Urban Profiling of Refugees Situations in Delhi
Urban Profiling of Refugees Situations in Delhi
Urban Profiling of Refugees Situations in Delhi
SEPTEMBER 2013
Joint IDP Profiling Service (JIPS) [email protected] 23 Avenue de France CH-1202 Geneva
twitter.com/JIPS_profiling facebook.com/JointIDPProfilingService
The full report and the profiling tools, including a tool for further analysis, can also be found at: http://www.jips.org/en/field-support/country-operations/india/india/delhi-urban-profiling Printed in Switzerland in September 2013
ACKNOWLEDGEMENTS
We are particularly grateful to the following for their excellent work and contribution to the profiling process
This report is a collaborative effort between the Joint IDP Profiling Service, the Feinstein International Centre (Tufts University) and UNHCR. The study was fully funded by the US Department of States Bureau of Population, Refugees and Migration. The profiling exercise was conducted in close collaboration with our implementing partner in Delhi the Development and Justice initiative (DAJI). We are particularly grateful to the following for their excellent work and contribution to the profiling process: UNHCR in Delhi and their implementing partners Bosco and SLIC; Ravi Hemadri, Ipshita Sengupta, Madhuri Sastry and Subodh Singh; the team of survey enumerators and focus groups facilitators; and all the survey respondents who generously contributed their thoughts and time. CartONG provided valuable support throughout the exercise on the mobile data collection process. The profiling process, including methodology design, training, data collection, analysis and report drafting was led by a team consisting of: Karen Jacobsen, Margharita Lundkvist-Houndoumadi, Eric Levron and Ivan Cardona.
TABLE OF CONTENTS
EXECUTIVE SUMMARY I. INTRODUCTION II. BACKGROUND AND CONTEXT REFUGEES IN INDIA III. METHODOLOGY IV. DEMOGRAPHIC AND HOUSEHOLD CHARACTERISTICS V. LIVELIHOODS OF REFUGEES IN DELHI
1. Employment security 2. Housing security 3. Financial security 4. Physical safety and experiences of discrimination and harassment
4 7 10 14 20 22
25 31 34 38
40
42 43 50 58 59
60
62 62 63 68
69 75
EXECUTIVE SUMMARY
EXECUTIVE SUMMARY
UNHCR currently assists over 24,000 urban refugees and asylum-seekers in New Delhi originating from non-neighbouring countries and Myanmar.
This work has enabled UNHCR to build up a strong understanding of the challenges facing these groups. This profiling exercise of Afghan, Somali and Myanmarese refugee households alongside their local Indian neighbours was designed to supplement existing knowledge with a new base of empirical evidence. This will contribute to effective programming and advocacy, in particular in relation to UNHCRs work on promoting self-reliance. Conducted between January and June 2013, the profiling study examined the key components of refugee livelihoods, namely employment, housing, and financial security, as well as physical safety. The study was based on mixed research methods, namely a survey of over one thousand households as well as a series of focus group discussions and key informant interviews. The analysis of the collected data has allowed us to build a valuable picture of the livelihood security of Afghan, Somali and Myanmarese refugees vis-vis their local Indian neighbours. Within the report we use our findings to examine the contributing and impeding factors to refugee groups securing a livelihood, looking at their legal status, social capital, human capital, and length of stay in Delhi. A particular focus was given to access to education. The study has been carried out through collaboration between UNHCR, the Joint IDP Profiling Service (JIPS) and the Feinstein International Centre (Tufts University), alongside our partner in Delhi, the Development and Justice Initiative (DAJI). In June 2013 our preliminary findings were presented at a workshop in Delhi, where recommendations were developed in close collaboration with UNHCR and their implementing partners, Bosco and SLIC.
EXECUTIVE SUMMARY
2. HOUSING SECURITY: Myanmarese, Somali and Afghan refugees all reported varying degrees of housing insecurity due to restricted access to accommodation, discrimination by landlords, and evictions. Primarily it was Myanmarese and Somali households who had experienced evictions and had less housing space. As a result of these findings we believe that BOSCO and SLIC should conduct sensitisation sessions with landlords, neighbours and neighbourhood associations in areas where refugees live to minimise and prevent conflicts. Particular focus should be given to areas with high concentrations of Somali and Myanmarese refugees.
3. PHYSICAL SAFETY, HARASSMENT AND DISCRIMINATION: Refugees from Myanmar and Somalia reported facing the most discrimination and harassment in the neighbourhood, by landlords, at work, in school, and by local authorities - and had the worst relations to the local communities. We found that they were counterbalancing this by strong intra-community support networks. Afghan refugee households experienced less discrimination and far fewer perceived their neighbourhood as unsafe. At the same time, they reported having less intracommunity networks. We recommend that UNHCR and its partners continue to bring together refugees and local Indian communities during cultural events to improve communication. On-going support of the formation of refugee community groups will support the strengthening of intra-community networks, as will the identification of key community members from the Indian community, who can facilitate relationship building with refugees. Finally, initiatives by UNHCR and its partners to bring together Indian women with women from refugee groups, and the establishment of joint youth workshops with Indians and refugees will further help to bridge gaps and increase communication.
4. EDUCATION : Adequate access to all levels of education, including higher education, was a concern for all the refugee groups included in the profiling. Myanmarese children were, however, in greater numbers than Somali and Afghan children, not attending any type of school. Harassment and discrimination was identified as a problem in government schools primarily by Myanmarese and Somali children. We recommend that UNHCR and BOSCO should continue to identify government schools in areas where refugees live, to advocate for admission of refugee students, and to strengthen engagement with local government schools and local communities to raise awareness about refugees. Distance learning opportunities should be combined with income generation activities for children unable to regularly study as a result of work commitments, a challenge reported primarily by the Myanmarese refugees. Furthermore, UNHCR advocacy with higher educational institutions could facilitate the admission of refugees at equal fees to Indian nationals, a topic which was brought up across all refugee communities. Some refugees possess higher education but no job as their certificates are not recognised. Their knowledge could be used by engaging them in the classes provided by BOSCO, in order to expand the subjects offered and support knowledge transfer and network building between refugee communities. Additional measures should include UNHCR and BOSCO offering vocational training and job placement based on the skills in high demand, such as car repair, IT repair and mobile repair, and continue to promote language courses in Hindi among refugee children, youth and adults.
I. INTRODUCTION
I. INTRODUCTION
In India, as of July 2013, UNHCR assists over 24,000 urban refugees and asylumseekers from nonneighbouring countries and Myanmar 1.
In India, as of July 2013, UNHCR assists over 24,000 urban refugees and asylum-seekers from non-neighbouring countries and Myanmar . Most of these refugees live in the National Capital Territory of Delhi, and comprise 0.001 percent of Delhis total population of about 23 million. In 2012 we decided to conduct a profiling exercise in Delhi to analyse different aspects of these refugees living situations, with the aim of drawing comparisons between the experiences of different refugee groups and those of Indian nationals living in the same neighbourhoods. This profiling exercise set out to identify specific areas that could be targeted for programming and advocacy. The objective to inform UNHCR programming meant the profiling exercise focused only on those refugee groups that were of concern to UNHCR. The profiling exercise in Delhi was also intended to develop a more systematic approach to urban profiling for the benefit of the wider humanitarian and development communities. The profiling exercise took place between January and June 2013 and was a collaborative project between UNHCR, the Joint IDP Profiling Service (JIPS) and the Feinstein International Centre (Tufts University). The data collection was conducted together with our implementing partner the Development and Justice initiative (DAJI). The project was funded by the US State Departments Bureau of Population, Refugees, and Migration. An important aspect of JIPS profiling philosophy is to work through a collaborative process both with respect to the development of study objectives and the dissemination of the findings. Such a collaborative process meant engaging with all stakeholders at an early stage, in order to agree on objectives, target groups, dissemination strategy and final recommendations. Through this collaborative process, we decided to focus on refugees from Myanmar, Afghanistan and Somalia. Among the Afghans, we decided to exclude Hindu-Sikh Afghans, most of who had come to Delhi more than twenty years earlier and were much better integrated into Delhi than non-Hindu Sikh Afghans. Our primary aim was to explore the livelihoods of the different refugee groups, in order to inform UNHCRs work on promoting selfreliance. We therefore agreed to collect the following information from the targeted refugee groups, as well as from a subsample of their Indian neighbours:
Demographic and household characteristics, such as age, sex, ethnicity and household composition. Migration patterns. Livelihood information related to employment security, financial security, housing security and physical safety. Human and social capital, such as education and skills, health, local networks and relations with other communities, particularly the Indian host population . In addition, we were interested in the access of refugee children and youth to education and schooling facilities in Delhi.
I. INTRODUCTION
This information was gathered using mixed research methods (see chapter III on Methodology). The analysis of the collected information, which is presented in this report, focuses on answering the following questions : How do livelihood situations differ between the refugee populations and the Indian neighbours ? What factors contribute to or impede the livelihood security of the different refugee populations ? To answer this question we explored the role of the human and social capital in each community. What is the relationship between Indians and the refugees ? What programming and advocacy recommendations can be derived from our findings ?
The report is organized as follows: After a presentation of the methodology, a background chapter sets out the refugee context in Delhi in terms of the legal framework of refugee rights in India and UNHCRs main programming priorities in Delhi. The report is then organised around four thematic chapters : Chapter IV outlines the demographic and household characteristics of the surveyed populations. Chapter V compares livelihood situations across the targeted groups. Chapter VI analyses the factors contributing to or impeding livelihood security. Chapter VII explores the education of refugee children and youth. Chapter VIII summarises the main findings and provides recommendations. Annex 1 includes all the data used throughout the report.
REFUGEES IN INDIA
10
II.
India registers and provides direct assistance to around 200,000 Tibetan and Sri Lankan refugees from neighbouring countries.
In the absence of a national legal framework for asylum, UNHCR registers, issues documentation to, determines the refugee status of, and assists over 24,000 urban refugees and asylum-seekers from nonneighbouring countries and Myanmar.
11
India is not a signatory to the 1951 UN Convention Relating to the Status of Refugees or its 1967 Protocol; neither does the country have a national legal framework for asylum.
Instead India has developed ad hoc policies to provide sanctuary to Tibetans and Sri Lankan Tamil refugees. As far as refugee status determination of asylum seekers from elsewhere is concerned, India is generally respectful of the decisions of UNHCR. At the same time India is a signatory to a majority of the international instruments that have a direct bearing on refugee rights and protection. These include The International Covenant on Civil and Political Rights and The International Covenant on Economic, Social and Cultural Rights, 1966, The Convention on Rights of the Child, 1989, The Convention to Eliminate all forms of Discrimination Against Women, 1980, The Convention against Torture, and Other Cruel, Inhuman or Degrading Treatment or Punishment, 1984.
Adherence to international obligations enshrined in these standard-setting instruments has been accorded the necessary legal significance by the Indian judiciary. The courts have actively used international legal principles to protect human rights in India when there is no domestic legislation to do so. In fact, the Constitution of India has placed an obligation on the Government to observe international law 2. Several of the treaties mentioned above have been relied on by the Courts to bridge the gap between international and domestic law. India allows refugees and asylum seekers access to healthcare, education and justice. The Right to Education Act 2012, for example, guarantees all children between the ages of 6-14 living on the territory of India, the right to free and compulsory primary education. One important concern, from a humanitarian perspective, is that in the absence of a national framework for asylum, refugees come under the purview of the Foreigners Act (1946), the main legislation that deals with the entry, stay and exit of foreigners in India. This act does not recognise refugees as a special category of aliens with a special status and governed by special circumstances. Therefore the Act treats refugees in the same way as other migrants and aliens. A potentially
important development is that the Government of India recently decided that refugees registered with UNHCR may apply for long term visas and work permits. The modalities of this process are currently being developed 3 and some refugees have already benefitted from this. This could result in improved access to the formal job market, especially for highly skilled refugees, and improved access to higher education for younger refugees. The UNHCR had already built a strong understanding of the situation of refugees in Delhi prior to this profiling study by conducting meetings with refugee communities and their representatives, and through protection outreach facilities, and regular participatory assessments. Through this work, for example, it was known, that racial discrimination, especially against the Somali and Myanmarese communities, obstructs access to national health and education services to varying degrees4. While India provides free access to education for all children regardless of nationality, issues like financial constraints, racial discrimination and security concerns lead to low enrolment and high dropout rates from government schools5. Our chapter on childrens access to schools explores these issues further.
3 UNHCR India Fact sheet 2013 : http://www.unhcr.org/50001ec69.pdf 4 Ibid 2 Article 51(c) provides that the state shall endeavour to foster respect for international law and treaty obligations in the dealings of organized peoples with one another. 5 See also : Pittaway E. and Maloney M., Protectors, Providers, Survivors: A Dialogue with Refugee Women and Girls, New Delhi, India, November 2010.
12
The Government of India has recently allowed UNHCR registered refugees to apply for long term visas and work permits, which will allow them to seek employment in the formal sector.
Before this refugees had access to employment in the informal economy without work permits. In the present context, UNHCR seeks to promote refugees self-reliance through skills training and practice (e.g. in tailoring, knitting, block printing, jewellery making). Employment support is also provided in cooperation with UNHCRs partner, BOSCO, and through local institutions and civil society, including job placement support and limited work opportunities at production centres and home based production.
UNHCR also provides grants to small businesses of 20,000 Rupees (315 USD) each, and job placement incentive schemes. The latter initiative aims to help refugees secure employment on their own, and to earn a salary of up to 6,000 Rupees (95 USD) per month. Refugees in the scheme are also paid a monthly incentive of 1,000 Rupees (16 USD) for a period of nine months to encourage them to continue working. UNHCR also addresses the requirements of refugees with specific needs by collaborating with community service providers, community-based organisations and local NGOs. These schemes are focused on time bound employment/child-care, and income generation opportunities for refugees with mental/physical disabilities or protection needs. Protection is provided through outreach centres in areas where urban refugees
live. For this work the UNHCRs key implementing partners in Delhi are BOSCO, which provides support on job placement, education, health, psychosocial support and youth groups; and the Socio-Legal Information Centre (SLIC), which provides legal aid and counselling. In 2012, a subsistence allowance was provided to refugees who were unable to work and had no other form of income or support. The monthly amount was 3,100 Rupees (49 USD) to the principal applicant and 950 Rupees (15 USD) per dependant (up to 7 persons)6. Refugees who receive subsistence allowance include the elderly, sick, persons with disabilities and unaccompanied minors. In 2012, approximately 1,700 vulnerable refugees received financial assistance.
6 As a comparison, the minimum wage in Delhi for unskilled work is approximately 4,000 Rupees (63 USD) per month (20 working days); see: http://labour.nic.in/upload/uploadfiles/files/ Divisions/wage_cell/Delhi.pdf
13
III. METHODOLOGY
14
III. METHODOLOGY
OVERVIEW
Our methodology utilised both a household survey and focus group discussions, and had to be adapted to the particular context and challenges of the refugee populations in Delhi, the fourth largest city by population in the world. The survey took place between March and April 2013, with a total of 1,063 households interviewed across the four target groups. The focus group discussions followed in April and May 20137. The profiling process entailed three main phases: preparation and planning, data collection, and finally data analysis and reporting. The preparatory phase included a scoping mission to Delhi to develop the profiling objectives and methodology. During the data collection phase, we trained staff, piloted and revised the survey tools prior to the data collection. The data analysis and reporting phase entailed tabulation and data processing in SPSS, systematisation of the qualitative data, and then merging the two types of data to produce a draft of the findings. Once the draft was ready, a workshop was held in Delhi (June 2013) to review and discuss our findings and develop recommendations to inform the final report. JIPS structured this workshop but could not attend. Instead, in keeping with our collaborative approach, the workshop was run by our local partners and UNHCR. Representatives from the following groups and organisations attended amongst others the workshop : UNHCR Implementing partner of the profiling exercise (DAJI) UNHCRs implementing partners (Bosco and SLIC) Members of the three refugee communities (Myanmar, Afghan, Somali), who had participated in the profiling exercise as enumerators and focus groups discussion facilitators Centre for Refugee Research from the New South Wales University Catholic Relief Services (CRS) BPRM/USAID After the workshop, the discussion, comments and recommendations were compiled and incorporated into this final report.
III. METHODOLOGY
All our respondents from Myanmar and Somalia had refugee status. Our respondents from Afghanistan included registered refugees, asylum seekers and some respondents who were not of concern to UNHCR, e.g. traders, students, tourists, patients seeking health care and not registered with UNHCR, or persons whose asylum claims had been finally rejected by UNHCR. For our purposes, we included only the registered Afghans and the asylum seekers in our analysis.
Sampling strategy We based our overall target survey sample of 1200 households on project resources and data analysis needs. We divided this roughly according to the proportions of Myanmarese, Afghans and Somalis in UNHCRs data base, and included about a third Indians. We interviewed 1,115 households, but after elimination 8 our final sample for data analysis comprised 1,063 households. These households were then distributed across the four refugee subsamples and one Indian subsample as follows:
Registered refugees (n=112) and asylum seekers (n=64) from Afghanistan: 15.8% (n=176) Refugees from Myanmar : 39% (n=435) Refugees from Somalia : 5.7% (n=64) Indians : 34.8% (=388) Defining the most appropriate sampling approach for the household survey posed some challenges. Firstly, the settlement patterns of the three target groups were highly heterogeneous. Myanmarese refugees were clustered
primarily in specific neighbourhoods of West Delhi, while Somalis were highly clustered in South Delhi and in Wazirabad, North Delhi. Afghans were more dispersed in South Delhi and in Wazirabad, North Delhi. Secondly, the size of the refugee populations varied a lot (from about 200 Somalis to more than 9,000 Myanmarese). A simple random sampling in each ward would therefore not yield enough refugee respondents. Moreover, a single sampling strategy was not appropriate for all the groups, so three different sampling approaches were devised, each one tailored to the three surveyed populations.
8 We had to eliminate 12 households from the original sample of 1,115, who were from other countries, as well as 41 Afghan households, who were not registered refugees.
16
III. METHODOLOGY
NORTH WEST
18 23
32
NORTH
NORTH EAST
225 435
WEST
14
CENTRAL
EAST
3 20
46 111 136
SOUTH
17
III. METHODOLOGY
The sampling approach for Myanmarese refugees entailed two steps. Firstly, we used UNHCRs registration database (proGRES) and a population proportional to size (PPS) strategy to randomly select 20 enumeration areas out of the 30 mapped Sample Sites9. PPS is useful when sampling units vary in size because it assures that households in denser enumeration areas have the same probability of getting into the sample as those in smaller sites, and vice versa. Secondly, we randomly selected 20 households in each of the 20 Sample Sites, for a total sample of 400. This selection occurred by randomly selecting 6-7 apartment buildings in each enumeration area, then going to these locations and randomly selecting the nearest 3-4 dwellings with refugees from Myanmar plus 2-3 Indian/non-refugee dwellings for interviews. Our final sample of Myanmarese totalled 434 households. The sampling approach for Somali refugees entailed enumeration. In view of the very limited number of Somali refugees living in Delhi (around 200 individuals in total) we targeted all of them. We identified their addresses
using UNHCRs registration database, beneficiaries lists of implementing partners and field visits with key informants. We ended up with a total of 64 Somali households. The sampling strategy for the Afghan communities had to address the very low densities in Delhis wards. We randomly identified addresses from UNHCRs registration database (using a random number generator). During the field visits, if no Afghan nationals were living in the selected apartment building, the enumerators selected up to two replacements from neighbouring blocks or buildings. Once an Afghan household was identified, enumerators used a snowball technique (often used in hidden populations which are difficult to access) by asking the identified Afghan respondent to point us to other Afghans in the neighbourhood. In order to limit bias we only selected up to four households through snowballing to complete the desired sample. Indian households were then randomly interviewed in the same neighbourhood. After that, enumerators moved to the next
randomly identified address and proceeded again with the interview and snowballing. We ended up with a total of 217 Afghan households (of which 176 households were registered as refugees with UNHCR. Indian households were randomly interviewed in the same neighbourhoods as our targeted refugees. Specifically, Indian households were selected either from within the buildings where refugee households were interviewed, or from neighbouring buildings 10. Mobile devices were used for the data collection, allowing a faster data processing and higher data quality11.
9 The Sample Site refers to the block, which was chosen as the statistical unit for this survey and identified through landmarks. 10 The questionnaire used can be found at: http://www.jips.org/en/field-support/ country-operations/india/india/ 11 The software Open Data Kit was used to code the questionnaire on the smart phones and store the collected data
18
III. METHODOLOGY
Focus Group Discussions and key informant interviews After concluding the survey, 12 focus group discussions (FGDs) were conducted with men and women and with young girls and boys from each refugee community12. The FGD participants were selected randomly from the surveyed households. We asked all survey respondents if they would be willing to participate in focus groups discussions and if they agreed, we noted their contact number. For the FGDs held with young refugees the participants were identified through the youth clubs of UNHCRs implementing partner, Bosco. The youth focus groups used a story board technique13, in which the participants were asked to make a series of drawings as part of a situational analysis approach. Some of the drawings from these sessions are included in this report. In addition, in order to focus the objectives of the study and to refine the questionnaires we conducted key informant semi-structured interviews with thematic experts from UNHCR and their implementing partners. During the qualitative data collection we interviewed individuals from the refugee communities to provide additional in depth information.
Limitations In addition to the sampling challenges mentioned above, following limitations are relevant to mention : While the analysis in this report provides a comparison of four target groups these groups have been sampled according to different strategies. This is often the case in urban settings, as target populations are usually diverse in terms of settlement patterns and numbers. Moreover the Indian sub sample was included for comparative purposes, meaning it was not composed with the aim of representing the Indian population in Delhi. Additionally, the sampling approach was not designed to estimate the total number of the three targeted refugee communities across Delhi, and the results cannot be used to validate the total number of refugees in UNHCRs refugee registration database. The focus group discussions and to some extent the survey responses may have been constrained or biased because the FGD facilitators and the survey enumerators were to varying degrees linked with UNHCR. This made it likely that some participants sought to use the discussion in a strategic way to communicate messages to UNHCR. Facilitators sought to reduce this by stressing their independence and that they were not in any way involved in UNHCR programming.
Furthermore, as is the case with household surveys, we had to rely on self-reporting, and could not verify answers.
Feedback to communities An important aspect of how we communicate the findings from the report is through feedback to the refugee populations we have profiled. This is important for ethical reasons - informing the refugees about the findings contributes to building trust and two way communication with the humanitarian community - and also enables the validation of our findings through cross-checks. A summarised version of the profiling report will be distributed to all refugee centres, which are run by UNHCR and their implementing partners across Delhi. Disseminating these findings to the communities will also help challenge some misconceptions the different refugee groups had about each other that were encountered during the research process.
12 The FGD question guides used can be found at: http://www.jips.org/en/fieldsupport/country-operations/india/india/ delhi-urban-profiling 13 This methodology was developed by Eileen Pittaway and Linda Bartolomei from the Centre for Refugee Research (CRR) in the University of New South Wales; See : http://www.crr.unsw.edu.au/research-projects/ reciprocal-research/
19
20
IV.
Sex, age and ethnicity of household heads Most of our respondents were heads of household: just under 70% of Afghans, 82% of Myanmarese and 94% of Somalis, but only half of Indians (49%). When the respondent was not the head of household, we asked for sex and age information about the head, as well as other details. Most non-head of household respondents across our sample were the spouse of the head. Our overall survey sample included a majority of male headed households, with a higher presence of female-headed households among refugee populations compared to Indians. Particularly Somali refugees showed a high percentage of female-headed households (43%), compared to Myanmarese (3%), Afghans (32%) and Indians, where only 14% had a female head.
14 This compares with 4.48 for all Delhi, as estimated by the Delhi state survey. See p.i of Employment And Unemployment Situation In Delhi, Based On Nss 66th Round Survey (State Sample) (July 2009 June 2010). Directorate Of Economics & Statistics, Delhi 110054. May 2012 Http://Des.Delhi.Gov.In
21
22
V.
Paying rent is a priority. Sometimes we dont eat at all. We save the money for paying rent.
- Myanmarese man
Understanding the living and working conditions of refugee populations in Delhi was a primary objective of this profiling exercise, and is crucial in helping identify effective responses and programming opportunities. In the following chapter we draw on the urban profiling research of the Feinstein International Centre 15 to identify four different aspects of refugee livelihood vulnerability and resilience employment, housing and financial security as well as physical safety and examine how these differed amongst our surveyed groups, and between households headed by men and women.
We cant get a job, only in Income Generation activities offered by UNHCR and Bosco, and the salary is not enough.
Somali woman
Using this framework we analyse the factors that lead to greater or lesser livelihood security among refugee groups. In addition, we provide a better understanding of the survival strategies refugees use to cope with difficulties, such as evictions and lack of income, with a view to supporting the design of program interventions that can support these strategies.
15 Karen Jacobsen and Rebecca Furst Nichols, 2011. Developing and Urban Profiling Methodology: Final Report. Feinstein International Centre Tufts University. (http://sites.tufts.edu/feinstein/files/2012/01/ Developing-a-Profiling-Methodology-final.pdf)
KEY RECOMMENDATIONS :
1. Employment security : Myanmarese refugees had a higher proportion of income earners. However, they worked in a limited number of professions and often faced discrimination at the work place and unstable work conditions. Afghan household heads had high unemployment levels, but reported better working conditions and access to a greater variety of jobs. Somali household heads had high levels of unemployment and a significant proportion of work related to NGOs, suggesting restricted access to the job market. 2. Housing security : Refugees from Myanmar and Somalia reported more crowded living conditions than Afghans or local Indians, and higher levels of evictions. 3. Financial security : Afghan and Somali households tended to have fewer income earners than their Myanmarese equivalents, but displayed greater access to unearned income sources such as savings, remittances, and UNHCR subsistence allowances.. 4. Physical safety : The proportion of refugees who had experienced theft, robbery or physical assault was much higher than that of Indians, with Myanmarese and Somali respondents reporting far higher levels of assault and lack of neighbourhood safety than Afghans.
23
EDUCATION
HUMAN CAPITAL
SOCIAL CAPITAL
LENGTH OF STAY
EMPLOYMENT SECURITY
Housing conditions
HOUSING SECURITY
Type of work
Employment conditions
FINANCIAL SECURITY
PHYSICAL SAFETY
Experience of physical assault
Household assets
24
1. EMPLOYMENT SECURITY
We define household employment security as a sufficient number of adult income earners (to ensure adequate income for the household), who are employed in decent work 16 which means not being exposed to capricious employers, physical risks or verbal abuse, and having time off to rest. Measures of employment security used in this study include : Number of income earners households with one income earner can be more prone to shocks, and having a second earner is a measure of employment security. Occupational status of household heads we looked at whether the household heads were working, looking for a job, or were inactive (i.e. not looking for a job, disabled or retired). Type of occupation we looked at whether income earners were self-employed or employees, which may indicate different degrees of entrepreneurship. Length of time income earners spent working over the past six months and reasons for not working to assess underemployment and overwork. Work stability, including possession of a work contract. Work conditions we looked at harassment at work as well as unstable working conditions illustrated by experiences of being fired or not paid. The information presented in this chapter refers to income earners and household heads. We defined income earners as people who contribute income to the household through a job. We collected information about up to two main income earners per household. At the same time we also enquired into the occupational status of the household heads, who most often were one of the two income earners. In this chapter we describe our findings, disaggregated by population group and sex of the household head, when relevant. Number of income earners The number of household income earners varied by group and sex. Indian and Myanmarese households had on the average the highest number of income earners: more than half the surveyed households had one income earner, and more than 30% had two income earners. There was no difference between male and female headed households. Afghan and Somali households had fewer income earners: 40% of Afghan households and 28% of Somali households had no income earners, whereas that was the case for less than 5% of Myanmarese and Indian households. About half of Somali households though had one earner, also with no difference between female and male headed households. In all groups with the exception of Somalis, female headed households were more likely not to have any income earners.
% 60 50 40 30 20 10 0
Female-headed HH Male-headed HH
25
26
In all groups, male household heads were more likely to be employed than female household heads, as shown in the Figure below.
27
Type of work
The occupation of the primary two income earners varied by group. Indians occupations were diverse: about 25% run a small business whereas a majority of the others worked as employees (salaried/ work for an organisation, clerk, teacher, driver, IT specialists, etc.). Refugee employment was less diverse. About a third of Myanmarese income earners (33%) worked in factories and another quarter in shops/ restaurants /beauty salons / call centres/ night parties (26%). Our focus groups with Myanmarese described factory work as washing and cutting clothes from 9:30 to 19:30 (including a 30 minute break), which earned them 100 Rupees (1.5 USD) per day. The advantage of this work was that no Hindi was necessary. Work in restaurants was paid more, 4,000 Rupees (63 USD) a month, but no holidays were given, not even Sundays, which posed a challenge for church attendance. In these jobs, however, Hindi and English skills were a prerequisite. Very few Myanmarese refugees (7%) were self-employed or had started their own business. Microentrepreneurs started a business after getting a grant from UNHCR and others after getting a loan from their Church, or a ROSCA (Rotating Savings and Credit Association). These small businesses were likely not to be secure livelihoods, as less than half (45%) were full-time activities. Afghans salaried occupations consisted of three main types: translation/interpretation, employment in shops/restaurants and income generating activities in the UNHCR refugee centres. Other employment included work as drivers, tailors and in clerk/office jobs. Entrepreneurship was more widespread: 40% of Afghan earners were self-employed, usually as interpreters and in activities that did not require initial capital (like service professions). Notably, half of the Somali earners were employed in income generation activities provided by UNHCRs implementing partners, primarily Bosco, and a few worked as interpreters (often in UNHCR or their implementing partners). That Somalis had not entered the local job market, indicates employment vulnerability and reduced self-reliance. One Somali woman reported during a focus group: we cant get a job, only in Income Generation activities offered by UNHCR and Bosco [] and the salary is not enough. Some women are working there, the rest cant work due to security problems [referring to discrimination experience in other jobs].
28
Employment conditions
When asked about work history within the past 6 months, more than half of all groups except Somalis had income earners who had worked on average between 5 and 6 months. Of Myanmarese income earners, 68% had worked most of the time compared to 94% of Indians. Those who had work contracts tended to have skilled jobs (IT workers, engineer, accountants), or work for government/organisations. Very few employed refugees had work contracts. This included interpreters, office workers/clerks, social workers and teachers (most of them were likely to work with a UNHCR implementing partner). When asked about conditions of employment, working Myanmarese and Somali refugees reported the most unsafe conditions: 32% of Myanmarese and 22% of Somali income earners felt physically unsafe at work; whereas 50% of Myanmarese and 46% of Somalis felt unsafe going to or returning from work. The equivalent figures for Afghans were less that 10% for both questions, and for Indians less than 4%. Myanmarese income earners also reported in higher numbers experiences of not having been paid or having been fired within the past 6 months. Discrimination and harassment at the work place was a recurrent topic in all focus groups with Myanmarese refugees, for adults and youth of both sexes. They reported that Indians were paid more than refugees for the same work -- in some cases Indians received 6,000 Rupees(95 USD) for work Myanmarese refugees would be doing for 4,000 Rupees (63 USD). In focus groups, Myanmarese women, both adult and youth, stressed their experience of harassment at work, which often forced them to stay at home with a resulting loss of salary. They said they could not afford to quit their jobs. The FGDs revealed general concern that little could be done to address the discrimination and harassment because discussants reported that nothing happened, even if the incident were reported to the police and SLIC.
% 50 40 30 20 10 0
Not received salary within past 6 months Fired within past 6 months
29
Conclusion
In sum, employment reflected mixed patterns. Myanmarese refugees had more income earners in the household and had spent more time working, similar to that of Indians. However, they faced more discrimination and harassment in the workplace and more tenuous job security (no work contracts). Myanmarese income earners faced more unstable and more hazardous employment conditions: they had a higher likelihood of being fired, increased likelihood of not being paid, and insecurity at or on the way to work. Afghan and Somali households had fewer income earners and more household heads not being employed. Entrepreneurship was more widespread among Afghan income earners when compared to Myanmarese. Afghan income earners also reported better working conditions with fewer incidents of being unsafe at work, or not receiving salary and/or being fired. Myanmarese refugees had the least diverse types of jobs indicating more restricted access to the job market-- whereas Afghans had almost as much diversity as Indians. Half of the Somali income earners were involved in subsidised Income Generation Activities at NGOs, suggesting restricted access to the job market and reduced self-reliance.
30
2. HOUSING SECURITY
Housing is a source of physical shelter and stability and a significant productive asset in urban settings because it underpins income and employment. Housing can generate income through home-based production activities, rental of a room, or it can provide secure storage of goods for vending or trade. Housing vulnerability could be defined as a situation where one is at risk of losing ones home, or where the households living situation is potentially risky for some members. An example of the latter is shared housing when households are obliged to share kitchen or latrine facilities with non-family members. In Delhi we measured tenancy and housing conditions, rent and frequency of evictions. In this chapter we first show differences between groups, disaggregated by sex of household head. Refugees were much more likely than Indians to share their dwellings: almost half (42%) of Somalis and 13% of Myanmarese refugees said they shared their dwelling with non-family members, whereas only one Indian respondent did. More than two thirds of the Myanmarese and more than half the Somali respondents said they shared toilets and/or kitchens with non-family members, reflecting a pattern of these two groups clustering in the same buildings. The Afghans lived in less clustered patterns; only 10% shared their dwelling with non-family members and 11% shared toilet and/or kitchen.
Housing conditions
Households from Myanmar (90%) and Somalia (81%) tended to live in only one room, whereas half of the Afghans occupied 2 rooms, as did about a third of the Indians. According to information shared in the focus groups, rent for one room was about 3,000 Rupees (47 USD) in West Delhi, where most Myanmarese people live. The average rent for Afghans in the Afghan residential areas was 10,000 Rupees (158 USD) for two rooms.
MYANMARESE SOMALI
31
More than half (51%) of Indian respondents owned their dwelling, but no refugees did. Almost all the refugees rented, with the exception of a few who lived in dwellings without paying rent.
MYANMARESE SOMALI
32
Conclusion
In sum, refugees from Myanmar and Somalia appeared to be more vulnerable when it came to housing than Afghans or Indians. They had less living space, with most households occupying only one room and were more likely to share toilets and kitchens with non-family members. This reflects the pattern we observed of the two refugee groups to cluster in the same buildings. By contrast, the vast majority of Afghans did not share any facilities. Refugee households from Somalia and Myanmar also had higher rates of evictions when compared to Afghan households.
33
3. FINANCIAL SECURITY
We define financial security as having enough financial resources to adequately fulfil the needs and some wants of a household. We assessed the financial security by looking at the following proxy indicators: number of income earners (analysed previously), access to unearned income sources such as savings, remittances, allowances and aid, as well as household assets. We also asked our respondents about their expenditures, specifically in relation to rent and medical bills. However, our results showed a high dispersion of values, indicating that these were not very reliable indicators; we therefore decided to omit them from the analysis.
34
Savings were mentioned primarily by the Afghan households (22%). One Afghan housewife described during an interview, how her family had arrived in Delhi with 15,000 dollars in savings. After two years the savings were coming to an end, as only one person in the household was employed, while the rent amounted to 10,000 Rupees (158 USD). Very few Indian households reported receiving government transfers. About 8% of Indians received a social protection programme benefit (or allowance), such as Old Age Pension Scheme (IGNOAPS) and Widow Pension Schemes (IGNWPS).
Unearned income is an indicator of household dependency on external sources and thereby of lower selfreliance. Whether allowances, as amongst Somali households, or savings and remittances, as amongst Afghan households, these sources provide only temporary financial security, and may not be permanent.
35
Household assets
Household assets can be divided into productive assets, which increase household productivity - such as a computer or electricity and which the household either owns or has access to; and transferable assets, which can be sold to increase cash flow (the household must own these in order to utilise their value). Assets are potentially an indicator of wealth or poverty, but it is important to understand the cultural and social context. The items found in refugee homes need not necessarily mean the household can afford to purchase them. They could be inherited from other refugees, who return to their countries or depart on resettlement. Cultural factors should also be taken into consideration. For example, Afghan homes could have carpets and no beds, but this could simply be a cultural preference. What has value in one setting may have less or none in another. In Delhi we asked about eleven types of assets: television, smartphone 17, table/chairs, computer, access to internet, car, motorcycle/scooter, DVD/VCD player, refrigerator and/ or washing machine, cooler, and air condition. We calculated an asset score for each household by assigning each type of asset one point. Thus, scores ranged from zero to 11. Almost half our Somali respondents (46%) had a score of zero, compared to only 9% of Indians. About two thirds of Myanmarese refugees (60%) had scores of 1-2 (indicating access to only one or two of the listed assets). More than half of the Afghan refugees scored between three and five. Indians had the highest scores with 42% having scores of 6-8, and 18% having scores of 9-11. These scores suggest that Somali and Myanmarese refugees are poorer in terms of assets than Afghans and Indians.
17 We did not ask about mobile phones, which most households possess. Asking about smartphones gave us more variation: 11% of Afghans owned a smartphone, compared with 41% of Indians, 1% of Myanmaree and 5% of Somalis..
SOMALI
36
Conclusion
When looking at unearned income sources, Somali households had the highest access to subsistence allowances and aid. Amongst the refugee groups, Afghan households received the most remittances, overtaken only by Indians, and had most often access to savings. Relatively few Myanmarese households reported having savings, receiving allowances, or remittances; though slightly fewer than half received aid (either from UNHCR or other sources- which could be their own community). Unearned income, whether in the form of subsistence allowances, savings or remittances is an indicator of household reliance on external sources and thereby of lower self-reliance. However, less self-reliance, as observed among Somali and Afghan households, is not equal to less financial security. As we saw in the previous chapter on employment security, Myanmarese household have more income earners, but they face insecure job stability and salaries. Therefore, relying on subsistence allowance and remittances or savings, as Somali and Afghan households respectively do, may offer more stability.
37
% 100 80 60 40 20 0
Indian Afghan Myanmarese Somali
38
Conclusion
The perceptions of physical safety in the different neighbourhoods, as reported by the refugee household heads, were closely linked to the relations between each refugee community and the Indian host population. The relations between Myanmarese and Somali refugees and their local neighbours were characterised by discrimination and harassment: Myanmarese and Somalis had for example experienced many more incidents of physical assault, theft or robbery, when compared to Afghans and Indians. Consequently, a great majority of Myanmarese and Somalis perceived their neighbourhoods as being unsafe, as opposed to far fewer Afghans.
39
40
VI.
No one from the community can help us. This is because everyone is in need. If one cant pay the rent, all they can do is pray.
Afghan woman
Everything around us [in Delhi] is changing to the better; new buildings but we dont change. In 10 years they [Indians] will be in heaven and we in hell.
Myanmarese woman
No one goes to the police, neither locals nor Somalis. After a fight, everyone just dusts off their boots and goes home.
Somali young man
41
42
2. HUMAN CAPITAL
Household human capital refers to the education, skills, knowledge, health and other productive capacities (such as the number of working age members) that contribute to a households livelihood security. Higher levels of human capital are likely to decrease vulnerability and poverty. Language abilities are particularly important in urban settings that are characterised by a mix of people of different origins. Speaking an international language can be useful for securing employment in international agencies, or in negotiating onward migration. Speaking the local or most widely spoken language increases access to employment and ability to engage in business. It also helps to enhance protection (for example, if a person is stopped by a police officer), and enables migrants/ refugees to join local organisations and networks. We therefore expect households to be more livelihood secure if an adult member of the household speaks the local languages18. In Delhi we measured education of household heads, skills of income earners (including languages and urban skills), and physical ability of household members to work, wee compared differences between households headed by men and women and explored whether these variables influenced employment, housing situation and financial security. Our survey focused primarily on household heads rather than all household members in order to save time and resources. This meant we missed other household members, whose human capital could have contributed to livelihood security. However, household heads can serve as proxies for the human capital of the household. Households that are headed by children, elderly, physically disabled or sick will be less secure. Similarly, households with large numbers of dependents, or those who cannot earn income, are less secure. Gender is a more complicated factor. Conventional wisdom portrays female headed households as more economically disadvantaged, but this is not always the case, particularly in urban areas where the risks of arrest and detention are faced disproportionately by young men. Race or ethnicity also influences access to livelihood assets when certain groups are subject to discrimination.
18 Often, children speak the local language if they are enrolled in school or more locally integrated than their adult family members. Childrens language skills can be an important asset to the household if they assist adults with translation; however, for the purposes of assessing livelihood security we consider adult household members language ability because of its relationship specifically to employment and protection.
43
Education
Our survey found that the education level of household heads varied by group. The lowest education levels of household heads were encountered among Somalis, where 35% had no education. The majority of Myanmarese (69%) had some primary education, whereas 21% had no education. Afghans had the highest proportion of household heads having completed secondary education (41%) and the most household heads, when compared to the other refugee groups, having completed a university degree (17%). We looked at whether there was a difference between female and male household heads and as expected, a larger proportion of female household heads had no education (over 30% vs. less than 15% of male heads).The situation was particularly uneven for Somali households, where 64% of female heads of households had no education compared to only 12% of male heads. In order to better understand how education is a factor of livelihood security, we cross correlated the education levels of household heads with different indicators of livelihood security, such as employment status, working conditions and experience of evictions. We found no relation between the education level of household heads and the likelihood for having experienced evictions. Evictions were primarily reported by Myanmarese and Somalis regardless of the education level of their household heads. We then explored whether the education level affected the employment conditions of household heads as expressed by two variables: experiences of being fired and of not being regularly paid. Again we found no strong correlations.
% 100 80 60 40 20 0
Indian Afghan Myanmarese Somali
44
Finally, we looked at the links between education level and employment of household heads. As the figure shows, higher education levels among refugees were not linked to higher likelihood of having work. Somali refugees were the exception, as the majority of employed household heads had a university degree. Afghans, for example, during the focus groups lamented the fact of not being able to use their education and certificates. The general lack of links between education levels and livelihood security, be it employment or housing security, imply that previous education has limited significance.
On the contrary, language skills, as we will see next, are closely linked to the ability to find work and access better paid work.
45
Languages
One possibility for refugees difficulty with finding employment could be lack of language skills. In Delhi the key working languages are Hindi and English. We asked our survey respondents whether any income earners in the household spoke these languages. Somalis had the highest proportion (89%) of Englishspeakers, but only 39% of them spoke Hindi. Myanmarese had the lowest proportion of English speakers (15%), but more than half spoke Hindi. About 94% of Afghan income earners had Hindi skills and 44% English skills. Afghan refugees told us that many learned Hindi in Afghanistan, often thanks to Bollywood movies. The focus groups discussed the link between language skills and salary levels, and agreed that better paid jobs required both Hindi and English skills. Similarly, refugees from Myanmar reported in the FGDs how limited Hindi skills often led to misunderstandings with the employer and at times resulted in receiving less in salary.
% 100 80 60 40 20 0
English Hindi
46
Urban skills
In addition to languages, urban life requires specific skills. In a commoditised urban economy, households coming from rural areas are likely to lack skills that are easily transferable to the urban setting. Such transferable skills include business, trade, a profession, office or factory work, or services. Of our respondents, 80% of Indians and 85% of those from Myanmar were from rural backgrounds compared to about a third of Afghans (38%) and fewer than 2% of Somalis. Nevertheless, as the graph below illustrates, very few Myanmarese reported having farming skills, which contradicts the usual assumption that a rural background implies farming skills. More than half of the Myanmarese and Somali respondents reported that no-one in their household had any job skills. A small proportion of Somalis reported having computer skills and translation skills. The Afghan population had a wider range of skills: IT, teacher, driver, tailor, interpreter, cook.
NATIONALITY SKILL No Skill Computer Cook Small Business Tailor Driver Farming Office job/clerck Teacher Interpreter Indian 7 19 22 24 3 12 8 12 4 0 Afghan 28 23 10 1 20 8 1 5 8 12 Myanmarese 55 6 4 5 13 2 8 0 3 2 Somali 52 19 5 2 3 0 0 2 5 27
47
% 150 120 90 60 30 0
Indian Afghan Myanmarese Somali
48
% 80 70 60 50 40 30 20 10 0
Indian Afghan Myanmarese Somali
Conclusion
Among the refugee groups we surveyed, Myanmarese and Somali households had the highest proportions of income earners and the lowest dependency ratios but they also had the lowest education levels and fewest skills. One explanation for this apparent contradiction could be that households with little human capital require more members to work. However, lack of education and skills mean they must accept insecure employment. We found that many Myanmarese and Somali refugees do not get paid regularly or have been fired (primarily Myanmarese). In addition, both Myanmarese and Somali refugees appear to experience harassment and discrimination at the work place as well as on their way to and from work. Afghans appear to have the most human capital, with the highest education levels, wider range of skills, better Hindi skills and few health issues preventing household members from working. At the same time, Afghan households had the lowest number of income earners. One explanation is that higher human capital allows more selective approaches to employment. For example, one Afghan respondent who was trained as a doctor reported that he would not accept to work as a cleaner.
49
Direct assistance is only provided from the community and the church. From Bosco and UNHCR it can take time, you ask for an appointment. Through the church, it is direct ; 2-3 days until the pastor convenes with the elders and grants money.
The money provided by the community and churches was sometimes a loan and sometimes a gift. Most Myanmarese contribute to these safety nets by paying 10% of their income to the church. One Myanmarese female informant said: if you are Christian you will pay, if you dont do it, it is like stealing money from God. Somalis also showed high levels of social capital, but although there was willingness to support each other, lack of resources prevented such support from being effective. One womans response captured the general view :
Refugees try between themselves to donate to each other; but it is difficult to get the required amount together. If one household cant pay their rent they have to vacate the house. Then they have to stay with other refugees.
50
The Afghan community demonstrated a relatively high degree of distrust, and most Afghan survey respondents said there would be no one to turn to in sudden need. Some Afghans during the FGDs said they would be unwilling to start a business with another Afghan as they would not trust them not to run off with the start-up capital. Others said they would keep it secret if not granted refugee status, as some Afghans could approach their employer to let them take over the job if they had the required documentation.
Afghans similarly reported not telling others if they got evicted as sharing such information might make them more vulnerable. However, it is interesting to note, that this lack of trust did not prevent Afghans from using their community when looking for a job (see below). Similarly, distrust seemed not as the only reason for lack of intra community safety nets, as one woman during an FGD said: No one from the community can help us. This is because everyone is in need. If one cant pay the rent, all they can do is pray.
We asked all income earners how they got their job. More than 70% of Afghan and Myanmarese employees said they found their job though the community, friends or relatives.
MYANMARESE SOMALI
51
In response to the question of where household heads would go for help, Afghans demonstrated the least confidence in their communities, and Myanmarese and Indians showed the highest levels. Of the Afghan respondents, almost a third said they would ask relatives abroad and half said they would have no options available to them for such help. Over half (56%) of our Myanmarese respondents said they would turn to their church or community association for assistance and 21% said they would ask neighbours or friends from the same ethnic group. About a third (34%) of Somali respondents said they would ask aid agencies for help. More than a quarter (28%) of Somalis said they would have no options
available to them for such help. About a third of Indian respondents said they would ask relatives in Delhi and another third (32%) said they would ask neighbours or friends. Given the importance of existing networks for new arrivals, we asked whether respondents had friends or relatives already in Delhi before they arrived. A much greater proportion of Indians, almost 64%, said they had relatives in Delhi before arriving, compared with less than 20% of all other groups, The focus groups shared diverse experiences of arriving in Delhi. Some Somali refugees said they were brought to Delhi by agencies that were not transparent about the destination they offered.
A Somali informant was not aware that she had been flown to Delhi until a local confronted her with a map. Several Afghan informants told us that on arriving at the Delhi airport, they took a taxi to the area with many private hospitals, where Afghans were known to work as interpreters for wealthy health tourists from Afghanistan. In those areas they expected to find other Afghans that could advise them on where to find an apartment and how to navigate in the first days.
NATIONALITY SOURCE OF SUPPORT Community association -Church None of the options would be available Relatives here in Delhi Neighbors-friends in general Neighbors/friends of same ethnic group Relatives abroad or other part of India Aid/humanitarian organization Indian 0 17 37 24 9 13 0 Afghan 1 51 2 6 2 32 6 Myanmarese 56 10 2 6 20 1 5 Somali 3 30 0 17 0 15 35
52
If we were more rich and could pay rent on time and were more educated and worked in better jobs, the problems would be fewer. We are no benefit for them; they [the locals] think, they are very poor, they are nothing, we are not afraid of them.
This feeling of inferiority may be another reason for avoiding engagement with locals. Some Myanmarese expressed a sense of stagnation in their community when compared to the developing Indian society :
Everything around us [in Delhi] is changing to the better; new buildings but we dont change. In 10 years they [Indians] will be in heaven and we in hell.
Afghan refugees said they felt discriminated against mainly because of their religion; landlords would refuse to rent apartments to Afghans for that reason. Some mentioned the physical harassment
53
Even though many fights took place during daytime, no bystanders interfered, and the Somalis being always in a minority were subjected to assaults. No one goes to the police, neither locals nor Somalis. After a fight, everyone just dusts off their boots and goes home. Most preferred not to report incidents to the police as this could lead to further discord and enmity with neighbours and landlords. Somalis were aware of Indians view of them:
Locals claim we are thieves. The media has been blaming Nigerians for a lot of criminality and locals cant tell the difference between Nigerians and Somalis, so they also blame us.
As a result, several Somalis expressed an unwillingness to learn Hindi. Like the Myanmarese, some older women said not speaking the language is an advantage, as then they dont understand what Indians say about them. A young man who had lived five years in Delhi, stated he didnt want to learn Hindi I hate the language, without any reason.
54
This group highlighted discrimination as a major challenge in the community (first drawing). This manifests itself, when locals comment on and make fun of their different dresses and touch their hair (second drawing). Somalis are often assaulted by locals; the other Somalis cant do anything but look and call UNHCR, SLIC and the police for help (third drawing). The boys
said that if they had the power to change things, they would ensure a just judiciary (fourth drawing). Their response to what they think could be done to address the discrimination and harassment was recommending more meetings between the local community, Somali refugees, UNHCR and the police (fifth drawing).
55
This drawing shows a Myanmarese young woman harassed at her workplace by colleagues.
The degree of harassment and discrimination experienced by the different refugee populations was quite varied. For Myanmarese and Somalis the issue came up in most of the topics we discussed during the focus groups, whereas it came up to a much lower extent for the Afghan refugees. These different relations with the local community reflected different degrees of bridging social capital. Somali refugees reported having almost no contact to the local community in order to avoid harassment, which was a daily experience for them. A Somali woman reported in a focus group that the reason for the very high clustering of Somali families in the same areas and even buildings was
an attempt to delimit their neighbours and surrounding environment so as to ensure greater safety. The suggestions, made by the focus group participants in order to address discrimination, were either to provide separate neighbourhoods and schools for refugees only, or, in a more optimistic vein, to conduct additional community mediation. The high degree of discrimination faced by Somalis could explain why most preferred to avoid the job market. Instead, over 50% of the Somali income earners were employed in the income generation activities of UNHCRs implementing partner and 23% were working as interpreters; again primarily for UNHCR and their implementing partners.
The Myanmarese refugees also experienced a high degree of discrimination and harassment in their daily lives; at work, in school, by landlords and neighbours. However, this did not prevent them from accessing the informal local job market though under very unequal conditions and with recurring experiences of being fired, not paid, and sexually harassed in case of women. A general perception in the focus groups was that there was little that could be done in order to improve the situation. Many felt that it was futile to report harassment and discrimination to UNHCR, SLIC or the police. The Myanmarese community networks were not capable of addressing this issue either, as they themselves had no voice in the local community.
56
Afghan refugees experienced less discrimination and were not stigmatised, perhaps because of their physical appearance and because they were better educated and spoke the local language. They were able to access apartments in better neighbourhoods among Indians without clustering in refugee neighbourhoods as Myanmarese and Somalis did. They were able to apply for better jobs, though lack of work permit and documentation was most often a barrier to actually getting those jobs. Some Indians told us that the discrimination that Afghans faced was similar to what Muslim Indians face in certain neighbourhoods.
Conclusion
Having access to social capital can increase the likelihood of securing employment, housing and financing. Social capital can potentially also protect against harassment or other forms of abuse. We looked at the social capital of the refugee groups in terms of bonding and bridging capital. Myanmarese refugees had more bonding capital in the form of strong safety nets. For example, they could access money from the church when in need, and they were hosted by neighbours if evicted. Afghans appeared to have much less social capital, with more than half reporting they would have nowhere to go if in need, other than turning to relatives abroad. During FGDs, many described how secretive Afghans were about problems such as eviction or not getting refugee status. Somali refugees did report higher levels of intra-community support. However, the community was small and had few resources, which made the support offered less effective. Bridging capital was weakened by the discrimination and harassment experienced by the refugee communities. Somali and Myanmarese refuges reported the highest degrees of discrimination and harassment in their daily lives in the neighbourhood, by landlords, in the school, by local authorities, at the workplace etc.
57
MYANMARESE SOMALI
To explore the impact of length of stay, we chose to look at the employment status of household heads in correlation to the duration of their stay in Delhi. Among the Indian migrants and the Afghan refugees and asylum seekers there is a similar pattern of increase in
the employment of household heads according to their stay in Delhi. However, a decrease is then noticed for those having been in Delhi the longest (this could be related to older age as well). Among Myanmarese and Somalis an increase is also clear, though to different degrees.
58
5. CONCLUSION
Not all forms of human capital lead to greater employment security. Despite good education levels among the Afghans, they had the lowest employment levels. One explanation is that high education leads to high expectations for employment. However, accessible urban jobs are low skilled, while better paid, high skilled jobs require Hindi language skills and documentation beyond the refugee identity card. Similarly, social capital in the form of intra community networks did not necessarily lead to greater employment security, but it did make the Myanmarese refugees more secure by offering strong safety nets, whereas Afghan households had to rely to a much higher degree only upon their own household and support from abroad in form of remittances. Thus, both Myanmarese and Somali households had to counterbalance their low human capital with stronger safety nets within the community. Myanmarese had effective safety networks whereas the much smaller community of Somalis had fewer resources to share. Relations to the Indian host population, so-called bridging social capital, were weakest amongst Myanmarese and Somalis. High degrees of discrimination and harassment affected their access to work and housing as well as their general physical safety. Only certain types of employment (e.g. badly paid jobs in factories) were accessible to Myanmarese; they experienced the highest degrees of evictions and were often harassed at the workplace, in the neighbourhood or at school.
59
VII. EDUCATION
60
VII.
the problem is the verbal abuse and being a minority, kids are being beaten, we worry about their security, not talent.
Somali woman
Through collecting information on all children in each household between 5 and 18 years our survey explored the types of education refugee children are receiving, the level of school attendance, and the reasons for low attendance. In addition, we held focus group discussion with the youth based on the storyboard technique (see methodology chapter III). Girls and boys between 15 and 24 years attended separate groups according to sex and origin.
Almost half of Myanmarese children (41%) are not enrolled in any type of school, a significantly higher number than Afghan refugee children (14%) and Somalis (13%). Reasons for non-attendance differed across these groups. While significant numbers of responses across all groups mentioned financial constraint as a factor limiting their school attendance, more Myanmarese children reported the need to work as a constraint, while Somali children reported language difficulties as a more prevalent factor. Discrimination at school was a problem, with Somali and Myanmarese girls and boys reporting that they were often verbally and physically abused by their classmates and teachers. Across the target groups we saw a continuity of human capital levels. For example, Afghan households with the highest education levels amongst household heads had the lowest percentage of children not attending school.
61
NATIONALITY EDUCATION TYPE Not enrolled anywhere Governmental school Private school English class Church school Vocational training Bridge school National Open School Indian 7 51 39 1 0 0 0 0 Afghan 16 17 24 38 1 13 10 13 Myanmarese 41 8 4 5 24 6 3 0 Somali 13 24 29 19 0 13 17 0
62
The survey responses showed that almost half (41%) of Myanmarese children were not enrolled in any type of school, compared with, 16% of Afghan refugee children and 13% of Somali children (only 7% of Indian children were not enrolled). More than half of the Indian children attended Government schools.
Among refugees, Somalis attended Government schools in higher numbers (24%). Private schools were attended primarily by Indian children (39%), followed by Somali (27%) and Afghan (24%) children. Very few Myanmarese children attended private school, whereas a significant number (24%) attended Church schools.
In general, there were no significant differences between male and female children in any population group, though there was a tendency for more male children in all population groups to be enrolled in private school. Of those children attending school, most did so, on a daily basis,
NATIONALITY REASONS FOR NOT ATTENDING SCHOOL financial constraints child is working medical reasons Language issues discrimination/hostility distance/no school drop out/failure security not safe lack of documentation Indian 56 3 12 0 0 9 18 0 0 Afghan 49 2 5 21 0 13 0 2 6 Myanmarese 63 38 10 1 6 1 0 2 1 Somali 38 13 0 63 13 0 0 13 0
63
When looking at the differences in access to education across the refugee communities interesting patterns arise. The Afghan and Somali community invested in the human capital of the future generation and sent their children to school in much higher numbers than the Myanmarese community. The relatively high attendance of Somali children in Government schools, despite the discrimination and harassment reported by all youth, is worthy of notice. At the same time, the disparity between age and grade, the reported exclusion by the teachers, and the difficulties in understating the class due to challenges with the Hindi language, all pose questions about the actual learning outcomes enjoyed by the Somali children attending Government schools. The Myanmarese community had the highest number of children not going to school. In the survey the main reasons reported for not attending school were financial constraints and the need to work. The focus groups reported discrimination and harassment by local children and teachers as an important factor as well. Although many Myanmarese children attended Church school, this would not provide them with a certificate recognised in India.
Some very interesting points emerged during the focus group discussion held with the youth from the three refugee communities. Afghan boys and girls expressed strong motivation to pursue higher education and lack of access to colleges was raised as a challenge by almost all Afghan youth in the focus groups. Several said they had not been able to enter college because they lacked the required documentation (often a valid visa was requested by the university). Many Afghan girls and boys explained that the Hindi language, particularly written, was too difficult and they preferred to attend the tuition classes in English at Bosco, which serve as a preparation for the Open School exams. Some Afghan boys expressed worries that the few hours of classes would not make them competitive enough in the final national exams, and that this would be a setback to being able to proceed with college education. There was little to no awareness among the Afghan students about the DAFI Scholarship and in general the youth seemed not to have any ideas as to how to enter college. The alternative though seemed clear, as an Afghan boy said: If we dont study or go to college, our future will be to become a waiter or a driver.
Afghan girls said a main challenge faced by youth was lack of awareness about the programs and courses available in India. They felt that there was no one to advise them about possible future career plans, and the tuition classes at Bosco, offered only limited subjects, such as economics, business, English, cooking etc. If students wanted to pursue a different direction they had to find the relevant books and teach themselves. In addition, some curricula were only partly taught, placing a great responsibility on the students for fully preparing for the exams. The Afghan girls stressed that Open School tuition classes were attended only by strong students or those with educated parents who could support their children. The weaker students would not be able to get through the curriculum with so little teaching.
64
Discrimination in the Government schools was reported most often by the Somali and Myanmarese youth during the focus groups. The Somali girls reported that they were made to sit at the back of the classroom or sometimes in a class without chairs, where they often did not receive instruction. Girls of 16 and 17 years in the focus group explained that they were placed in classes with children of 6-7 years of age due to their inadequate Hindi skills. It is interesting to note that Somali youth did not attend the Open school, which is in English. One explanation is that many Somali women were illiterate and therefore not able to support their children attending the more demanding Open School process (as indicated by Afghan girls previously).
Somali and Myanmarese girls and boys reported that they were often verbally and physically abused by their classmates and teachers. A Somali boy reported that the local Indian students told Somali children that they would never be like them, no matter how they dressed or how much they studied. A Somali woman expressed her concerns very succinctly: the problem is the verbal abuse and being a minority, kids are being beaten, we worry about their security, not talent. Women reported that their kids often came home from school with injuries. Both women and youth in the focus groups said UNHCR should maintain a school just for refugees.
Despite the reported discrimination against Somali and Myanmarese children and youth, many, primarily Somalis (24%), kept attending Government schools. The frustration and sadness caused by these daily experiences though were clearly communicated by the youth in the focus groups.
65
The girls identified lack of access to higher education as an important concern they had (first drawing). The results of lacking access for refugees are illustrated in the difference between literate and illiterate persons, who accordingly either get a good job in an office or need to work in a warehouse (second drawing).
Women will then [when illiterate] stay at home, work in warehouses or not even get a job.
The Afghan refugee community was not able to respond to this challenge (third drawing).
66
The girls suggested that a one good idea would be to introduce some exams that could place the students at the right level and thereby recognise their previous education (fourth drawing).
What the community can do, in order to better respond to the challenge of not being able to access college, is to attend the Open School classes ; nevertheless, this does not address the challenge of not accessing college (fifth drawing).
67
4. CONCLUSIONS
All communities showed a tendency to transfer the existing either high or low human capital levels to the next generation (even if Afghan parents in the focus groups expressed worries about their children being less educated than themselves). In general, we see a continuity of human capital levels in the different target groups (sometimes a decrease). Thus the Afghans, with the highest education levels amongst household heads, had the lowest percentage of children not attending school, and the Afghan youth expressed very strong motivation to enter college, considering it decisive for their future. The Myanmarese community, with the lowest education levels of household heads, also showed the largest percentage of children not attending school. The Myanmarese youth in FGDs mentioned only once the lack of access to college as a serious concern. The Somali community was somewhere in between: the adults had low levels of education; however, the children did attend school in relatively high percentages. However, it is important to add, that the Somali youth in the focus groups did mention that their attending school often brought few actual learning outcomes and that their primary concern was discrimination in school and not future plans for entering college.
68
69
70
BOSCO should strengthen their engagement in sensitisation of employers with the aim of improving the working conditions and job security. Mediation between employees and employers could also prevent misunderstandings over work and salary arrangements. This should be targeted specifically at work places where many refugees from Myanmar are employed. BOSCO should explore work placement of groups, as this could potentially decrease the risks to personal safety faced on the way to and from work, as well as in the work place.
Refugees from Afghanistan: Afghan refugees often attempt to access the more formal job market. Possessing accepted documentation is therefore often a challenge. Employers often require a valid visa, not recognising the UNHCR issued refugee certificate. In addition, the certificates and diplomas of refugees are not acknowledged.
BOSCO should sensitise employers on the significance of the UNHCR issued refugee card. UNHCR should advocate with the Government to recognise certain diplomas or opportunities for recertification.
Refugees from Somalia: Somali refugees face serious difficulties in entering the labour market in Delhi, often the main reason for this being discrimination and the lack of valid visas. Most Somali refugees are therefore engaged in income generation activities offered by UNHCR and partners.
SLIC should sensitise local communities with the aim of increasing local employment options. BOSCO should encourage Somali refugees to start small businesses and apply for small grants. This could possibly provide a longer term path to entering the local labour market.
CROSS-COMMUNITY RECOMMENDATIONS :
To establish community focal points for livelihood issues in order to enable a continuous dialogue between refugee communities and UNHCR and partners. BOSCO should continue to educate employers on refugee rights and the value of UNHCR documentation, in order to improve work conditions and stability. UNHCR and BOSCO should explore the option of linking refugees with public and private sector initiatives to build their skills. UNHCR should directly engage with industry and commerce associations such as Federation of Indian Chambers of Commerce and Industry (FICCI) and Confederation of Indian Industry (CII). BOSCO should explore revolving funds for groups of refugees for small businesses such as tea shops and restaurants or other vending enterprises.
71
2. HOUSING SECURITY All three refugee groups reported varying degrees of housing insecurity due to restricted access to housing, discrimination by landlords, and evictions. Primarily it was Myanmarese and Somali households who had experienced evictions.
BOSCO and SLIC should conduct sensitisation sessions with landlords, neighbours and neighbourhood associations in areas where refugees live to minimise and prevent conflicts. This is particularly needed in the areas where refugees from Somalia and Myanmar are residing.
3. PHYSICAL SAFETY HARASSMENT AND DISCRIMINATION Refugees from Myanmar and Somalia reported facing the most discrimination and harassment in the neighbourhood, by landlords, at work, in school, and by local authorities - and had the worst relations to their local communities. Myanmarese were counterbalancing this by strong intra-community support networks. Afghans experienced less discrimination from the local community, however they reported having less community networks to access and had to rely more on support from abroad. Somali refugees had few established community networks to provide a safety net, and faced a high degree of racial discrimination and harassment. Somali men during a focus groups discussion mentioned the difficulties faced by Somali refugees upon arriving to Delhi, and they wished for some support in organising a receiving and introducing mechanism within their community.
UNHCR should provide information to newly arrived asylum-seekers about available services in Delhi. This can be done partly by supporting community groups/initiatives to aid newly arrived refugees. UNHCR and partners should continue bringing together refugee and local communities during cultural events and festivals, in order to lessen misunderstandings and improve communication. UNHCR should continue to support the formation of committees/community groups with good communication structures that can discuss and promote refugee interests. UNHCR should identify leaders and/or key community members of the local Indian community, who can facilitate relationship building with refugees and mobilise local communities to reduce harassment and discrimination against refugees. UNHCR and partners should bring together Indian and refugee women on a common platform to bridge gaps, and to build communication to address gender based violence. UNHCR and partners should organise joint youth workshops with Indians and refugees. BOSCO should explore facilitation of basic skills training facilities in the refugee neighbourhoods, also open to Indian youth.
72
4. EDUCATION Access to education - primary, secondary as well as higherand education quality was a concern cutting across the three refugee communities. Myanmarese children were, however, in greater numbers than Somali and Afghan children, not attending any type of school. Harassment and discrimination was identified as a problem in government schools primarily by Myanmarese and Somali children. Work commitments were reported by Myanmarese as the primary explanation for not attending school, while among Somalis the main reason was the language barrier. The alternative to Government Schools, the Open School classes provided by UNHCRs partner, seemed to cover only a restricted array of subjects and target primarily stronger students.
UNHCR and BOSCO should continue to identify government schools in areas where refugees live and to advocate for admission of refugee students. UNHCR and BOSCO should strengthen engagement with local government schools and local communities to raise awareness about refugees, address issues of discrimination and harassment and enhance parentsteachers associations. UNHCR and partners should organise joint events for refugee and Indian children. BOSCO should offer more frequent classes and a greater variety of subjects for Open School students. Some refugees possess higher education but no job as their certificates are not recognised. Their knowledge could be used by engaging them in the classes offered by the Open School, in order to expand the subjects offered and support knowledge transfer and network building between refugee communities. UNHCR and BOSCO should combine open schooling opportunities and income generation activities for refugee children and youth, who are unable to attend regular schools due to financial problems and engagement in paid work (primarily a concern among Myanmarese youth). UNHCR should continue to advocate with tertiary educational institutions to facilitate admission of refugees and charge fees at par with Indian nationals. BOSCO should organise information campaigns to increase awareness and knowledge of the DAFI scholarship. UNHCR and BOSCO should offer vocational training courses and job placement based on the skills in high demand such as car repair, IT repair and mobile repair. UNHCR and BOSCO should continue to promote language courses in Hindi among refugee children, youth and adults.
73
74
75
76
Myanmarese somali 111 225 32
Darod % % % % % n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. Ashraf Madiban Tuni
IndIcator / dIsaggregatIon
Unit
Indian
a. saMPle dIstrIBUtIon
Hawiya Bantu % n.a. n.a. n.a. 1.6 1.6 14.1 15.6 10.9 54.7 % n.a. n.a. n.a. 1.6
sample size, by district 136 14 23 3 176 16.6 40.9 6.0 435 64 0 0 0 18 435 0 0 46
South Delhi
West Delhi
East Delhi
all districts
B. DEMOGRAHICS afghan
afghan
Indian
Other
head of household 27.9 15.2 2.1 0.0 0.5 0.3 1.0 1.3 2.3 0.3
Bihar
% 23.3 5.1 0.6 0.6 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3.1
Assam
70.5 14.7 1.4 0.7 0.2 0.0 0.0 0.2 0.9 0.2 0.0 1.6 0.0 0.0 0.0 0.0 1.6 0.0 0.0
81.6
93.8
husband-wife of head
son-daughter of head
brother-sister of head
father-mother of head
grandfather-grandmother of head
grandson-granddaughter of head
Friend-not related
son-in-law- daughter-in-law
distribution of Indian household heads according to state of original residence % % % % % Jammu and Kashmir Jharkhand % Madhya Pradesh Maharashtra Mizoram % % % Punjab Rajasthan Sikkim % % % 0.5 14.2 42.4 7.2 0.3 0.5 1.0 0.3 0.5 1.6 0.8 0.3 n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a.
other
distribution of household heads according to sex 14.5 85.5 67.6 62.8 57.1
Haryana
Female
Male
Female HoH 55.3 58.0 38.5 36.5 36.0 40.5 37.3 37.1
Mean
Yrs
Median
Yrs
Male HoH 46.8 47.0 39.0 36.0 27.0 40.9 37.4 29.3
Mean
Yrs
Median
Yrs
all household heads 48.0 48.0 39.0 36.0 40.7 37.4 32.6 29.0
Mean
yrs
% % %
Median
yrs
(B. DEMOGRAHICS)
Unit
Indian
(B. DEMOGRAHICS) Indian 16.4 2.1 1.3 0.7 0.0 8.9 8.0 1.4 afghan Myanmarese somali 45 to 60 years % % More than 60 years
Unit
Mean
all HoH Under 18 years 18 to 45 years % % % Unit Indian afghan 2.4 1.6 0.8 Myanmarese 16.2 7.8 7.8 52.7 41.7 47.6 % 28.7 48.9 43.8 35.7 61.2 3.1 0.0 somali
Median
5.0
4.0
4.0
4.0
Mean
Median
5.0
6.0
5.0
1.5
All HoH 5.3 5.0 5.0 4.0 2.0 5.4 4.7 3.1
Mean
distribution of households according to number of income earners, by sex of HoH Female HoH 0 1 % % % 8.9 30.4 53.6 % 7.1 52.6 36.8 10.5 0.0 8.6 56.8 30.2 4.3 25.9 51.9 22.2 0.0
Median
Female HoH 1.5 2.8 1.7 2.0 2.1 2.5 1.9 2.2
Under 18 years
45 to 60 years
Under 18 years
3+ all HoH 0 1 2
2.8
2.5
2.4
1.7
45 to 60 years
all HoH
Under 18 years
3+
18 to 45 years
distribution of households heads (HoH) according to occupational status, by sex of HoH Female HoH Self-employed - Own business Employed-Salaried Unemployed but looking for job % % % Unemployed but not looking for job % Working for relatives (unpaid) % 7.1 32.1 5.4 21.4 0.0 14.0 24.6 42.1 15.8 0.0 6.8 39.5 4.9 5.6 0.0 0.0 3.7 11.1 44.4 0.0
45 to 60 years
Female HoH 55.2 37.5 5.0 2.3 1.0 7.4 47.1 47.2 4.1 0.0 44.6 48.8
28.6
18 to 45 years
45 to 60 years 4.7
14.8
% % %
13.9
Under 18 years
28.8
Disabled
84.7
0.0
0.0
0.0
0.0
18 to 45 years
52.8
77
78
Myanmarese 41.4 0.0
creative arts (musician, dancer, theater, film) % 0.4 0.0 0.3 0.0
Unit
Indian
somali
Indian
afghan
Myanmarese
somali
Self-employed - Own business 63.0 2.6 0.0 0.4 0.4 0.7 0.4 27.1 0.4 0.0 2.8 0.0 0.0 36.1 0.0 13.9 22.2 22.2
religious cleric
36.9
5.1
2.8
tailor
Employed-Salaried 1.8 4.2 0.0 0.0 1.2 8.2 1.2 0.3 0.0 0.8 0.0 0.0 7.6 0.0 9.2 31.9
46.2
Domestic work
Student
Retired
Disabled
self-employed - own business 54.3 3.4 2.1 0.2 0.5 0.5 0.7 32.4 0.2 0.0 1.6 0.0
Cook
32.6
5.7
1.6
employed-salaried 2.3 6.7 0.0 2.1 1.0 9.0 1.8 0.3 0.0 1.1 0.0 0.0 5.7 0.0
44.2
domestic work
student
retired
disabled
small business
24.9
teacher
2.2
7.5 Proportion of employed income earners with negative experiences (multiple answers) 54.1 1.5 2.9 16.0 4.0 6.6
50.2
46.2
engineer
craftsman
d. HoUsIng secUrIty
Unit
Indian
Female HoH
4+
Male HoH
4+
all HoH
4+
20.2
number of rooms occupied by household 2.5 Female HoH Male HoH All HoH % % % 2.0 2.0 1.0 1.0 1.9 1.1 1.1
Proportion of households that have been evicted, by sex of HoH 5.4 3.3 3.6 5.3 0.8 2.3 80.2 80.2 80.2 66.7 58.3 61.9
Mean
Median
Female HoH 0.0 7.1 0.0 10.5 69.1 55.6 8.8 82.1 51.9 3.5 16.0 63.0
Proportion of households that have been evicted according to reasons for eviction (multiple answers) Could not pay rent any longer Could not pay rent on time Owner did not want us here Development projects in the area Because of discrimination/ harassment Because of children/visitors/ big family Increased rent/overcharged by owner Other describe % % % % % % % % 7.1 28.6 57.1 0.0 0.0 0.0 0.0 7.1 50.0 0.0 50.0 0.0 0.0 0.0 0.0 25.0 35.9 50.0 18.7 0.6 11.8 10.1 5.2 6.6 47.4 65.8 52.6 0.0 2.6 0.0 7.9 5.3
Sharing dwellings
Sharing toilet
Sharing kitchen
Male HoH
Sharing dwellings
Sharing toilet
Sharing kitchen
all HoH
sharing dwellings
sharing toilet
sharing kitchen
share rent
share food
100.0
share income
79
80
Myanmarese somali
F. PHIsycal saFety Unit Indian afghan Myanmarese somali
e. FInancIal secUrIty
Unit
Indian
38.6 92.6
physical assault % 1.8 19.3 76.5 44.4
17.9
22.2
theft or robbery
3.6
5.3
53.7
14.8
Assistance
Savings
33.6
12.5
22.2
Assistance
Savings
% 35.2 5.7 2.3 22.5 2.1 0.0 2.3 58.7 46.4 79.4 14.5 22.2
assistance
Benefits/allowances
savings
17.5
23.3
53.8
1 to 3
4 to 6
7 to 11
Male HoH
1 to 3
4 to 6
7 to 11
all HoH 8.5 15.5 27.5 48.4 3.4 0.2 32.4 8.8 4.8 0.0 50.0 69.1 47.6 14.2 21.8 47.6
1 to 3
4 to 6
7 to 11
number of assets owned by households 5.8 6.0 3.0 2.9 1.5 1.0 0.9 1.0
Mean
Median
Unit
Indian
afghan
Myanmarese
somali
H. socIal caPItal
Proportion of head of households working, by education level 71.0 70.5 82.6 86.3
completed secondary completed university or postgrad % % % % % % % % 16.4 16.4 38.3 38.3 24.3 24.3 21.1 % % % 21.8 19.3 18.8 18.8 23.3 23.3 40.9 40.9 17.0 27.4 21.8 47.1 19.3 37.5 27.4 23.5 47.1 13.2 37.5 10.1 23.5 14.7 72.8 72.8 9.8 9.8 2.6 2.6 20.9 20.9 69.0 69.0 7.3 7.3 2.8 13.2 10.1 14.7 16.7 12.3 3.1 5.6 16.7 28.1 12.3 3.1 university or postgrad Malecompleted HoH Maleno HoH school no school some or completed primary
42.4 0.0
Female HoH no school % 35.2 36.8 31.1 64.0
53.9
9.5
distribution of households according to education level of household heads, by sex of HoH Female HoH
completed secondary
Proportion of households according to language spoken by income earners (multiple answers) 30.3 99.5
some or completed primary completed secondary completed secondary completed university or postgrad completed university or postgrad all HoH all HoH no school no school some or completed primary
43.7
15.4
88.9
Hindi
Proportion of households with rural background, by sex of HoH 64.7 82.2 80.5 38.2 85.1 1.6 38.8 86.4 2.9 36.4 82.7 0.0
Female HoH
Male HoH
All HoH
Proportion of households according to skills of its members (multiple answers) 7.0 19.1 21.7 23.8
Youth organization Womens group % % Womens group Neighborhood committee Neighborhood committee group Savings/ Credit/Self-help Savings/ Credit/Self-help group Refugees/migrants association Refugees/migrants Religious group association Religious group Community based organization Community based organization None None Other
No Skill 22.5 9.8 0.6 19.7 7.5 1.2 4.6 8.1 11.6 1.8 26.6 3.5 4.7 0.5 1.6 7.6 0.0 2.1 0.0 13.2 3.1 5.3 1.6 3.7 4.7 5.8 18.8
completed secondary completed university or postgrad
28.3
54.5
51.6
Computer
Cook
completed universityaccording or postgradto community % 21.1 17.0 2.8 Proportion of households groups participation of its members (multiple answers) Proportion of households according to community groups participation of its members (multiple answers) % 0.8 11.5 3.5 Youth organization
Tailor
Driver
Farming
Office job/clerck
Teacher
Interpreter
% to how they got their 0.5job Other of employed income earners according distribution distribution of employed income earners according % to how they got their 31.7job Networks/friends/relatives
no school
completed secondary
Mean
Median
Proportion of households with members unable to work due to health, by sex of HoH 25.0 9.4 11.6 25.0 21.8 31.6 56.2 64.5 61.4 55.6 30.6 41.3
37.2 24.1 % % % Relatives abroad or other part of India Aid/humanitarian organization Aid/humanitarian organization % % 24.1 9.4 9.4 12.6 12.6 0.0 0.0
0.0 16.7 16.7 0.0 0.0 15.0 15.0 35.0 35.0 Neighbors-friends in general Neighbors/friends of same ethnic group-origin Neighbors/friends of same ethnic Relatives abroad or other part of India group-origin
Female HoH
Male HoH
All HoH
81
82
Indian afghan Myanmarese somali church school Vocational training Bridge school national open school Proportion of children (5 -18 years) not attending school according to reasons (multiple answers)
2.8 38.4 0.5 9.7 18.4 16.9 afghan Myanmarese somali 15.2 9.8 17.6 8.6
Proportion of households with friends or relatives already in delhi before they arrived
Female HoH
58.8
Male HoH
65.0
All HoH
64.4
I. lengtH oF stay
Unit
Indian
Female HoH financial constraints % % % % % % % % % 0.0 0.0 8.3 0.0 15.6 0.0 0.0 0.0 0.0 0.0 0.0 18.8 25.0 6.3 0.0 0.0 36.1 11.6 0.0 5.4 2.0 0.7 2.0 0.7 58.3 50.0 64.6 child is working medical reasons Language issues 20.0 0.0 0.0 80.0 20.0 0.0 0.0 20.0 0.0
7.9
36.0
35.4
45.2
4-10 years
32.6
56.7
Proportion of heads of households working, by length of stay in delhi 15.2 50.0 28.6
40.0
discrimination/hostility distance/no school drop out/failure security not safe lack of documentation Male HoH financial constraints % % % % % % % % % 54.5 4.5 4.5 0.0 0.0 13.6 22.7 0.0 0.0 48.4 3.2 3.2 22.6 0.0 9.7 0.0 3.2 12.9
1-3 years 70.3 57.1 afghan Myanmarese somali 69.2 100.0 60.5 14.8
85.7
62.9
58.4
10.7
4-10 years
98.3
81.2
J. cHIldren edUcatIon
Unit
Indian
child is working medical reasons Language issues discrimination/hostility distance/no school drop out/failure security not safe lack of documentation Male HoH
Governmental school 19.9 22.9 38.8 0.5 11.9 9.0 12.4 0.0 0.0 2.8 18.9 5.7 13.5 22.2 0.0 4.8 16.2 4.5 24.3 9.4 29.7
56.3
34.0
English class
Church school
Vocational training
Bridge school
Male HoH
financial constraints child is working medical reasons language issues discrimination/hostility distance/no school drop out/failure security not safe
% % % % % % % %
Governmental school
46.1
42.8
English class
Church school
Vocational training
Bridge school
All HoH 7.0 17.3 24.4 37.6 8.4 4.2 5.1 16.2 40.8 12.7 23.8 28.6 19.0
lack of documentation
0.0
6.3
0.7
0.0
governmental school
50.6
38.8
english class
83