Estimating the numbers of refugees in London

Journal of Public Health Medicine

Vol. 22, No. 3, pp. 406±412 Printed in Great Britain

Estimating the numbers of refugees in London Martin Bardsley and Marian Storkey

Abstract Background Local populations of refugees and asylum seekers are growing in many urban areas in Western Europe and it is increasingly important to develop health and welfare services that are appropriate for these groups. However, in the United Kingdom there are no routine data sources at local level that give even the most basic information such as the numbers of refugees within a given area. Methods The total number of asylum seekers entering the United Kingdom was calculated using cumulative data on asylum seekers at national level. This population was then apportioned ®rst to Greater London and then to the boroughs within London. The apportionment to London boroughs was based on analysis of four datasets. In the absence of any better evidence, an average of these four approaches was used to produce the ®nal borough level estimates. Results The total numbers of refugees and asylum seekers in London who have entered the United Kingdom over the past 15 years was estimated to be between 240 000 and 280 000. At borough level the estimates of refugee populations ranged from under 1000 to values up to 20 000. There were statistically signi®cant associations between the four data sources when the proportions of the London total in each borough were compared. However, for some boroughs there could be large differences between estimates based on different data sources. Conclusion The estimates provided give an indication of the size of the refugee population in London. None of the data sources used to apportion the London total were ideal and all were proxy values with their own strengths and weaknesses. This work points to the importance of developing information systems that in future will allow better estimates of the size of the refugee populations. This is particularly important in view of the UK national policy of dispersal proposed in the latest Immigration and Asylum Act. Keywords: refugees, asylum seekers, population estimates, service planning

Introduction Refugees and asylum seekers form signi®cant minority populations in many of our towns and cities. In 1997 there were nearly 300 000 applications for asylum in Western European countries;1 half of all applications were made in Germany, and the United Kingdom and The Netherlands together accounted for another quarter of all applications. Between 1983 and 1997, 396 000 people have applied for asylum in the United Kingdom. For many urban areas the need

to develop appropriate health and welfare services for refugees and asylum seekers is a matter of growing concern.2±4 However, estimates of the numbers of refugees and asylum seekers at a local level can still be dif®cult to obtain.2 Refugees and asylum seekers are not homogeneous and it is dif®cult to make generalizations across wide range of ethnic groups, age, class and history.5±7 Many of the health problems of refugees are not necessarily linked to refugee status but overlap with problems of deprived or excluded groups, ethnic minorities or new entrants to the country. However, a number of local studies have pointed to some characteristic health concerns amongst refugee communities.8±11 These include: (1) Physical after-effects of war, torture, the displacement period and journey into the United Kingdom.12,13 (2) Refugees may generally be at slightly greater risk from a range of communicable diseases,14 of which the most important condition from a public health perspective is tuberculosis.15,16 (3) Mental health problems that are the after-effects of war and torture can be signi®cant for some groups and individuals, and are seen as serious and prolonged problems following trauma.17,18 In addition to these there are more widespread social and psychological problems arising from coping with a new culture, loss of home, etc.19 This may also include problems that are a result of the uncertainty around the process of claiming asylum and settling in the United Kingdom. One of the key issues for health concerns accessibility to appropriate services and in particular the need to overcome barriers that may result from: different languages,7,10 poor quality information on what services are available20 and poor understanding by professional staff of refugees' rights to care.21 Only a small proportion of those applying for asylum in the United Kingdom are given refugee status. The umbrella term `refugees' used in this paper encompasses: asylum seekers, i.e. people who have applied for asylum but who have not yet had a Health of Londoners Project, East London and The City Health Authority, 81±91 Commercial Road, London E1 1RD. Martin Bardsley, Project Manager London Research Centre, 81 Black Prince Road, London SE1 7SZ. Marian Storkey, Principal Research Of®cer Address correspondence to Dr M. Bardsley. E-mail: [email protected]

q Faculty of Public Health Medicine 2000

E ST IMAT ING NU M B E R S OF RE FUGE E S IN L ONDON

®nal decision; people who have been granted refugee status; people granted exceptional leave to remain (ELR); people refused refugee status but who have not left the country It is extremely dif®cult to obtain demographic data on refugees and asylum seekers at a local level in the United Kingdom. The most basic requirement is for information on numbers of refugees and asylum seekers within an area, and ideally an indication of countries of origin, age and gender. The available data tend to be national ®gures, data gathered from small local surveys, or data from records of people using existing services. However, it is possible to combine data sources to build up a partial picture to help estimate the size of local refugee and asylum seeker communities.22 This paper discusses the approach taken to estimate the size of the population of refugees and asylum seekers in London.

Methods The approach used was based on two stages. The ®rst stage was to use national data to estimate the numbers in the whole of the United Kingdom from published Home Of®ce sources.1 The second step was to apportion a share of this population to Greater London and then to individual areas within London using a range of published statistics. All these published statistics represented approximations to the total population. To make them comparable, they were all expressed in terms of the percentage of the total London refugee population in each of the 33 local authorities (32 boroughs and the City of London).

Estimating total populations in London from national data The main data at a national level in the United Kingdom are the numbers published annually by the Home Of®ce's Immigration and Nationality Directorate.1 These provided values for the cumulative number of applications (including special asylum cases) between 1983 and 1997 that formed the basis for the model. From this pool was subtracted the numbers of people whose applications were withdrawn or who were removed from the country. Of those who are refused refugee status, only a small proportion is known to leave (those escorted by the Home Of®ce). It is not known how many of those refused stay in the country. This model assumes that between 25 and 50 per cent of those refused refugee status or ELR leave the country, and therefore that between 50 and 75 per cent stay in the country illegally. This is a critical assumption. Unfortunately, there are no objective data on this issue ± the values chosen are based on the views of people working with refugee communities and represent our `best guess'. One of the most comprehensive studies of people granted refugee status in the United Kingdom suggested that 85 per cent live in London23 and that two-thirds of all those identi®ed as seeking asylum in the United Kingdom had given a ®rst address in Inner London.

407

The following datasets were used to estimate the relative size of refugee populations in each borough.

Information from Social Services Since 1997 the Association of London Government has been collating statistics on asylum seekers receiving social service support under the terms of the National Assistance Act 1948 and the Children Act 1989. Although this is only a subset of the total refugee population, it does give an indication of the relative caseload in individual boroughs as recorded in three separate categories: families with children; single adults; unaccompanied children. The data are recorded as weekly caseload ®gures and an average between January 1998 and February 1999 was calculated. To estimate the total number across the three categories it was assumed that the average size of families with children was 3.3, on the basis of statistics recorded by the Home Of®ce. Bed and Breakfast Information Exchange The Bed and Breakfast Information Exchange (BABIE) is part of the Housing and Social Research Department of the London Research Centre and collects information on people in temporary accommodation. Two data sources were used to obtain a distribution of asylum seekers.24 The ®rst was the number of homeless households recognized under Housing Acts found to be asylum seekers in each year from 1990 to 1997. The second was the cumulative number of asylumseeking households in temporary accommodation including those under the Housing Acts of 1985 and 1996, the National Assistance Act 1948 and the Children Act 1989. An average of these two values was taken.

Estimates of refugee children in London The Refugee Council has estimated the numbers of refugee schoolchildren in Britain ± and in London in detail.25 The method used appears to have included a combination of approaches including refugee surveys in some schools, Local Authority statistics from language surveys cross referenced with Home Of®ce data, Census data and information from community groups.

Of®ce for National Statistics mid-year population estimates The Of®ce for National Statistics (ONS) uses estimated numbers of asylum seekers as a component in their mid-year population estimates.26 These are based on national ®gures on the numbers and country of birth of asylum seekers from the Home Of®ce, which are allocated to each district or borough in England and Wales. As comprehensive of®cial data are not available at a local level, a proxy variable is used to calculate the distribution. It is assumed that 85 per cent of asylum seekers come to London and that these asylum seekers are then distributed to London boroughs using the 1991 Census distribution of residents by country of birth. Individual

408

JOURNAL OF PUBLIC HEALTH MEDICINE

countries are not used, but the world is split into ten regions (excluding the Old Commonwealth, the European Community and the United States). The proportion of the London total in each borough was estimated using these four data sources. Analysis of variance on the percentages and rank values was used to test for consistency between the various methods.

Results The results in Table 1 show the cumulative estimates based on national asylum applications. They suggest that between 240 000 and 280 000 refugees arrived in London between 1983 and 1997. The London total estimate would be 320 000 if it were assumed that all of those refused asylum remained in the United Kingdom (rather than the range of 50±75 per cent assumed in the model). Alternatively, it would be 160 000 if it were assumed that all those refused left the country. Of the estimated 240 000±280 000 recent (i.e. last 15 years) refugees in London, just under one-half will have been granted of®cial refugee status (including special cases) or ELR. One-®fth will be awaiting a decision over their asylum application, and just over one-third are estimated to be those who have been refused status but have stayed in the United Kingdom. A little over onethird of the London total will have applied for asylum in the years 1995±1997. The distribution of refugees across London boroughs according to the four data sources is shown in Table 2. The information from the social services departments shows wide differences between boroughs in terms of absolute numbers and the proportion of the London total. On average, in any week during this period the social services caseload of asylum seekers in London included 8844 families with children, 15 136 single adults and 1919 unaccompanied children. The proportion of this total London caseload at borough level ranged from values below 1 per cent in Bexley, Havering, Richmond and Sutton, up to over 8 per cent in Haringey and 10 per cent in Newham.

Table 1 Estimating the cumulative population of refugees from asylum applications since 1983 in the United Kingdom Category

Numbers of refugees 1983±1997*

Total applications Applications withdrawn Cases removed Estimate 25±50% of refusals leave Special asylum cases Special SE Asian cases National total London (85% of national)

396000 20000 29000 47000 to 94000 5000 25000 282000±330000 240000±280000

*All ®gures rounded to nearest 1000. Source: Authors' model incorporating data from Carey-Wood et al.;23 Home Of®ce.1

The data on homeless asylum-seeking households and asylum-seeking households in temporary accommodation show a similar range of distribution across London. Three boroughs had less than 1 per cent of the distribution within London ± Havering, Bexley and Sutton. Boroughs with the highest numbers from distribution were Westminster, Islington and Camden. The Refugee Council data suggested that there were 38 544 refugee children in London. The spread across boroughs ranges from values close to zero in Havering, to over 10 per cent in Hammersmith. The distribution given by the ONS adjustments shows slightly less variation between extremes. The lowest share of the total was 0.5 per cent in Barking and Dagenham, whilst the highest was 6.6 per cent in Barnet. There were signi®cant correlations between the results using these methods, and the ranking of boroughs within London showed even more consistency. The results using the BABIE data and ONS were closest (correlation coef®cient of 0.695) and there was least consistency between the ONS data and the social services (r 0:376). When analysis of variance was used the amount of a variation explained by boroughs was around 66 per cent (F 5.6, p < 0.01). Although there is, in general, a signi®cant correlation between the results using the various methods, there may be a spread of values for any one borough, as Table 2 shows. Table 3 shows the estimated number of refugees and asylum seekers in each borough. These estimates suggest that the size of asylum seekers and refugee populations ranged from 500± 600 in Havering to between 15 000 and 20 000 in Newham, Haringey and Westminster. When expressed as a number of asylum seekers and refugees per 1000 residents, the highest values are around 70±80 per 1000 in Hackney, Haringey and Newham. The lowest values, in Bromley and Sutton, were below 10 refugees per 1000 residents.

Discussion The results are indicative of the problems local areas face when trying to estimate how many refugees live in their area. The value of 240 000±280 000 for London is very much an estimate and represents our best guess. Of course, by excluding refugees who arrived in the country more than 15 years ago, these ®gures may be lower than those that look at the population of people who have ever been refugees. However, these ®gures highlight the numbers of more recent refugees. It is likely that the speci®c needs of refugees will change with time in the United Kingdom. In terms of estimating the overall numbers in London, the most sensitive part of the calculation is the assumption that between 25 and 50 per cent of those refused refugee status or ELR leave the country, and therefore that between 50 and 75 per cent stay in the country illegally. This is a critical assumption. Unfortunately, there are no objective data on this issue ± the values chosen are based on the views of people working with refugee communities and represent our `best

409


Table 2 Percentage of the total refugee population in the London boroughs using four approaches Borough

Social services contact

BABIE

Refugee Council survey of children

ONS MYE data

Average of these four sources

Barking and Dagenham Barnet Bexley Brent Bromley Camden Croydon Ealing En®eld Greenwich Hackney Hammersmith Haringey Harrow Havering Hillingdon Hounslow Islington Kensington and Chelsea Kingston Lambeth Lewisham Merton Newham Redbridge Richmond Southwark Sutton Tower Hamlets Waltham Forest Wandsworth Westminster Total

2.1 2.7 0.5 4.0 0.9 4.5 2.7 2.2 2.3 3.3 4.4 3.8 8.4 1.1 0.2 2.0 2.5 4.9 5.1 0.7 5.5 4.2 1.4 10.4 2.0 0.9 5.5 0.4 1.1 3.4 2.2 4.5 100.0

2.7 2.6 0.4 4.9 1.1 7.7 3.7 4.9 2.8 2.0 4.9 3.9 5.4 1.9 0.2 2.2 1.6 6.6 4.1 1.0 5.5 3.6 1.4 4.1 1.6 1.4 3.3 0.6 1.3 1.7 2.1 8.5 100.0

0.5 5.4 1.3 7.3 0.2 3.8 1.9 5.9 6.2 3.8 10.2 1.9 6.5 1.8 0.0 1.1 2.8 2.2 2.3 0.4 0.4 5.4 3.6 9.5 2.3 0.3 3.3 0.3 2.4 2.3 1.8 3.1 100.0

0.5 6.6 0.8 6.1 1.6 4.6 3.0 6.3 2.7 1.6 4.1 3.2 4.7 2.9 0.5 1.7 3.0 2.6 5.5 1.5 4.2 2.2 2.3 3.8 2.3 2.0 3.2 1.0 2.8 2.2 3.9 6.5 100.0

1.4 4.3 0.7 5.6 0.9 5.1 2.8 4.8 3.5 2.7 5.9 3.2 6.2 1.9 0.2 1.7 2.5 4.1 4.2 0.9 3.9 3.9 2.2 7.0 2.0 1.1 3.8 0.6 1.9 2.4 2.5 5.7 100.0

guess'. Similarly, this model assumes that all of those successfully applying for asylum in the last 15 years stayed in the United Kingdom and that none have died or returned home. The model does not take into account any migrants who have entered the country illegally and who have not applied for asylum. Although there will be a number of illegal migrants, the total is likely to be small given the United Kingdom's strict entry requirements and the geography of the islands (making entry more dif®cult). The ®gure of 85 per cent of asylum seekers settling in London was based on the analysis by Carey-Wood. Although this study is 3±4 years old, there is no reason to suggest that the high proportion of UK refugees living in London has changed. A more recent Refugee Council study of refugee schoolchildren27 calculated that 88 per cent of all refugee children in Britain were attending schools in London. Recent analysis of social security data (Department of Social Security, personal communication, 1999) looking at a category of urgent case payments (for which the majority are believed to relate to asylum seekers) suggested that 91 per cent of the asylum seeker and other urgent case payment

recipients in Britain lived in London. These two other sources suggest that the 85 per cent ®gure is realistic. When it comes to distributing this population to areas within London, the various data sources all have their strengths and weaknesses. None of them exactly corresponds to the measure we want and all are considered a proxy for total refugee distributions. The data from social services are based on average numbers in contact with social services per week. There may be variability in completeness between social service departments in identifying, classifying and recording cases, and in the time that asylum seekers stay on the caseload. It is increasingly common for some asylum seekers to be placed in temporary accommodation outside of the borough, and more recently, outside of London. The Refugee Council survey is based on the numbers of children, who may be distributed differently from adults. National statistics indicate that 68 per cent of asylum seekers are single people. It may be the case, for example, that this method allocates more asylum seekers to areas where more family housing is available.

410


Table 3 Borough estimates of numbers of refugees (entering United Kingdom since 1983) Borough

Estimated number of asylum seekers and refugees

Asylum seekers and refugees per 1000 residents

City of London Barking and Dagenham Barnet Bexley Brent Bromley Camden Croydon Ealing En®eld Greenwich Hackney Hammersmith Haringey Harrow Havering Hillingdon Hounslow Islington Kensington and Chelsea Kingston Lambeth Lewisham Merton Newham Redbridge Richmond Southwark Sutton Tower Hamlets Waltham Forest Wandsworth Westminster Total

370±430 3500±4100 10400±12100 1800±2100 13400±15600 2200±2600 12300±14400 6800±7900 11600±13500 8400±9800 6400±7500 14200±16500 7700±8900 15000±17500 4700±5400 500±600 4200±4900 6000±6900 9800±11500 10100±11800 2200±2500 9400±11000 9300±10800 5300±6200 16700±19500 4900±5700 2800±3200 9200±10700 1300±1500 4600±5300 5800±6700 6000±7000 13600±15900 240000±280000

76±76 23±27 32±37 8±10 54±63 7±9 65±76 20±24 39±45 32±37 30±35 74±86 49±57 68±80 22±26 2±3 17±20 29±33 55±65 62±72 15±17 35±41 39±45 29±34 73±85 21±25 15±17 40±46 7±8 26±29 26±31 23±26 64±75 34±39

The information from BABIE is based on the areas where people are located, but it applies to only a relatively small subset of the total refugee population. It is perhaps more indicative of the distribution over recent years. The numbers in temporary accommodation will be in¯uenced by the amount of temporary accommodation within each area, as well as the number of asylum-seeking households needing that accommodation. The basis of the ONS approach is that asylum seekers live in areas where there are other members of their community. However, in practice there are two key problems. First, asylum seekers are allocated using the 1991 Census country of birth data distributions, not actual distributions of asylum seekers. Second, very broad geographical areas of the world are used so that, for example, Somali asylum seekers will be allocated in the statistics to boroughs that have a large non-Commonwealth population born in Africa, and Bosnian asylum seekers will be allocated in proportion to their populations from Eastern Europe and the former USSR. Although the ONS acknowledges the de®ciencies with the

data, there are no better comparable local data with which this exercise can be undertaken. These totals feed into the government's of®cial population estimates for England and Wales, and have important resource implications where central funding allocations are determined by the resident population. When the results of these methods are compared with some local-based studies, typically amongst refugee communities, then the estimates here may appear low. However, one of the problems of these local studies is that they are based on communities that may not ®t administrative boundaries, or the methods may lead to some double counting. There are also other data sources that may be available at local level that we could not access across all London. For example, in education services, some areas collate information from individual schools. Although the results using these methods are not entirely consistent, together they do point to the uneven distribution of asylum seekers across boroughs and to the higher concentration of asylum seekers in Inner London. The percentage of refugees


in Inner London closely matches the 66 per cent found by Carey-Wood et al.23 Where there are signi®cant differences between the methods we have no reasons for choosing one over another, hence the rather crude option of simply taking an average. Unfortunately, we do not have enough information to look at the distribution of a nationality of origin. Yet we know that the pattern of settlement by nationality is not even across London, and a study of the health authorities of ®rst address for a subset of people passing through Heathrow Health Control Unit shows that some nationalities tend to settle in one area, others elsewhere.2 In fact, we would expect newly arriving refugees to be drawn towards existing communities from the same parts of the world. Earlier experience looking at the national programmes suggests that this happens.28 This has important implications for the new asylum legalization in the United Kingdom,29 which is based on a policy of dispersal away from London and south-east England. One of the important lessons of past experience is the importance of community groups to help refugees to settle in the United Kingdom.30,31 The estimate of the size of the refugee population in London of between 240 000 and 280 000, based on admission to the United Kingdom in the past 15 years, gives an indication of the scale of populations. The absence of even basic information about the numbers of asylum seekers and refugees at a local level is one of the handicaps to developing services that are accessible for these groups. The new legislation in the United Kingdom suggests new arrangements for asylum seekers. It is important that these changes are used to improve our understanding of some of the basic questions about local refugee communities. A better understanding of the size and nature of these population is essential to better planning and provision of the appropriate services to one of the most disadvantaged groups in our community.

Acknowledgements

411

5 Kemp C. Health services for refugees in countries of second asylum. Int Nurs Rev 1993; 40: 21±24. 6 Clinton-Davis, Lord, Fassil Y. Health and social problems of refugees. Social Sci Med 1992; 35: 507±513. 7 Gammell H, Ndahiro A, Nicholas N, Windsor J. Refugees (political asylum seekers): service provision and access to the NHS. London: College of Health, 1993. 8 Brent and Harrow Health Agency, Brent and Harrow Refugee Groups, Northwest London Training and Enterprise Council. Brent and Harrow refugee survey. Harrow: Brent and Harrow Health Agency, 1995. 9 Bariso EU. The Horn of Africa health research project: an assessment of the accessibility and appropriateness of health care services to the Horn of Africa (Eritrean, Ethiopian, Oromo and Somali) communities in Camden and Islington. London: Healthy Islington, 1997. 10 Haringey Council. Refugees and asylum seekers in Haringey. Research project report. London: Haringey Council, 1997. 11 Hendessi M, Bank-Anthony B. Health needs and experiences of refugees in Croydon. Report of a research project. Croydon: Linx Social Policy Consultancy, 1997. 12 Peel MR. Effects on asylum seekers of ill treatment in Zaire. Br Med J 1996; 312: 293±294. 13 Gorst-Unsworth C, Goldenberg E. Psychological sequelae of torture and organised violence suffered by refugees from Iraq. Trauma-related factors compared with social factors in exile. Br J Psychiat 1998; 172: 90±94. 14 Dick B. Diseases of refugees ± causes, effects and control. Trans R Soc Trop Med Hyg 1984; 78: 734±741. 15 Gushulal BD. Tuberculosis control in refugees ± policy and practices. Int J Tubercul Lung Dis 1998; 2: S87±S93. 16 Hayward A. Tuberculosis control in London. The need for change. A report for the Thames regional directors of public health. A discussion document. London: NHSE, 1998. 17 Goldstein RD, Wampler NS, Wise PH. War experiences and distress of Bosnian children. Pediatrics 1997; 100: 873±878. 18 Silove D, Sinnerbrink I, Field A, Manicavasagar V, Steel Z. Anxiety, depression and PTSD in asylum seekers: associations with premigration trauma and post-migration stressors. Br J Psychiat 1997; 170: 351±357.

The authors would like to thank the Home Of®ce, the Refugee Council, The Association of London Government and LB Hammersmith and Fulham, London Research Centre and the Of®ce for National Statistics for supplying data, and the Health of Londoners Task Group for discussing results.

19 Baker C, Arseneault AM, Gallant G. Resettlement without the support of an ethnocultural community. J Adv Nurs 1994; 20: 1064±1072.

References

22 Wassbein C. Report of refugee populations in Lambeth, Southwark and Lewisham. Refugee Outreach Team. London: Lambeth, Southwark and Lewisham Health Authority, 1998.

1 Home Of®ce. Asylum statistics 1997. London: Home Of®ce, 1998. 2 Aldous J, Bardsley M, Daniell R, et al. Refugee health in London. Key issues for public health. London: Health of Londoners Project, 1999. 3 Health Education Authority's Expert Working Group. Promoting the health of refugees. London: Immigration Law Practitioners Association, 1998. 4 Johnson MRD, Akinwolere OA. Refugees and primary health care in the West Midlands. Research Paper in Ethnic Relations No. 25. Coventry: University of Warwick, 1997.

20 Rutter J. Refugee children in the classroom. Stoke-on-Trent: Trentham, 1994. 21 Karmi G. Refugees. In: Rawaf S, Bahl V, eds. Assessing health needs of people from minority ethnic groups. London: Royal College of Physicians, 1998.

23 Carey-Wood J, Duke K, Karn V, Marshall T. The settlement of refugees in Britain. Home Of®ce research study No. 141. London: HMSO, 1995. 24 London Research Centre. Homelessness in London bulletin No. 14. February 1999. London: London Research Centre, 1999. 25 Refugee Council. Section 11 and refugees. A policy paper on how Section 11 funding should be used to meet educational needs for refugee children. London: Refugee Council, 1997.

412


26 London Research Centre. ONS mid-year estimates 1997. Population advice note 98-3. London: London Research Centre, 1998.

30 Compass Partnership. The Bosnia project: stage 2 evaluation. London: Compass Partnership, 1997.

27 Rutter J. Section 11 and refugees. London: Refugee Council, 1997.

31 Carey-Wood J. Meeting refugees' needs in Britain: the role of refugeespeci®c initiatives. London: HMSO, 1997.

28 McFarland E, Walsh D. Refugees in Strathclyde: the challenge of resettlement. Crit Social Pol 1989; 9: 101±106. 29 Home Of®ce. Fairer, faster and ®rmer ± a modern approach to immigration and asylum. Cm 4018. London: HMSO, 1998.

Accepted on 3 December 1999