Official Statistics

Sex ratios at birth in the United Kingdom, 2017 to 2021

Published 12 October 2023

Executive summary

Aim

This report provides statistics on the analysis of male to female birth sex ratios in the United Kingdom using the most recent data for the period 2017 to 2021. This report was previously published for Great Britain until the 2020 publication, when it was extended to include Northern Ireland, covering the time period 2014 to 2018.

The purpose of these official statistics is to update and inform the public and Parliament on whether there is evidence of sex selective abortions happening at scale within specific communities in the United Kingdom. Data is sourced from birth registration data in England, Wales, Scotland and Northern Ireland. This analysis uses a generally accepted birth sex ratio upper limit of 107 (see ‘Trends in birth sex ratios’, that is, we are looking for birth sex ratios with significantly more than 107 males born for every 100 females within a group).

Where the ratio of males to females for a particular ethnicity or mother’s country of birth is significantly greater than 107 (after adjusting for multiple testing and the birth order of the child), this may indicate that people in this group have been involved in sex selective abortions. A lower birth sex ratio limit is not used, as we are not investigating if there are many more females born than males born in the United Kingdom.

This analysis has been repeated annually since 2013 and over this time, there has been no evidence of sex selective abortions occurring in the UK. We want to better understand how this publication is used, and by whom. Please see the ‘Further information’ section, below, for more details.

Key statistical terms used in this document are defined in the glossary, below, and should be read to ensure correct interpretation of the findings. Key terms relating to the source data, such as ethnicity and mother’s country of birth, are defined in the technical appendices.

Key findings from 2017 to 2021 data

This report presents statistics on the analysis of male to female birth sex ratios in the United Kingdom for the period 2017 to 2021. The key findings are outlined in the paragraphs below in this section.

This analysis finds evidence of a statistically significant imbalance in the ratio of males to females at birth for children of Indian ethnicity after 2 or more previous births, in England and Wales for the period of 2017 to 2021. This may indicate that sex selective abortions are taking place. If so, it is estimated that approximately 400 sex selective abortions may have taken place to female foetuses over the 5 year period from 2017 to 2021.

No other evidence suggesting sex selective abortions occurring in the United Kingdom over the period 2017 to 2021 has been found in this analysis.

There were 3.6 million births registered in the United Kingdom in this period with a ratio of males to females of 105.4, below the accepted upper limit of 107 (see Table 1 under ‘Results’).

Latest analyses by ethnicity of the child and mother’s country of birth for overall birth sex ratios, and by birth order, finds the birth sex ratio of Indian children where the mother has had 2 or more previous children (113 males to 100 females) significantly higher than a male to female birth sex ratio of 107. No other ratio was found to be significantly higher than a male to female birth sex ratio of 107 (see Tables 2 and 3 in the accompanying statistical data spreadsheet).

Background

The Department of Health and Social Care (DHSC) made a commitment to publish birth sex ratio analysis annually, in line with the recommendation of the Council of Europe Parliamentary Assembly that member states should ‘collect the sex ratio at birth, monitor its development and take prompt action to tackle possible imbalances’ and ‘encourage research on sex ratios at birth among specific communities’. If more males than females are being born than is expected, this could imply some pregnancies are being terminated with an abortion due to the sex of the fetus. This process is known as a sex selective abortion. However, a number of factors can influence the sex of a child including maternal and paternal age, coital rates, number of children and sex of previous children.

The sex ratio is defined as the ratio of males to females in a population and is generally expressed as the number of males per 100 females. The sex ratio at birth is usually expressed as the number of males born alive per 100 females born alive (OECD, 2010). The term ‘child sex ratio’ is used to refer to the ratio of males to females in a defined age group – typically 0 to 6 years but occasionally 0 to 4 and 0 to 5 years (UNICEF, undated; JSK, undated). In most countries, the normal sex ratio at birth varies between 102 and 106 males per 100 females (NationMaster, 2009^{[footnote 1]}).

The legal context

Concern has been raised in some countries about the occurrence of sex selective abortions^{[footnote 2]}, ^{[footnote 3]}.

Sex is not itself a lawful ground for abortion in England, Wales and Scotland (Abortion Act 1967). DHSC guidance, in May 2014, states that abortion on the grounds of sex alone is illegal (see point 25 in the Guidance in Relation to Requirements of the Abortion ACT 1967).

However, under the Abortion Act, it is lawful to abort a fetus where 2 registered medical practitioners (RMPs) (that is, doctors) are of the opinion, formed in good faith, “that there is a substantial risk that if the child were born it would suffer from such physical or mental abnormalities as to be seriously handicapped.” This would be classed as a Ground E abortion. There are some serious conditions which are known to be related to a person’s sex.

From 31 March 2020 a new framework for access to abortion services for Northern Ireland came into force. As with England, Wales and Scotland, sex is not itself a lawful ground for abortions in Northern Ireland.

In early 2015 the Serious Crime Act contained a requirement that the Secretary of State should arrange for an assessment to be made, within 6 months of Royal Assent, of the evidence for sex-related abortions occurring. A report was subsequently laid before Parliament in August 2015, which contained the assessment of evidence of abortions being undertaken on the grounds of the sex of the fetus, and a statement and explanation of why the Secretary of State for Health considered a formal plan under sub-section (3) (a) of the clause was not required.

Identifying the sex of a foetus with technology

Antenatal sexing of the foetus is not a routine part of antenatal care. Scans are undertaken to support the clinical care of the woman and foetus for purposes such as:

• ascertaining the number of foetuses
• ascertaining the age of the foetus
• undertaking screening for foetal anomalies

It is usually only possible to identify the sex of a foetus at the second ultrasound scan, which takes place at around 18 to 21 weeks gestation. Disclosing the sex of a foetus is a local decision, adhering to local policies, and should be based on clinical judgment about the certainty of the assessment and the individual circumstances of each case.

The introduction of new and emerging technologies (such as non-invasive prenatal testing (NIPT) which can identify sex of the fetus after 7 weeks gestation is currently available in private clinics and on the internet) provide further context from which the monitoring of birth sex ratios needs to be considered. NIPT testing is available on the NHS depending on eligibility criteria but will not include assessment of the sex of the foetus^{[footnote 4]}.

The majority of abortions take place in the first trimester of pregnancy - for example, 94% up to 12 weeks and 70% up to 7 weeks gestation in England and Wales (see Abortion Statistics, England and Wales, 2021 and Termination of pregnancy statistics for Scotland). The majority of abortions are therefore taking place nearly 2 months before most women could have been told the sex of the fetus in the antenatal screening pathway. It is acknowledged though that some people may seek information on sex using other methods such as tests available on the internet or in private clinics. However, data is not available on the extent to which other methods of sex determination may be being used. See Annex A, below, for further information on abortions by gestation and ethnicity.

Trends in birth sex ratios

Within large populations, we can expect the birth sex ratio to vary, due to external factors including wars and economic crises. Figure 1 shows the fluctuation within the United Kingdom since 1910. The chart shows that the birth sex ratio for the UK has never been above 107 over that period. The minimum ratio was 103.5 in 1914 and the maximum ratio was 106.7 in 1973. The male to female birth sex ratio has been consistently around 105 since 1980.

Figure 1: live male births per 100 live female births, England and Wales, Scotland and Northern Ireland, 1910 to 2021

Source: Office for National Statistics, National Records of Scotland and Northern Ireland Statistics and Research Agency.

The issue of sex ratios of newborns is the subject of numerous academic articles, where there is a range of evidence. A 2011 World Health Organisation report stated a biologically normal sex ratio at birth ranges from 102 to 106 male births, per 100 female births^{[footnote 5]}, although other studies have stated that 105 or lower is a more ‘normal’ ratio of males to females^{[footnote 6]}. Evidence suggests that a number of factors can influence the sex of a child, including maternal and paternal age, coital rates, number of children and sex of previous children^{[footnote 7]}. It is important to note that the interaction between factors that could influence the sex of the child has not been controlled for or taken into account in this analysis.

For this publication a birth sex ratio upper limit of 107 males to 100 females is used as a threshold for comparisons. This is based on a review of available literature ^{[footnote 8]}, advice from academic experts and on examination of data on birth sex ratios in more developed countries.

Calculations and statistical tests used

The aim of this analysis is to investigate if any of the birth sex ratios considered are statistically significantly higher than 107, in other words, if any group has statistically significantly more than 107 males born for every 100 females.

Birth sex ratios above the 107 threshold do occur. These could be a chance finding or they could be due to a real difference. Statistical significance testing is carried out to determine whether any differences observed between the birth sex ratios and the 107 threshold are likely to be ‘real’ or whether they are consistent with chance fluctuations.

This publication uses a number of methodologically approved techniques to test whether ratios over 107 are statistically significant. The first stage of the process is to calculate the probability (‘p values’) that the differences observed could arise by chance as opposed to there being a real difference. We have used the common acceptable level of 5% significance level in this analysis, which means that a statistically significant result is found for any p values less than 0.05 (5%). For more detail, see section ‘Glossary’ below or the technical appendices.

However, this methodology presents some difficulties when there are many tests involved. For example, the ‘mother’s country of birth’ analysis involved testing the significance level for 176 countries and 5 birth orders, theoretically equivalent to 880 statistical tests (In fact, 823 tests were carried out as 57 countries in the unknown birth order category had no data available). With so many tests, it would be expected that some results appear statistically significant due to chance alone. Even when there was no real difference from a ratio of 107, on average, a 5% significance level means that 1 in 20 (or 5 in 100) results would be found to be significant. When applying this across these mother’s country of birth tests there is a high chance of a ‘false positive’ result, therefore inferring evidence about sex selective abortions incorrectly.

To address this issue, known as the ‘multiple testing problem’, a statistical technique called the Benjamini-Hochberg procedure was applied using p values already calculated as part of our method to assess statistical significance. Following this, the Storey technique is used to estimate the false positive discovery rate as a form of sensitivity analysis. The application of these techniques is discussed in detail in the technical guidance.

Limitations of the analysis

Using birth registration data to calculate birth sex ratios is an indirect method for investigating evidence of sex selective abortions.

The relatively small number of births within many of the groups in this analysis are such that large differences between birth sex ratios and the expected upper limit of 107 would need to be observed for the ratio to be identified as statistically significantly higher than the expected upper limit of 107. Therefore, evidence would only be identified through this means if sex selection was taking place on a significant scale.

Any differences in the birth sex ratios seen could be due to a number of factors, not just sex selective abortions. There is evidence that paternal and maternal age, coital rates and the number and sex of previous children can influence the sex of a child.

Results

Coverage of this analysis

This report presents statistics on the analysis of male to female birth sex ratios in the United Kingdom for the period 2017 to 2021. This report was previously published for Great Britain until 2020 where it was extended to include Northern Ireland.

Sex ratios are examined overall and by birth order (first born, second born and so on) for:

• all children born in 2017 to 2021 for the United Kingdom
• ethnicity of child born in 2017 to 2021 for England and Wales
• the mother’s country of birth for children born in 2017 to 2021 for the United Kingdom

All births

In the period 2017 to 2021, there were 3.6 million live births in the United Kingdom and an overall ratio of males to females of 105.4 which is below the upper limit of 107. (Analysis covered countries of mothers’ birth where the total number of births for 2017 to 2021 was 100 or more to ensure adequate sample sizes, so excluding some countries.)

The birth sex ratio did not vary significantly across the different birth orders with a ratio between 105 and 107 males to 100 females (see Table 1 below).

The birth sex ratio for unknown birth order is above the 107 threshold but was not found to be statistically significant when tested. Therefore, it does not show evidence of at scale sex selective abortions.

Table 1: birth sex ratios by birth order, United Kingdom, 2017 to 2021

Birth order	Number of births	Birth sex ratio
All births	3,572,980	105.4
First born	1,530,451	105.7
Second born	1,242,506	105.3
Third born or more	788,519	105.0
Unknown	11,504	107.3

(Unknown birth order represents those children whose birth order was unknown.)

Births by ethnicity of the child and birth order

Data on ethnicity of the child is not available for Scotland and Northern Ireland, therefore this component of the analysis is limited to England and Wales.

There were 3.2 million births included in the analysis of birth sex ratios by ethnicity of the child and birth order, for England and Wales. The majority of births in England and Wales were for children in the white British ethnic group (62% of stated ethnicities).

An analysis of birth sex ratios by the ethnicity of the child for England and Wales both for overall birth sex ratio and by birth order, found one significant result. The birth sex ratio for children of Indian ethnicity of the birth order 3 or more was 113 and found to be significantly higher than 107 (see Table 3 in the accompanying statistical data spreadsheet).

Although some other birth sex ratios were higher than 107, such as children of unknown birth order for Bangladeshi, black African, white British and not stated ethnicity, they were found not to be statistically significant.

Figure 2: birth sex ratios by ethnicity: England and Wales, 2017 to 2021

Figure 3: birth sex ratio by ethnicity and birth order: England and Wales, 2017 to 2021

When interpreting the birth sex ratios (see Table 3 in the accompanying statistical data spreadsheet), it is important to keep in mind that this analysis covers 12 ethnic groups, in line with the ethnic groupings used by the Office for National Statistics, for 5 categories of birth order (all, first born, second born, third born or later, and unknown birth order). This breaking down of the overall number of births means that some of these groups have very small numbers, such as 112 births in ‘any other black background’ for unknown birth order. We would therefore expect to see high birth sex ratios for some groups simply as a result of random variation and small sample sizes (see figure 4 below).

Figure 4: plot of male to female birth sex ratio by ethnicity and birth order: England and Wales, 2017 to 2021

The above chart (figure 4) is a funnel plot. That is a chart that shows the measure of interest (birth sex ratio) on the vertical (y) axis and sample size on the horizontal (x) axis. There is a line showing the acceptable maximum ratio (107), and another line showing ‘funnel limit’ lines, or the 95% confidence intervals. A point outside these funnel limits (confidence interval) indicates a value that would be very unlikely to occur only by chance based on a single statistical test for significance.

In figure 4, the majority of sex ratios above the 107 threshold have a smaller number of live births. The Indian ethnicity, where birth order is 3 or more, is the exception as it has a sex ratio of 113 for 15,401 births, which is above the funnel limits (confidence interval). When tested, this was found to be statistically significant. The finding of statistical significance for Indian ethnicity where birth order is 3 or more is maintained when accounting for multiple testing, unlike the finding for mixed ethnicity which, when accounting for multiple testing is no longer statistically significant.

As described above in ‘Trends in birth sex ratios’, a birth sex ratio over 107 males to 100 females may indicate sex selective abortions. The evidence suggests there may have been approximately 400 sex selective abortions to female fetuses of Indian ethnicity, after 2 or more previous children, in England and Wales over the 5 year period from 2017 to 2021. More detail on estimating the number of sex selective abortions can be found in the technical appendices.

In the previous publication, Sex ratios at birth in the United Kingdom: 2016 to 2020, the birth sex ratio for Indian ethnicity where birth order is 3 or more was also above the 107 threshold at 111, however this was not found to be statistically significant.

Figure 5: number of legal abortions by gestation: Indian ethnicity, 2 or more previous live or still births, England and Wales, 2017 to 2021, Abortion Notification System

Source: Abortion Notification System, DHSC.

Figure 5, above, shows abortion data from the Abortion Notification System rather than data on births. Abortion data includes the ethnicity of the woman having the abortion while birth data includes the ethnicity of the child.

Figure 5 shows that between 2017 and 2021 in England and Wales, there were 13,843 abortions where the woman was of Indian ethnicity and had 2 or more previous live or stillbirths. The majority (87.5%) of these abortions occurred before the point at which is it possible to identify the sex through NIPT testing in private clinics (after 7 weeks gestation). There were 4,133 abortions between 8 and 17 weeks gestation. Just 298 abortions were at 18 weeks gestation or more, where it’s possible to find out the sex of a foetus in the NHS pregnancy patient pathway.

Births by mother’s country of birth

The majority of births in the United Kingdom were to mothers born in England, Wales, Scotland and Northern Ireland (73%).

The analysis for the United Kingdom on the overall birth sex ratio and birth order by the mother’s country of birth, showed no result to be significantly higher than 107. It is important to note that the analysis of births by mother’s country of birth did produce some birth sex ratios higher than 107 (see Table 2 in the accompanying statistical data spreadsheet). For example, Togo’s birth sex ratio for first childbirth is 242 males to 100 females. However, this is likely the result of random variance on a small sample size of 41 births. Further analysis conducted on the data using the Benjamini-Hochberg and Storey statistical techniques were not suggestive of sex selective abortions.

Table 2 (in the accompanying statistical data spreadsheet) shows the birth sex ratios for all the countries included in the analysis. When interpreting the birth sex ratios in Table 2, it is important to keep in mind that this analysis covers 176 country of birth groups for 5 categories of birth order (all, first born, second born, third born or later, and unknown birth order). We would therefore expect to see high birth sex ratios for some groups simply as a result of random variation and small sample sizes. To illustrate, between 2017 to 2021, the birth sex ratio for women born in Djibouti, there were 24 babies who were the first born with a birth sex ratio of 100 while 31 babies were the second born with a birth sex ratio of 244. However, this has not produced a statistically significant result. This wide variation exists, particularly where sample sizes are smaller (see Figure 6 below).

Figure 6: plot of male to female birth sex ratio by mother’s country of birth: all births, United Kingdom, 2017 to 2021

The above chart (Figure 6) is a funnel plot. That is a chart that shows the measure of interest (birth sex ratio) on the vertical (y) axis and sample size on the horizontal (x) axis. There is a line showing the acceptable maximum ratio (107), and another line showing ‘funnel limit’ lines, or the 95% confidence intervals. A point outside these funnel limits (confidence interval) indicates a value that would be very unlikely to occur only by chance based on a single statistical test for significance.

Figure 6 shows the greater variation in the birth sex ratio for countries with lower numbers of live births. Over half the countries (64%) are below the 107 ratio threshold, with most of those that are higher having small sample sizes, for example Benin (182 births) and Paraguay (112 births) where the ratio is 140 and 129 males to 100 females respectively. Similar variation is seen below the 107 line, with Barbados where there were 89 males born for every 100 females. As noted above, when testing using the Benjamini-Hochberg technique, there were no countries found to be statistically significantly higher than the 107 ratio threshold.

Abortions by gestation and ethnicity

To supplement the birth sex ratio analysis, further analysis has been done to investigate the relationship between gestation (and thus when sex can be identified) and abortion rates, by ethnic group of the woman. This shows that the number of abortions vary by ethnicity of the woman, gestation period and number of previous abortions. The majority of abortions occur before the women would know the sex of the fetus. See Annex A, below, for further detail.

Conclusion

Following extensive statistical testing, this analysis of birth sex ratios finds a statistically significant imbalance in the ratio of males to females at birth for children of Indian ethnicity after 2 or more previous births, in England and Wales for the period of 2017 to 2021. This suggests that sex selective abortions are taking place. This is the first evidence of sex selective abortions since these statistics were first published.

No further evidence of sex selective abortions occurring in the United Kingdom have been found in these statistics.

Annex A: abortions by gestations and ethnicity

Data on the sex of the foetus at an abortion is not available, which is why birth registrations data are used in this analysis. We can however analyse abortion data by ethnicity of the woman (abortion data does not include mother’s country of birth) and gestation to provide wider context around the birth sex ratio analysis. Any evidence for unusual birth sex ratios might imply sex selective abortions were taking place in the population.

This section adds some additional context on abortions by gestation and ethnicity from the annual abortion statistics publication, Abortion statistics for England and Wales, 2021.

This annex refers to abortions for residents of England and Wales in 2021.

All abortions, by gestation

The total number of abortions for residents of England and Wales in 2021 was 214,256. This was an increase of 2% from 2020, and the highest number recorded since the Abortion Act was introduced. The majority of abortions took place in the early stages of pregnancy: 94% up to and including 12 weeks, and 70% up to and including 7 weeks gestation. NHS antenatal sexing of a fetus typically takes place much later in the pregnancy at 18 to 21 weeks gestation and Chart 1 shows 98% of abortions took place at 17 weeks gestation or less.

The majority of abortions are therefore taking place nearly 2 months before most women could have been told the sex of the fetus in the antenatal screening pathway. It is acknowledged, though, that some people may seek information on sex using other methods such as tests available on the internet or in private clinics. However, data is not available on the extent to which other methods of sex determination may be being used.

Chart 1: proportion of abortions by gestation week, England and Wales, 2021

All abortions by ethnic group of woman

For residents of England and Wales ethnicity was recorded on 91% of the forms received for 2021. Of all ethnicities recorded in the Abortions Statistics publication, 78% of abortions were reported as relating to women of white ethnicity, 9% as Asian or Asian British and 7% as black or black British, 5% as mixed and 1% as Chinese or other ethnic group (see Chart 2).

Chart 2: Legal abortions by ethnic group of woman, England and Wales, 2021

All abortions, by gestation and ethnic group of woman

There are variations between the gestation period of an abortion when comparing different ethnicities. The proportion of abortions before 17 weeks range from 97% to 99% depending on the ethnic group of the woman. For abortions carried out before 12 weeks the range between different ethnic groups is from 90% to 94%, with a wider range for abortions carried out before 7 weeks from 47% to 55% depending on the ethnic group of the woman.

Table 2: abortions by gestation period and ethnic group, residents of England and Wales, 2021

Ethnic group of woman	Total number of abortions	Less than 7 weeks gestation	7 to 11 weeks gestation	12 to 16 weeks gestation	17 to 21 weeks gestation	22 weeks and over gestation
All ethnic groups	214,256	50%	42%	6%	2%	1%
White - British	134,000	50%	42%	5%	2%	1%
White - Irish	1,194	49%	42%	6%	2%	0%
White - any other white background	16,445	48%	42%	7%	2%	1%
Mixed - white and black Caribbean	4,031	47%	44%	6%	2%	1%
Mixed - white and Asian	1,339	52%	42%	4%	1%	0%
Mixed - white and black African	1,339	49%	43%	5%	2%	1%
Mixed - any other	2,744	48%	43%	6%	2%	1%
Asian or Asian British - Indian	6,945	48%	43%	6%	2%	1%
Asian or Asian British - Pakistani	4,389	51%	41%	5%	2%	1%
Asian or Asian British - Bangladeshi	1,741	49%	44%	4%	1%	0%
Asian - any other Asian background	4,176	51%	40%	5%	2%	1%
Black or black British - African	8,337	51%	41%	5%	2%	1%
Black or black British - Caribbean	3,689	48%	44%	6%	2%	1%
Black or black British - any other	1,810	50%	42%	6%	2%	1%
Chinese	838	47%	44%	6%	3%	0%
Any other ethnic group	1,209	55%	36%	5%	3%	0%
Not known or not stated	20,030	54%	38%	5%	2%	1%

Note: percentages are rounded and may not add up to 100.

For abortions under 7 weeks, the ‘Any other ethnic group’ women have the highest proportion (55%). The ethnic groups with the lowest proportion of abortions under 7 weeks gestation were: ‘Mixed - white and black Caribbean and Chinese’ (47%).

Abortions over 17 weeks account for 2% of all abortions. There is some variation across ethnicities for this gestation. ‘Asian or Asian British-Indian’, ‘Asian or Asian British-Pakistani’, ‘Asian or Asian British-Bangladeshi’, ‘Chinese’, ‘White - any other white background’ and ‘Any other ethnic groups’ have 3% of abortions over 17 weeks. Women who have stated their ethnicity as ‘Mixed - white and Asian’ have 1% of abortions over 17 weeks. 

Repeat abortions and ethnic group

Of all abortions in 2021, 43% were to women who had previously had one or more abortions. Again, amongst women who have had a previous abortion, there was variation across ethnic groups. For example, 49% of Mixed ethnicity women having abortions in 2021 had previously had at least one abortion, compared with 33% of Asian ethnicity.

Table 3: percentage of women who had one, 2, 3 or more previous abortions, by ethnic group of the woman, Residents of England and Wales, 2021

Ethnic group of woman	No previous abortions	1 previous abortion	2 previous abortions	3 or more previous abortions	Total
All ethnicities	57%	30%	9%	3%	100%
White	57%	31%	9%	3%	100%
Mixed	51%	33%	11%	5%	100%
Asian	67%	26%	6%	2%	100%
Black	52%	33%	11%	4%	100%
Chinese or other	66%	24%	7%	3%	100%
Not known or not stated	60%	29%	8%	3%	100%

Note: percentages are rounded and may not add to 100.

Conclusion

This section shows that abortions vary by ethnicity of the woman, gestation period and whether the woman has had previous abortions. There is a similar pattern across all ethnicities with limited variation.

Glossary

Key terms relating to the source data, such as ethnicity and mother’s country of birth, are defined in the technical appendices.

Variation

Variation is a measure of how spread out the data is. One way to measure the variation of the data would be to calculate the range (the difference between the highest value and the lowest value).

Statistical significance

To say that something is statistically significant means that there is a high probability that what we are observing is real and not just a random occurrence. The commonly accepted threshold for statistical significance is 5%. This means that the likelihood of an event happening by pure chance is less than 5%. We conclude that there must be a reason other than chance to determine the event.

Sample and population

In this analysis, the population of interest is all births (children) in the United Kingdom (and separately in England and Wales) between 2017 and 2021. We also focus on a specific subset of the population data, child’s ethnicity and mothers’ country of birth, which is a sample within the larger population.

Sample size

In this analysis, the sample size refers to the number of births in a given population group. For example, a sample size would be the number of births where the child’s ethnicity is white British. With a small sample size there is a greater risk of an unusual result occurring by chance. We expect to see very high (or low) birth sex ratios for some groups simply as a result of random variation and small sample sizes. Larger samples more closely approximate the population one is interested in studying.

Discovery rate

The ‘false discovery rate’ is the proportion of results that are statistically significant by chance, also referred to as a false positive result. It is calculated by taking the number of false positives and dividing by the number of observations.

Sensitivity analysis

‘Sensitivity analysis’ is a way to see how sensitive or responsive something is to changes. In statistics it is a technique used to assess how variations or changes in the input or assumptions of a statistical model impact its output or results. It involves systematically testing different scenarios or values for specific variables to understand the model’s robustness and how sensitive its conclusions are to different conditions.

Confidence interval

A confidence interval is a way of expressing the precision of an estimate (or the uncertainty surrounding it). The most common interval (the 95% confidence interval) shows where we confidently expect the true result from a population to lie 95% of the time. This analysis uses a 95% confidence interval in figure 4. This shows where a particular value of sex ratio is outside of the expected range for a given population size.

Further information

As part of the ongoing development of this publication, we are reviewing its content and methodology. If you have any feedback relating to this publication or are a user of these statistics and would like to be consulted about its content and/or methodology, please get in touch at: birthratios@dhsc.gov.uk.

Enquiries

Enquiries about the data or requests for further information should be addressed to:

Abortion Statistics Team
Department of Health and Social Care
10 South Colonnade
Canary Wharf
London
E14 4PZ

Email: birthratios@dhsc.gov.uk

Extracts from this publication may be reproduced provided a reference to the source is given.

Links

See Sex ratios at birth: statistics for birth sex ratio analysis for 2016 to 2020, and previous reports.

Related links

See Abortion Statistics 2021

---

1. Source: NationMaster, 2009, quoted in: ‘Preventing gender-biased sex selection’ an interagency statement OHCHR, UNFPA, UNICEF, UN Women and WHO, ISBN 978 92 4 150146 0 (NLM classification: QS 638) © World Health Organization 2011 (viewed on 25 June 2009) ↩

2. Hesketh, T and Xing, Z W (2006). Abnormal sex ratios in human populations: causes and consequences. Proceedings of the National Academy of Sciences, 103(36), 13271-13275. ↩

3. Chahnazarian, A (1988). Determinants of the sex ratio at birth: Review of recent literature. Social biology, 35(3-4), 214-235. ↩

4. In this test a small blood sample is taken from the mother’s arm. The UK National Screening Committee made a recommendation to use NIPT, a contingent test in the NHS Fetal Anomaly Screening Programme only for Trisomy 13,18 and 21 and not for any other genetic marker, including sex. (A trisomy is a disorder characterised by an additional chromosome.) ↩

5. World Health Organization (2011). Preventing gender-biased sex selection: an interagency statement. OHCHR, UNFPA, UNICEF, UN Women and WHO. ↩

6. Hesketh, T and Xing, Z W (2006). Abnormal sex ratios in human populations: causes and consequences. Proceedings of the National Academy of Sciences, 103(36), 13271-13275. ↩

7. Jacobsen, R, Møller, H and Mouritsen, A (1999). Natural variation in the human sex ratio. Human reproduction, 14(12), 3120-3125. ↩

8. Chahnazarian, A (1988). Determinants of the sex ratio at birth: Review of recent literature. Social biology, 35(3-4), 214-235. ↩