ban. In 1993 the ratio was 114 male per 100 female births; by the early 2000s this had increased to 118:100.

India: it introduced the Pre-Conception and Pre-Natal Diagnostic Techniques Act (PNDT) in 1994 which prohibited sex selection practices, including pre-screening to determine the sex of a fetus. In the decade which followed the introduction of the act, the sex ratio at birth did not improve.

Click to open interactive version

At first glance, this would suggest that the banning of sex-selection practices was unsuccessful. The sex ratio continued to increase after their implementation.

But, as many researchers have highlighted: we don't know the counterfactual scenario without a ban.82 Without a ban the sex ratio might have increased even further. In fact, some studies have argued that bans were effective in preventing a worsening of gender imbalance.83

So these policies may have had some impact on reducing the increase, but they clearly did not come close to ending the practice.

What has been effective in reducing skewed sex ratio?

Data Sources

UN World Population Prospects
  • Data Source: UN Population Division (2017 Revision)
  • Description of available measures:
    • Population, by Five-Year Age Group and Sex
    • Population Sex Ratio (males per 100 females)
    • Sex ratio at birth
  • Time span: 1950-2100
  • Geographical coverage: Global, by country
  • Link: https://esa.un.org/unpd/wpp/Download/Standard/Population/

World Bank

Chao et al. (2019)
  • Data Source: Chao, F., Gerland, P., Cook, A.R., Alkema, L. (2019). Systematic assessment of the sex ratio at birth for all countries and estimation of national imbalances and regional reference levels. PNAS.
  • Description of available measures:
    • Sex ratio at birth (modelled estimates)
  • Time span: 1950-2017
  • Geographical coverage: Global, by country
  • Link: https://www.pnas.org/content/early/2019/04/09/1812593116