Population estimates by marital status and living arrangements, England and Wales: 2002 to 2014

• In 2014, 51.5% of people aged 16 and over in England and Wales were married or civil partnered while 33.9% were single, never married.

• Between 2002 and 2014 the proportions of people aged 16 and over who were single or divorced increased but the proportions who were married or widowed decreased.

• The increase between 2002 and 2014 in the percentage of the population who were divorced was driven by those aged 45 and over, with the largest percentages divorced at ages 50 to 64 in 2014.

• In 2014 around 1 in 8 adults in England and Wales were living in a couple but not currently married or civil partnered; cohabiting is most common in the 30 to 34 age group.

• More women (18.9%) than men (9.8%) were not living in a couple having been previously married or civil partnered; this is due to larger numbers of older widowed women than men in England and Wales in 2014.

This bulletin presents annual estimates of the population by legal marital status and living arrangements for England and Wales. The estimates cover the years 2002 to 2014, broken down by age group and sex, and are consistent with the annual Mid-Year Population Estimates.

Statistics for 2011 to 2014 have been published today for the first time. Statistics published today for the years 2002 to 2010 have been revised from previous publications. The new statistics for 2002 to 2014 have been calculated using a new methodology. Analysis of the impact that the new methodology has had on the estimates for the years 2002 to 2010 is available in this bulletin.

Reference tables are available alongside this bulletin. Reference table A contains two tables; Table 1 presents estimates of the population by legal marital status, age group and sex. Table 2 presents estimates of the population by living arrangements, broken down further by grouped legal marital status, age group and sex. These reference tables present estimated numbers, to allow more in-depth analysis by users. However this bulletin focuses mainly on the percentages of people in each marital status or living arrangement category by age and sex, to allow quick and meaningful understanding of the trends.

The Population Estimates by Marital Status and Living Arrangements can be used in the production of marriage and divorce rates and also in the estimation of pension liabilities by insurance and pension firms. See ‘Who uses the Population Estimates by Marital Status and Living Arrangements?’ section for more information.

The method used to produce the estimates has changed to incorporate estimates of the cohabiting population and to include estimates of people in civil partnerships - more details of the change are provided in the ‘Methods’ section. Analysis showing the differences in the estimates when using the old or the new methodology is detailed in the section ‘Impact of the change in methodology on the population estimates by marital status’ section.

Marital status indicates whether a person is legally married or not. This publication uses 5 categories of legal marital status:

• Single, never married

• Married, including separated (this category includes those in both opposite and same sex marriages). See ‘Note about marriage of same sex couples’ for more information about the statistics within this publication.

• Civil partnered, including separated (civil partnerships were introduced in England and Wales in December 2005)

• Divorced, including legally dissolved civil partners

• Widowed, including surviving civil partners

Living arrangements indicate whether people are living with a partner or not, irrespective of their legal marital status.

Just over half (51.2%) of adults in England and Wales in 2014 had a legal marital status of married (Table 1). The next largest group was single, with one-third (33.9%) of people in it. The proportions of people divorced and widowed were similar, at 8.2% and 6.5% respectively. The smallest group was the civil partnered (0.2%); this is because this marital status is relatively new in England and Wales, having been introduced in December 2005.

People aged 16 to 19 were the most likely to be single (99.6% in 2014). Those aged 65 to 69 were the most likely of all the age groups to be married (71.8%) in 2014; this was closely followed by those aged 55 to 59 and 60 to 64 at 70.5% and 70.2% respectively. The oldest age group contained the largest proportion of widowed people (60.1% of people aged 85 and over were widowed). In contrast, the 55 to 59 age group had the largest percentage of divorced people (15.7%), closely followed by those aged 50 to 54 (15.5%) and 60 to 64 (15.2%).

Figure 1: Population estimates (age 16 and over) by marital status, age group and sex, 2014

England and Wales

Source: Office for National Statistics

Notes:

1. See the 'Definitions' section for details about each marital status category.
2. Married and civil partnered are combined in this chart due to the small numbers of civil partners.
3. Click the chart to enlarge.

Figure 1 illustrates the age and sex structure of the population of England and Wales by marital status. The length of each bar shows the size of the population in that age group. The coloured bands on each bar represent each marital status.

The number of people who are single (never married) is largest at the youngest ages. Those aged under 16 are all considered to be single in line with the legal age of marriage in England & Wales. Most men and women under the age of 25 remained single in 2014.

The number of single people reduces sharply for those in the 25 to 29 and 30 to 34 age groups as the size of the married and civil partnered population increases. This reflects those marrying or civil partnering for the first time in their late 20s and 30s; marriages data for 2011 shows the average (mean) age at first marriage was 32.2 years for men and 30.2 years for women.

The divorced population becomes visible on the pyramid for those in the 25 to 29 age group but remains small for those in their 20s and 30s, partly reflecting the relatively small size of the population at risk of divorce or dissolution (i.e. the married or civil partnered population). The numbers of people who are divorced increase in the 35 to 39 age group and above, with the largest number of divorced people in the 50 to 54 age group. There were larger numbers of divorced women than men at all ages over 20 in 2014.

Very small numbers of women aged under 45 and men aged under 50 are widowed, so the widowed population is only visible on the pyramid at older ages. The differences in marital status of men and women are most apparent at older ages. In 2014 there were more widowed women than men at older ages – for example at age 70 and over there were 1.75 million widowed women compared with 0.48 million widowed men. This difference is because women have higher life expectancy than men, and women have traditionally married an older spouse.

Among the total adult population (16 and over), the proportion of men who were single in 2014 (37.3%) was larger than the proportion of women (30.7%); this partly reflects the older average age at first marriage for men. Overall, similar proportions of men and women aged 16+ were married, as we would expect, while women are slightly more likely to be divorced and much more likely to be widowed than men. Part of the reason for this is simply the larger numbers of older women than older men in the population.

Table 2 shows that in England and Wales, 48.7% of people in 2014 were ‘living in a couple: married or civil partnered’. This is slightly lower than the percentage legally married or civil partnered in Table 1, because a small proportion of married people will be separated or not living together for other reasons such as one partner being in a care home or in prison.

The next largest group of people - one quarter of adults in England and Wales - were ‘not living in a couple: never married or civil partnered’ (24.9%). This living arrangement was most common among the 16 to 29 age group, where 71.6% were single and not living in a couple.

Around 1 in 8 adults in England and Wales were living in a couple but not currently married or civil partnered. Three-quarters of these cohabiting people were ‘never married or civil partnered’, while one-quarter were ‘previously married or civil partnered’. Living in a couple (never married or civil partnered) was most common at younger ages, with 22.2% of people aged 30 to 34 and 17.4% of people aged 16 to 29 having this living arrangement. This reflects the trend for young adults to cohabit before or instead of marrying or civil partnering.

In contrast, the proportions of people ‘cohabiting – previously married or civil partnered’ were highest (5.5%) in the 45 to 49 and 50 to 54 (6.0%) age groups. These are likely to be mainly divorced people who are living with a new partner and have not remarried.

In all age groups, people who were previously married or civil partnered were more likely to be ‘not living in a couple’ than ‘living in a couple’. Overall, 14.4% of adults were ‘not living in a couple: previously married or civil partnered’ in 2014. However this rose to 38.5% among people aged 70 and over, due to the high proportion widowed and small numbers who cohabit in this age group.

Figure 2: Percentage of people (age 16 and over) in 2014 by living arrangements and sex

England and Wales

Source: Office for National Statistics

Notes:

1. Click the chart to enlarge

Figure 2 shows similar proportions for men and women in all 3 types of ‘living in a couple’ arrangement. For example, 50.0% of men and 47.6% of women were ‘living in a couple: married or civil partnered’ in England and Wales in 2014. The proportions of men and women in couples will not match exactly, as these proportions are affected by differences in the numbers of men and women not in couples.

The main difference in living arrangements between men and women is among those not living in a couple. For women 18.9% were ‘not living in a couple: previously been married or civil partnered’ compared with only 9.8% of men. This is because of the larger numbers of women who are widowed than men, particularly at older ages. Conversely a higher proportion of men (27.9%) were never married and not living in a couple compared with women (22.1%); this was the case at all ages.

Figure 3 shows how marital status has changed between 2002 and 2014 in England and Wales. The percentage of people who are married has decreased over time, from 54.8% in 2002 to 51.2% in 2014, though stabilising over the last 3 years.

The proportion of people who are single has increased steadily over time, from 29.6% in 2002 reaching 33.9% in 2014. This coincides with an increase in people cohabiting who are never married or civil partnered (Figure 5) as cohabitation has become more common as an alternative to marriage at younger ages. The rise in singleness also reflects more people in middle age groups remaining unmarried.

Figure 4 shows that the percentage of people who are divorced has increased slightly over the time period; this is being driven by the population aged 45 and over, who have seen larger increases in the percentage divorced than younger age groups. For those aged 45 and over this could be partly explained by a reduction in remarriages so people are remaining divorced. It also coincides with the percentage of people who were widowed decreasing from 8.1% in 2002 to 6.5% in 2014; this is associated with continuing increases in life expectancy, particularly for men. This means that married people are living longer than before and so may divorce rather than become widowed.

The percentage of people who are civil partnered has increased from 0.06% of the population in 2006 to 0.25% of the population in 2014.

Figure 5 shows how living arrangements in England and Wales have changed between 2002 and 2014.

In 2002 people who were ‘cohabiting – never married or civil partnered’ represented 6.8% of the population; this had increased to 9.1% by 2014. This may be explained by an increasing trend to cohabit instead of marrying or before marriage particularly at younger ages. Cohabiting couple families are the fastest growing family type in the UK (2004 to 2014).

Focussing on people that have previously been married or civil partnered; the percentage of people who are cohabiting changed little between 2002 and 2014, remaining around 2.9% of the population aged 16 and over. This is also true of people who are ‘not cohabiting: previously married or civil partnered’ who remained at around 15% of the population aged 16 and over between 2002 and 2014.

In 2002, people who were ‘not cohabiting: never married or civil partnered’ represented 22.9% of the population. By 2014 this had increased to around a quarter of the population at 25.1%. This could be explained by younger people delaying entering cohabiting relationships or delaying leaving the family home.

International comparisons of marital status are not straightforward for a variety of reasons. First, the types of legal marital status available to residents vary across countries, particularly with regards to same sex unions and registered partnerships of various types. Second, some countries collect data on the separated or those with unknown marital status whereas others will combine these within other marital status categories. Third, percentages in different marital status groups are likely to be affected by countries’ age structures, given that marital status varies considerably by age.

Bearing these differences in mind, a crude comparison of the main marital status groups (Table 3) shows England and Wales sitting near the middle of the table of European countries. For example, 41% of people (of all ages) in England and Wales were married in 2012. This is lower than countries such as Romania with 50% married and Italy with 49% married but higher than Sweden and Iceland, where 34% were married in 2012. Similarly, the widowed population in 2012 varied from 4% or less in Turkey and Iceland to nearly 10% in Hungary and Lithuania. This compares with 5.5% in England and Wales, a similar level to Denmark and Finland.

The highest proportions of divorced people are found in Latvia (12% in 2012) and the Czech Republic (11%), with the lowest in Turkey (2%) and Romania (5%). Again England and Wales sits in the middle with 6.8% of people divorced.

These variations between countries reflect a range of factors including cultural differences in marriage and divorce behaviour, the prevalence of cohabitation, and differences in life expectancy which affect the proportion who are widowed.

Following the 2015 change to the methodology used to produce the population estimates by marital status, users will notice some differences when comparing the estimates for the years 2002 to 2010 using the old methodology and the new methodology. Here we summarise these differences. This section only applies to the population estimates by marital status, not the population estimates by living arrangements which are new.

For more information about the change in methodology see the; ‘Methods’ section of this bulletin, the Quality and Methodology Information (QMI) document or the Population Estimates by Marital Status and Living Arrangements information document.

A reference table (195 Kb Excel sheet) has been published alongside this bulletin which illustrates the differences between the estimates for 2002 to 2010 using the two different methodologies.

Differences in the estimates for 2002 to 2010 exist for two reasons;

• The data source and method used to construct the estimates have both changed

• The estimates for 2002 to 2010 produced using the old methodology have not been rebased following the 2011 Census; the estimates for 2002 to 2010 using the new methodology do take into account the rebased Mid-Year Population Estimates that incorporate the 2011 Census results

Separating the differences between the estimates using the old and new methodologies is complex. The following tables try to illustrate the effects that each of the two reasons has on the estimates but there will be some interaction between the two.

The following section of the bulletin focuses on the total population rather than just those aged 16 and over.

Table 4 shows the marital status distributions for the total population using the old method and the new method. Comparing the distributions allows some understanding of the effect that the new methodology has. The proportion of the population who are single is smaller using the new method compared with the old method. In 2010 the single population represented 47.6% of the population using the old method but only 45.7% of the population using the new method.

In contrast, the married population represents a larger percentage of the total population using the new method compared with the old method. For example, in 2007, 40.1% of the population were estimated to be married using the old method, while using the new method 42.5% were estimated to be married. Survey based estimates tend to show larger proportions of married people than other sources such as the Census because married families are more likely to respond and continue to respond across waves of surveys, compared with people living alone.

The percentage of the population who are divorced is slightly smaller using the new method, while the percentage widowed is consistent between the two methods. Population estimates for those in civil partnerships were not available using the old method. The majority of the population who are civil partnered would have been included in the single category in the old estimates; this reduces the gap between the old and the new ‘single’ estimates for 2006 onwards.

The first column in Table 5 shows the overall change in mid-year population estimates that is due to rebasing following the 2011 Census. For example in 2010 the population estimate for England and Wales was 452,000 higher following rebasing than previously. Because the 2011 Census showed the population to be higher than previously estimated, the new marital status estimate totals simply reflect these new population estimates, rather than the change in marital status data and methodology.

Looking at the differences between the old and new marital status estimates in the remainder of Table 5, the populations who were single or divorced in 2002 to 2010 were smaller using the new method compared with the old method, while the married population was larger using the new method. The widowed population fluctuated from being smaller in 2002,using the new method to being larger in the years between 2004 and 2010. These differences mainly reflect the differences in marital status distributions shown in Table 4.

There will also be some interaction between the effect of the population rebasing and the overall marital status distributions of the population using the old and new methods. This is because the size and direction of population rebasing following the 2011 Census varied by age and sex. For example the 2011 Census found more females aged 25-29 but fewer males aged 25-29 than had been previously estimated. These differences in the population age and sex structure following rebasing will also have some impact on the overall marital status distribution, given that marital status varies by age and sex.

ONS previously published data for 1971 to 2010 using a different methodology. We ran a consultation between 3 June 2014 and 11 July 2014 which outlined the new methodology to be used and the proposed new tables. The design of the new output was widely accepted by users and the response to the consultation was published. The differences in the data produced by the two methodologies are explained within this bulletin and within a reference table (195 Kb Excel sheet) alongside the release.

The new methodology uses estimated proportions of marital status and living arrangements from the Labour Force Survey (LFS) and applies these to the Annual Mid-Year Population Estimates. The LFS is a large household survey of people in the UK. The statistics used to produce the population estimates by marital status and living arrangements are derived from responses to the LFS in the April to June quarter of each year. The Quality and Methodology Information document (QMI) and the Population Estimates by Marital Status and Living Arrangements information document provide more detail on the new methodology.

We have updated the methodology for estimates relating to the year 2002 onwards. This is because of:

• societal changes, such as the introduction of civil partnerships in England and Wales in December 2005, which had not been able to be incorporated into the old method

• the rise in cohabitation either as a precursor to, or an alternative to marriage. Estimates of the cohabiting population were not possible using the previous method as there is no requirement to formally register cohabiting relationships

• concerns over quality of the survey estimates of the number of marriages occurring abroad and the quality of marital status information, including civil partnerships, collected from migrants

The methodology change also creates efficiencies in the time taken to produce and publish the Population Estimates by Marital Status and Living Arrangements. It will also make it easier to incorporate any new marital status or living arrangement categories into the estimates in future.

The estimates are of interest to government departments, local government, commercial companies, special interest groups, academia, and the general public.

Population Estimates by Marital Status and Living Arrangements have a number of uses; these include:

• informing policy and legislative decisions

• denominators in the calculation of rates, for example marriage and divorce rates published by ONS

• forecasting around pensions and insurance funds

Specific examples of the use of the Population Estimates by Marital Status and Living Arrangements include:

• relationship advice groups use the estimates to help estimate the costs of national policy recommendations and also to understand their client base and the products that they require

• several government departments require the estimates so they can provide advice surrounding the GB National Insurance Fund (NIF) and occupational pension schemes

• insurance and pension firms use the estimates to help with the estimation of pension liabilities

The Marriages (Same Sex Couples) Act 2013 made provision for the marriage of same sex couples in England and Wales from 29th March 2014 onwards. No married same sex couples were randomly selected in the interview sample in the April to June 2014 quarter of the Labour Force Survey (LFS). Therefore statistics on same sex married couples are not reported on within this publication.

In future, an annual estimate of same sex couples who are married, by sex, will be published when at least three people in same sex couples who are married have been surveyed in the LFS (to protect their privacy). Until sample sizes become large enough to produce robust estimates, data are unlikely to be available by age group. This will be reviewed annually.

Changes in the Population Estimates by Marital Status and Living Arrangements over time are not described as being statistically significant or not within this bulletin.

For users to gain an understanding of the level of accuracy of the Population Estimates by Marital Status and Living Arrangements it may be possible to produce confidence intervals around the survey estimate component only. ONS will consider producing confidence intervals for the survey component in the future depending on the user demand for these. Research into producing confidence intervals around population estimates is continuing; should these become available in the future, users would be able to gain an understanding of the overall accuracy of the Population Estimates by Marital Status and Living Arrangements rather than only the survey component.

Please contact us by emailing families@ons.gov.uk if you would like to express your interest in having confidence intervals around the survey estimate component of the Population Estimates by Marital Status and Living Arrangements.