Official Statistics

Sex ratios at birth in the UK: 2019 to 2023 - technical appendices

Published 16 September 2026

Introduction

This report is a supplementary document to the main commentary section of the ‘Sex ratios at birth in the UK: 2019 to 2023’ publication. This document provides technical detail on the methodology and statistical tests used to analyse sex ratios at birth in the UK.

Users of the statistics

The sex ratio at birth statistics are of interest to the Council of Europe who originally requested their collation. Following the amendment in the Serious Crime Act 2015, Parliament used the statistics within their remit to assess the legality of the Abortion Act 1967 and assess the birth sex ratios within the population in Great Britain.

Academics and journalists reviewing evidence for sex selective abortions also have an interest in these statistics. The United Nations Population Fund reviews sex ratios at birth at a global level.

Data sources

Birth registrations

Analysis is presented by mother’s country of birth and birth order (first born, second born, third born or later, and unknown birth order) for the UK. Data on birth order and the mother’s country of birth is collected through birth registrations in England, Wales, Scotland and Northern Ireland. This data was supplied to the Department of Health and Social Care (DHSC) by the:

• Office for National Statistics (ONS)
• National Records of Scotland
• Northern Ireland Statistics and Research Agency

NHS birth notifications

Analysis is presented for ethnicity of the baby (as stated by the mother) for babies born in England and Wales, in line with the ethnicity groupings used by ONS. 

For data on ethnicity of the baby (as stated by the mother), birth registrations are linked with NHS birth notifications, which capture the ethnicity of the baby. These birth notifications are linked with birth registrations, and aggregated data is supplied to DHSC by ONS.

Ethnicity of the baby is not recorded on birth notifications or birth registrations in Scotland and Northern Ireland. Therefore, this analysis covers England and Wales only.

Calculations

Ratios

The sex ratios at birth were calculated by dividing the number of male births by the number of female births and multiplying this value by 100 to achieve a ratio of the number of males born per 100 females. For the publication, these ratios were then rounded to the nearest decimal place. This calculation was applied to:

• all births, by mother’s country of birth and birth order, in the UK
• all births, by ethnicity of child and birth order, in England and Wales

Confidence intervals

For the first time, 95% confidence intervals have been published alongside each ratio in the accompanying data tables. These were calculated using the Wilson’s score method.^{[footnote 1]} 

These confidence intervals should only be used for understanding the uncertainty and precision of each ratio and must not be used for identifying statistically significant results. Statistical analysis has been conducted separately using a method that has been reviewed and quality assured by ONS.

Rounding

In the accompanying data tables, all numbers have been rounded to the nearest multiple of 5. The purpose of this is to protect sensitive information and to prevent individuals from being identified. Ratios and confidence intervals are calculated from the unrounded figures and have been rounded to the nearest decimal place.

Statistical analysis

Independent review of methodology

In 2016, DHSC asked the Methodology Advisory Service at ONS to review the methodology and provide assurance that it was a robust approach for analysing sex ratios at birth.

ONS made the following recommendations, which have been implemented in the analysis:

• when implementing the Benjamini-Hochberg procedure (see below), the process should involve calculating the probabilities and then ranking these results in descending order in one operation, rather than doing separate tests by all births and birth order
• the Benjamini-Hochberg procedure may be supplemented with an analysis using Storey’s (2001) approach (see below) to estimate the local positive false discovery rate (pFDR)
• continue to aggregate 5 years of data in the analysis to ensure that the sample size is adequate to be able to detect a specified difference
• one-sided tests against a ratio of 107 males per 100 females are appropriate - this was based on a review of available literature, advice from academic experts and examination of data on birth sex ratios in more developed countries
• birth sex ratios for second-born child or more should no longer be reported (previously, DHSC analysis reported on male to female birth sex ratios for 2 or more children - however, as the birth order data for second-born or more children is closely related to third-born or more children, the recommendation was to no longer report the birth sex ratios for second-born children or more children)
• ratios and analyses for the unknown birth orders should be reported (it is possible that any evidence of sex selection could show up in the unknown birth order category - therefore, given that the birth order is of primary policy interest, the methodology review recommended reporting birth sex ratios and analyses for the unknown birth order from 2016)

Threshold used in the analysis

This analysis uses an upper threshold of 107 male births per 100 female births. This threshold was selected based on:

• a review of academic literature^{[footnote 2][footnote 3]}
• advice from academic experts
• an examination of sex ratio at birth data from developed countries

The analysis investigates whether any groups have sex ratios at birth that are statistically significantly higher than this threshold. A sex ratio at birth above 107 means that more than 107 males were born for every 100 females. Where a ratio is found to be statistically significantly above 107, this may be consistent with sex selective abortion having taken place.

A lower threshold was not used because the purpose of this analysis is to investigate whether there is evidence of unusually high numbers of male births relative to female births. It does not investigate whether there are unusually high numbers of female births relative to male births.

Testing whether birth sex ratios exceed the threshold

Sex ratios at birth are analysed for all births and by birth order (first born, second born, third born or later, and unknown birth order), the mother’s country of birth and the baby’s ethnicity. To ensure reliable analysis is carried out, all mothers’ countries of birth with fewer than 100 total births were excluded from the analysis and from the publication.

Some groups have sex ratios at birth above the expected upper threshold of 107 males per 100 females. However, differences from this threshold can occur by chance and do not necessarily indicate a real difference.

Statistical tests are used to assess whether the observed sex ratios at birth are likely to differ from the threshold of 107 by more than would be expected through chance variation alone. These tests produce p-values, which indicate how likely it is that an observed difference could occur if there were no real difference from the threshold. A significance level of 5% is used in this analysis. This means that results with a p-value below 0.05 would normally be considered statistically significant and unlikely to have arisen by chance alone.

In this analysis, we carried out many statistical tests across different groups defined by the mother’s country of birth and the ethnicity of the baby.

Analysis for the mother’s country of birth involved testing 178 countries and 5 birth orders, which would theoretically equate to 890 statistical tests being carried out. However, an additional 5 tests were also carried out to analyse sex ratios at birth for the total number of births in the UK across all 5 birth orders. In addition, 38 countries in the unknown birth order category were excluded due to having either zero males or zero females, making it impossible to calculate a birth sex ratio. In total, 857 statistical tests were carried out in the analysis of mother’s country of birth and birth order.

The analysis of ethnicity of the baby and birth order involved testing 12 ethnic groups and 5 birth orders, equivalent to 60 statistical tests. An additional 5 tests were also carried out to analyse birth sex ratios for the total number of births in England and Wales across all 5 birth orders. Due to having large numbers of births, no ethnic groups or birth orders were excluded from statistical testing in this analysis. In total, 65 statistical tests were carried out in the analysis of ethnicity of the baby and birth order.

However, because a large number of statistical tests were carried out across different groups, additional methods were needed to account for the increased likelihood of finding statistically significant results by chance.

Adjusting for multiple statistical tests

When large numbers of tests are performed, the likelihood of finding at least one statistically significant result purely by chance increases. This is known as the multiple testing problem.

To account for this, we used the Benjamini-Hochberg procedure, a statistical method designed for situations where many tests are carried out at the same time. The method considers the results from all tests together and calculates a set of Benjamini-Hochberg critical values. Each test result is then compared with the relevant critical value to determine whether it remains statistically significant.

A limitation of the Benjamini-Hochberg procedure is that its ability to identify statistically significant differences depends on the size of the groups being analysed. In groups with large numbers of births, relatively small differences in sex ratios at birth may be identified as lying outside the expected range. However, many of the groups in this analysis contain relatively few births, meaning that only large deviations from the expected upper limit of 107 male births per 100 female births would be identified as statistically significant.

Sensitivity analysis: Storey technique

Given the limitations of the Benjamini-Hochberg procedure, an alternative statistical analysis was conducted to check the validity of the results.

Storey^{[footnote 4]} and Storey and Tibshirani^{[footnote 5]} suggested an alternative procedure, where the false discovery rate is estimated for a fixed region called the ‘critical region’ - that is, the range of values under which we would reject the hypothesis of there being no countries or groups with a ratio above 107, and so the result would be statistically significant. This area is called the q-value and can be compared across different rejection regions as evidence for what proportion of false discoveries is actually seen across the series of tests.

Therefore, the Storey technique is used to estimate how many of the statistical tests performed were ‘true positives’ at the 5% significance level. This differs from the Benjamini-Hochberg procedure which makes adjustments to the critical values for the group of tests being used, in such a way as to control the false discovery rate (that is, to limit the proportion of outcomes where the test says that a result is significant, but no effect is actually present).

Statistical power

The ability of a statistical test to detect a difference depends partly on the number of births included in the analysis. In general, larger groups provide more information and make it easier to detect small differences from the expected sex ratio at birth of 107 male births per 100 female births.

Statisticians refer to this ability to detect a real difference as the ‘power’ of a test. A test with higher power is more likely to identify a statistically significant result when a group’s true sex ratio at birth is above 107.

The analyses presented in this publication involve a large number of statistical tests. Because the Benjamini-Hochberg procedure is used to account for multiple testing, it is not straightforward to calculate the statistical power of the overall testing process.

For illustrative purposes, table 1 shows how large an observed sex ratio at birth would need to be for a single statistical test to identify a result as statistically significant at the 5% level.

The table demonstrates that small groups require larger departures from the threshold of 107 before a statistically significant result can be identified. For example, where a group contains 100 births, the observed sex ratio at birth would need to be at least 149 males per 100 females to be identified as statistically significant. By contrast, where a group contains 100,000 births, an observed sex ratio at birth of 108 males per 100 females would be sufficient.

This means that the analysis is better able to detect relatively small differences from the threshold in groups with large numbers of births than in groups with relatively few births.

Table 1: required observed sex ratio at birth for a result to be identified as statistically significant at the 5% level for the number of births shown

Number of births	Observed sex ratio at birth required for a statistically significant result in a single test
100	149
500	124
1,000	119
5,000	112
10,000	111
50,000	109
100,000	108

For comparison with the data in the 2019 to 2023 publication, table 2 shows how many mothers’ countries of birth had a number of births in the shown range.

Table 2: number of mothers’ countries of birth by total number of births, UK, 2019 to 2023

Number of births	Number of mother’s countries of birth
0 to 99	0
100 to 499	53
500 to 999	29
1,000 to 4,999	48
5,000 to 9,999	16
10,000 to 49,999	24
50,000 to 99,999	5
100,000 to 499,000	2
500,000 or more	1

Estimating the number of additional females born

The number of additional female births required to achieve a sex ratio at birth of 107 males per 100 females can be estimated by performing the following calculation:

1. Divide the number of male births by 107 and multiply by 100 to calculate the estimated number of female births.
2. Subtract the observed number of female births from the estimate to calculate the number of additional female births required.

Table 3: estimated number of additional female births required to achieve a sex ratio at birth of 107 males per 100 females

Sex ratio at birth	100 total births	500 total births	1,000 total births	5,000 total births	10,000 total births	50,000 total births	100,000 total births
107	0	0	0	0	0	0	0
108	0	2	4	22	45	225	449
109	1	4	9	45	89	447	894
110	1	7	13	67	134	668	1,335
111	2	9	18	89	177	886	1,772
112	2	11	22	110	220	1,102	2,204
113	3	13	26	132	263	1,316	2,633
114	3	15	31	153	306	1,529	3,057
115	3	17	35	174	348	1,739	3,478
116	4	19	39	195	389	1,947	3,894
117	4	22	43	215	431	2,153	4,307
118	5	24	47	236	472	2,358	4,716
119	5	26	51	256	512	2,560	5,121
120	6	28	55	276	552	2,761	5,523

Contact

As part of the ongoing development of this publication, we are reviewing its content and methodology. We would like to hear from our users about how you use our products, how well these products meet your needs and any feedback on how they could be improved.

Contact us at birthratios@dhsc.gov.uk.

1. See DHSC Technical guidance: proportions and odds (accessed 10 September 2026). ↩

2. Hesketh T and Xing ZW. Abnormal sex ratios in human populations: causes and consequences. Proceedings of the National Academy of Sciences 2006: volume 103, issue 36, pages 13,271 to 13,275. ↩

3. Chahnazarian A. Determinants of the sex ratio at birth: review of recent literature. Social Biology 1988: volume 35, issues 3 and 4, pages 214 to 235. ↩

4. Storey JD. A direct approach to false discovery rates. Journal of the Royal Statistical Society: Series B (Statistical Methodology) 2002: volume 64, issue 3, pages 479 to 498. ↩

5. Storey JD and Tibshirani R. Statistical significance for genomewide studies. Proceedings of the National Academy of Sciences 2003: volume 100, issue 16, pages 9,440 to 9,445. ↩