Child and adult hands

Home / publications / briefings /

Non-European Migration to the UK: Family Unification & Dependents

03 Jul 2018

This briefing examines data on non-EU/EEA migration to the UK of people whose basis for entry to the UK is through ties to a family member. Family migration includes what we identify as ‘family unification’ migration (those coming to join family members already living in Britain as citizens or settled residents) and ‘dependents’ who accompany or join a migrant with only temporary leave to remain in the UK.

  1. Key Points
    • Non-EU family migration has declined since the mid-2000s
    • Asia remains the most common region of origin for non-EEA family migrants to the UK
    • A majority of non-EEA family unification migrants are women
    • More than nine in ten non-EEA migrants with permission to enter the UK as spouses or fiancé(e)s are spouses (i.e. already married before they arrive)
    • Tier 1 and Tier 2 migrants bring the most dependents, per capita
  1. Understanding the Evidence

    Alongside work, study, and asylum, family migration is one of the four main categories of non-European migration to the UK, though there are multiple measures and definitions of it. In the most expansive definition, family migration includes anyone coming to the UK for the stated purpose of joining or accompanying a family member. Data from the Office of National Statistics, including the International Passenger Survey (IPS) and Long-Term International Migration (LTIM) estimates, are based on this broad definition. However different data sources show different levels of family migration. Some reasons for these discrepancies are well-known. For example, IPS data are limited to people planning to stay in the UK for a minimum of one year, but border admissions data and visa data are not.

    In data based on non-EEA passengers entering the UK or entry clearance visas issued to non-EEA nationals, finer distinctions can be made among different sets of family migrants.

    First, some migrants gain entry to the UK because they are the family member of a British citizen, or of a non- EEA national with indefinite leave to remain in the UK (i.e. a ‘settled’ migrant). These may be designated as “family unification” (or ‘family formation and reunification’) migrants.

    Second, other migrants, called ‘dependents’, gain entry to the UK as family members of people who are themselves migrants to the UK with limited (temporary) leave to remain.

    Third, some non-EEA migrants enter the UK as family members (almost always spouses or partners) of EEA nationals. Such individuals may obtain EEA family permits, but are not required to do so, and they may or may not be EEA nationals themselves. This briefing does not generally include EEA family permits in counts of non-EU family migration.

Non-EU family migration has increased less than other categories

According to IPS estimates, non-EU family migration to the UK increased from an average of 35,000 per year in the 1990s to 53,000 in 2016, or 24% of all non-EU immigration that year. These estimates include both dependents and family unification migrants. Family migration, like overall migration to the UK, increased from 1997 to the mid- 2000s, peaking at 74,000 in 2006 (see figure 1). Also similar to other categories of migration, family migration declined in the second half of the 2000s. These shifts in family migration were smaller in magnitude than similar shifts in migration for work. As a result of these trends, family migration comprises a smaller share of overall migration now than it did in the 1990s (see our briefing on Immigration by Category: Workers, Students, Family Members, Asylum Applicants).

Although this briefing focuses on non-EU migration, it is worth noting that, including EU nationals, total family migration is now at similar levels to the 1990s. ONS estimates of family migration were 90,000 in 1991 and averaged 75,000 during the 1990s. ONS estimates had increased to 105,000 by 2006 but fell to 83,000 in 2016.

Figure 1

Similar to IPS data, Home Office data on visas issued also show a decline in non-EEA family migration in the late 2000s. IPS data show non-EU family migration has fluctuated around 50,000 annually from 2008 onwards, while visa data suggest a slight decrease over the same period. Because the data are collected differently and the IPS is a survey estimate with margins of error, such discrepancies between data sources in the short term are not unusual. Home Office data show much larger numbers than IPS estimates. The inclusion of short-term migrants in the Home Office data, as well as people who were issued visas but did not actually make the trip to the UK, are one reason behind this difference.

Return to top

Work dependents outnumber family unification migrants

Figure 2 shows that border admissions of those admitted as spouses, partners, or children declined from 53,300 in 2006 to 37,400 in 2010 and declined further to 20,300 in 2016. Dependents of work permit holders (including, since 2008, labour migrants in the Points-Based System) increased from 13,400 in 1990 to a peak of 49,000 in 2006 before levelling off around 33,000 annually from 2008 onwards. Border admissions data on student dependents show a decrease from 23,300 in 2010 to 11,500 in 2012. The number of asylum seekers’ dependents decreased from 20,600 in 2001 to 6,700 in 2004, then fluctuated between 4,700 and 6,200 from 2005 to 2013 and increased to 8,610 in 2016.

Figure 2

Return to top

Asia most common region of origin

Passenger entry data, compiled from samples of landing cards submitted at ports of entry, provide further information on the components of family and dependent migration. Among the regions used in Home Office classifications, Asia is the source of more family migrants and dependents than any other area, as shown in Figure 3. In 2016, according to border admissions data, 37,900 family migrants and dependents were from Asia (classified by nationality rather than country of birth), or 62% of the total. This included 26,000 people admitted as dependents of migrants and 12,000 admitted as spouses, partners, and children of British citizens or migrants previously granted settlement.

Figure 3

Asia has been the most common region of origin for family migrants since at least 2004, as shown in figure 4 (which excludes dependents and migrants granted settlement on arrival). Family unification migration from Asia has dropped from 30,400 in 2007 to 11,900 in 2016, which is below 2004 levels.

Figure 4

Return to top

Women outnumber men, spouses outnumber fiancé(e)s, adults outnumber children

Examining family migration by relationship type, again using passenger entry data, reveals several key points. First, women outnumber men among family migrants, in contrast with labour migrants and student migrants (Figure 5). There are more wives than husbands entering the UK for family unification. Women comprised 75% of migrants admitted as spouses or fiancé(e)s in 2016, and at least 60% in every year since 1997.

Among admitted marital partners, 94% of 2016 arrivals were spouses compared to 6% fiancé(e)s. The percentage who were spouses remained between 87% and 94% in each year since 1997.)

In addition, children comprise an increasing share of family unification migration, increasing from 7% (1,890 children) in 1997 to 11% (2,240) in 2016. Migrants entering as civil partners, unmarried partners, and proposed civil partners were 1,560 (8% of family unification migration) in 2016.

Unfortunately, passenger entry data do not clearly distinguish people admitted as parents, grandparents, and other relatives of UK citizens or residents. A Home Office (2011) Occasional Paper disaggregated family migrant visas granted in 2010, showing that 2,665 visas were granted to adult and elderly dependents, or 5% of family route visas. Passenger entry data also do not show how many spouses or fiancé(e)s are (to be) married to British citizens, as opposed to migrants with indefinite leave to remain in the UK. Again, this information is available for grants of settlement since 2006, and shows at least 69% of such settlement grants in 2016 went to migrants who married British citizens.

Data do not show how many of those British citizens were foreign-born themselves. The Home Office (2011) examined a selection of 531 records of migration through marriage in 2009 among applicants from selected countries to analyse this question, although the data have limitations: samples are small for each country and not all the samples were randomly selected. For those that were randomly selected, according to Migration Observatory correspondence with the Home Office, the data show variation in the share of sponsors who were themselves foreign born, from 11% of people sponsoring American spouses to 43% of those sponsoring Indian spouses and 67% of those sponsoring Bangladeshi spouses.

Figure 5

The composition of family unification migration varies with migrants’ nationality, as shown in Figure 6 (gender proportions) and Figure 7 (proportion of marital partner migrants who arrive in the UK as already-married spouses). The percentage of arriving spouses and partners who are female is highest among Middle Eastern and non-EEA European nationals. The share of women is lowest among nationals of countries in Oceania and Africa.

Figure 6

Migrants from different regions also show differences in the ratio of spouses to fiancé(e)s – or, in other words, the share of marriage-related migrants who are already married when migrating to the UK, compared with those coming to the UK to get married. Comparing regions of origin, the share of already-married partners has consistently been highest for migrants from Africa and Asia, at respectively 96% and 97% in 2016. The lowest percentages of spouses, meanwhile, are found among migrants from Oceania and the Americas both around 85%.

Figure 7

Return to top

Tiers 1 and 2 migrants bring most dependents, per capita

Shifting from family unification of settled residents and citizens to dependents of people with limited leave to remain in the UK, available data show the number of dependents attached to each main category of migration (work, study, family, and asylum), and to each subcategory with the Tiers of the PBS. These data show which groups of migrants more often bring family members with them, often expressed as a ratio of dependents to main applicants.

As seen in Figure 8, Tier 1 and Tier 2 work migrants bring more dependents per main applicant with them than other groups. For every visa granted to main applicants in the five year period from 2012 to 2017, there were 2.2 dependent visas in Tier 1 and 0.7 dependent visas in Tier 2. Ratios were lower in other categories: 0.1 dependents per student, 0.05 per Tier 5 temporary worker, and 0.3 dependents per asylum applicant.

Figure 8

Return to top

High likelihood of settling

Family and dependents comprise a larger share of settlements than of initial in-flows of migrants to the UK. A Home Office study found that 83% of family migrants entering the UK in 2010 had been granted settlement by the end of 2015, while 13% no longer had valid leave to remain in the UK (“Statistics on changes in migrants’ visa and leave status”, 2018) (formerly known as “The Migrant Journey”)).

Dependents are less likely to settle than family unification migrants, since their stay in the UK depends on the length of stay of the “main” migrant in their family. Among main applicants, rates of settlement are lower in the work and study routes than in the family route, which should apply to dependents in those paths as well, since their ability to settle depends on the main applicant in the family being granted settlement. Additional details, including the only available data on parents, grandparents, and other family members, can be found in the Migration Observatory briefing on Settlement in the UK.

Return to top

Evidence gaps and limitations

There is no single clear way to measure family migration and thus estimates of family migration vary not only across different data sources, but also across different definitions of family migration. Although the differences these measures and sources are well known, issues still arise with some datasets.

For example IPS data are from a sample survey that is subject to a margin of error. Also, the IPS classifies family migrants based on migrants’ own self-reports, rather than the basis of their legal right to enter the UK. The LTIM estimates used in this briefing have been only partially revised by the ONS. In their ‘Quality of Long-Term International Migration estimates from 2001 to 2011’ report published on 10th April 2014, the ONS revised the total net migration estimates for 2001-2011; this suggests that the total net migration between 2001 and 2011 was underestimated by 346,000 net migrants. This revision is based on evidence of underestimation by the LTIM, especially of migration from the A8 Eastern European countries. However, a revised version of inflows and outflows as well as breakdowns by citizenship or reason for migration, etc., is not available.

Similarly Home Office data’s inclusion of short-term migrants as well as those given leaves to enter, but who did not arrive into the UK, causes discrepancies between IPS. Similarly most Home Office data gathered through control of immigration include only non-EEA/Swiss nationals, whilst the IPS collects data on migrants of all nationalities, dividing them between EU and non-EU, but this has been accounted for in this briefing by focussing on Non-EEA. The Admissions data used in this briefing do not include statistics on parents and grandparents, though this is available from the Home Office’s Settlement data.


  • Home Office. “Statistics on changes in migrants’ visa and leave status.” Home Office, London, 2018
  • Home Office. “Family Migration: evidence and analysis, 2nd ed.” Occasional Paper 94, Home Office, London, 2011

Further readings

Related material

Next Update

June 24, 2019


Dr Scott Blinder

Press Contact

If you would like to make a press enquiry, please contact:

Rob McNeil

+ 44 (0)7500 970081

 Contact Us 


This Migration Observatory is kindly supported by the following organisations.

  • University of Oxford logo
  • COMPAS logo
  • Esmee Fairbairn logo
  • Barrow Cadbury logo
  • Paul Hamlyn Foundation logo