Impact of Social Network, Reunion, Health and Adjustment on Intention to Return: a Study of Sri Lankan Tamil Refugees in India

Home , Tamils

Impact of Social Network, Reunion, Health and Adjustment on Intention to Return: A Study of Sri Lankan Tamil Refugees in India

1. Introduction

Migration from one area to another in search of improved livelihood is a key feature of human history. Forced migration though has the elements of migration but its causes are different. The legal definition of a refugee is enshrined in the 1951 United Nations Convention Relating to the Status of Refugees. Article 1 of the Convention defines a refugee as a person residing outside his or her country of nationality, who is unable or unwilling to return because of a ‘well-founded fear of persecution on account of race, religion, nationality, membership in a political social group, or political opinion.

India has been receiving refugees from its neighbouring countries including Sri Lanka. The recent figure shows that in 2009, India has hosted around 456,000 refugees and asylum seekers that include 96,000 Sri Lankan refugees who are mostly Tamils (USCRI, 2009). Like many other countries, multi-ethnic phenomena had led to the rise of ethnic conflict between Sinhalese and Tamil minority group. Disagreement between Sinhalese and Tamil ethnic communities flared up when drawing up Sri Lanka’s first post-independence constitution in 1948. Several instances later occurred that discriminated Tamils such as the language policy of Sri Lanka, 1956, banning of importing of Tamil media literature in 1970, change of country name, etc.

In 1983, the ethnic conflict between Tamils and Sri Lankans in its serious discord led to civil and political unrest. After this the Tamil minority group started leaving their country seeking refuge in other countries. The first mass exodus was during 1983-87 where around 134,000 Sri Lankan Tamils arrived in Tamil Nadu. Due to the civil war in 1990, another 122,078 Sri Lankan Tamils came to Tamil Nadu during 1989-91. The third mass exodus started in 1996. During the period 1996-05, more than 22,000 Sri Lankan Tamils came to Tamil Nadu. The fourth mass exodus during 2006-07 recorded influx of another 19,680 Tamils. During 2007, there were about 97,708 refugees reside in Tamil Nadu of which 74,219 refugees reside in about 117 camps (Government of Tamil Nadu, 2007). The Sri Lankan Tamil refugees in Tamil Nadu are broadly categorised into following three divisions.

i) Camp Refugees – those who live in camps meant for refugees; ii) Non-Camp Refugees – those who have personal resources in Tamil Nadu and obtain special provision for accommodation to stay in Tamil Nadu; and iii) Special Camp Refugees – those who found to be a part of any militant outfits.

The study was conducted among camp refugees. Details about the camps selected for the study is presented in section 2.

Theories associated with migration cannot be directly related to the context of forced migration as they by and large state the economic motive behind migration. Whereas, theories associated with forced migrants are micro theories that are individual and family centric where survival is perhaps only motive of the migrants. The very definition of UNHCR

1 highlights that any movement for economic motive should not be considered as refugees. Hence, the general migration theories though cannot be directly associated with forced migration, to some extent they can be associated with the factors of forced migration such as push-pull factors, decision making, networking, social capital and ties (Lee, 1966; Bordieu & Wacquant, 1992; Massey, 1990; Thomas Faist, 1997 & 2000).

Particularly, Faist (1997) pointed out in his network theory where migrants are interdependent as the effect of one person or group of persons will have impact on another person or group. In addition, Lee’s migration theory (1996) can be seen in the context of forced migration where push and pull factors play role in determining the movement, (re)settlement and repatriation. Fussell (2012) highlights migration networks contribute to understanding the dynamics of differential migration and also help to predict future migration. Refugees look for opportunity and a ripe time to return to their origin and they get connected with their social connections and network of people to understand the environment in the origin and assess its conduciveness.

Fleeing from one’s country to another can involve stressful and challenging processes, which strongly influence the health, well-being and life quality of immigrants who have come to a new country with their desire to have a better and safer life (Beiser & Hou, 2006; Mui & Kang, 2006). They must deal with stressors associated with the processes of adapting to a new environment, which presents a multitude of challenges in their daily lives. Having psycho-social-cultural implications, those adaptation processes appear greatly stressful and demanding for refugees who are considered minor and less established because they do not seem to have established a strong and well organized community support and network of their own within the community in which they have settled.

Literature on the importance of social networks for the process of migration focuses on three main areas of migration experience. It examines the role of social networks in the migration decision-making process (Hugo, 1981; Ritchey, 1976); in the choice of destination (Massey et al., 1994); and in the adaptation of migrants in host societies (Caces, 1987). Koser (1997) defines these three areas of migration experiences as the three stages of the asylum cycle: pre-flight, flight, and exile. Social capital has been recognized as the networks of social relations that can provide people and groups with the access to resources and support. According to Granovetter (1973), these social relations can be understood as strong ties made up of family and close friends, and weak ties that are comprised of networks of acquaintances (or, using another terminology, bonding and bridging resources). Most people find themselves part of a dense social group, made up of family and close friends, as well as part of a circle of acquaintances. Each acquaintance will have his or her unique circle of close family and friends. Granovetter argues that the existence of one’s circle of acquaintances (weak ties) is crucial in bridging two or more densely knit groups of close friends and family.

In many countries, refugee and migrants’ associations are regarded as important for maintaining links with the native culture as well as for ‘voicing’ the needs and interests of specific groups within the multicultural milieu of receiving societies (Eastmond, 1998). The former process and the establishment of ‘ethnic networks’ in receiving societies are considered as essential at the early stages of settlement. They provide refugees with

2 emotional support, and a sense of roots and continuity (Eastmond, 1998; Bloch, 2002). Networking and communication with friends and relatives in Sri Lanka about the prevailing safety had helped the refugees to decide their return (Giammatteo, 2009). Thus it is necessary to review critically the relation between social network and social connections the refugees have in the host country and in the country of origin with their return outcome.

Communication is one of the basic necessities where human beings express their feelings with one another. Various studies highlight communication as strong factor that determine the adaptation of migrants and refugees in host society. Also, communication acts as a viable solution for the migrants who are stressed due to loneliness and difficulties. Through communication relationships and interpersonal networks, immigrants seek help for loneliness, stress and the difficulties that they encounter (Fogel, 1993; Jou & Fukada 1995). Interpersonal networks are also helpful in finding additional contacts (Jou & Fukada 1995). These contacts often tend to expand ones network within refugee circle as well as outside refugee circle.

The article 12 of The Refugee Convention (1951) states the importance of family relationship among refugees and urges the states to work towards bringing the refugee families together. The Convention states “the personal status of a refugee shall be governed by the law of the country of his domicile or, if he has no domicile, by the law of the country of his residence; and rights previously acquired by a refugee and dependent on personal status, more particularly rights attaching to marriage, shall be respected by a Contracting State, subject to compliance, if this be necessary, with the formalities required by the law of that State, provided that the right in question is one which would have been recognized by the law of that State had he not become a refugee.”

Refugee settlements can be understood in terms of the triangular relationships of what the refugee left, what developed and how they view their situations (Keller, 1975). For instance, the repatriation of Sri Lankan refugees who arrived in India during 1983-87 was widely criticised as it was political decision and not based on ground realities. Also, certain refugees were forced to return reluctantly because of the condition of their camps (Sukumaran et al., 1992). Black and Gent (2006) argues that post conflict return is a highly politically charged process in a number of context, both for returning and those who did not migrate or flee, leading many observers to question the notion of an unproblematic return home. They further note, especially doubt remains both about the conditions and voluntariness of return, the ability of individuals’ return to reintegrate in their home countries and regions and the wider sustainability of the return process.

According to UNHCR, local integration is a complex and gradual process which comprises distinct but related legal, economic, social and cultural dimensions and imposes considerable demands on both the individual and the receiving society. Finding a home in the country of asylum and integrating into the local community could offer a durable solution to the plight of refugees and the opportunity of starting a new life. Families in which members were left behind in war zones, camps or unsafe environments often experienced adjustment and attachment difficulties when reunited.

Need for the study

Studies relating to forced migrants including refugees globally and in India are abundant. However, studies on post conflict measures are scarce. In particular, studies linking factors such as adjustment, health status, social capital and the factors influence their decision to return to the country of origin where conflict has been halted. For instance, normalcy declared in Sri Lanka in the mid 2009 after the extinct of militant group. It has initiated talks about repatriation of Sri Lankan refugees. Being developing country India has limitations to accommodate refugees from different parts of the world and thus cannot afford to provide citizenship to Sri Lankan Tamil refugees despite of their ethnic ties with India. So this acts as a push factor in the host country. On the other hand, the Sri Lankan government has ensured sustained peace for refugees and displaced. Humanitarian agencies are also taking proactive measures in resettlement process and ensuring peace. These all acts as a pull factors in the origin.

Given this situation where the protracted refugees are in indifference between choosing their country of origin and host, it is essential to study the factors which affect their decision to return. Also, these people are expected to have well integrated with the host population which is identical in terms of culture, language, etc., it would be an opportunity to study their decision to return. Further, social relations in the host acts as a social capital in feeling the belongingness of the refugees in the host. Studies show that when refugees settle in host, they keep a close watch on the conditions in the origin and a conducive environment if they have intention to return. Also, those with adequate wealth and family members left behind in the origin would more likely to look for such opportunity to return. Given the adverse experience they had during their first phase return and the dejection caused of it would influence their decision to return. Given the experience the refugees have in the past, their decision towards return will also be rational and mindful of the situation in the origin. Thus, the study is aimed to investigate the aspects that influence the decision on return.

Objectives

1. To examine the socioeconomic relationship between individuals within the refugee population itself, and the socioeconomic relationships between the refugees and the surrounding host population; 2. To investigate the family reunification among refugees with their social connections; 3. To examine the mental health status among refugees; 4. To understand the adjustment pattern among refugees; 5. To examine the factors determining intention to return to origin and stay back in the host.

Hypothesis

- Refugees with longer duration of stay in host country have better social network. - Refugees with more number of movements have better social network. - Refugees with better social network would have good health status. - Refugees with longer duration of stay in the host country have better adjustment. - Refugees with strong social network have better adjustment.

- Refugees with better adjustment with host population have lower intention to return.

Conceptual framework

Reunion, mental Adjustment to health status host society

Social Intention Other factors connections and and confounding network to return variables

Study design, sample selection and techniques used for data collection

This section presents the detailed methodology used to carry out the study including research design, study area, sampling design, sample size, selection of respondents, techniques of data collection and types of tools, data collection and field experience, data entry and processing, analysis and statistical techniques applied and indices used and section plan for the thesis. This section also portrays in the detail the field experience and approvals obtained for the study.

The study used descriptive research design using blend of quantitative and qualitative capsules of data collection. In 2007, the department published an operational manual for various Government Officials, who are associated in any way with the refugees, to lay out a clear pathway to the service of refugees. A guideline for the officials working for refugees has been brought out by Department of Rehabilitation, Government of Tamil in 2007. This provides information on the number of camps spread across in various districts of Tamil Nadu along with basic demographic information. Refugee population universe is obtained from this guideline and sample camps were selected out of it. The samples collected for quantitative data is 288 refugee households from each one individual in the adult age has been selected. For qualitative data, In-Depth Interviews (IDIs) were conducted among 17 respondents. Quantitative technique includes semi-structured questionnaire while the qualitative capsule uses in-depth interview method based on few guiding questions with necessary probing into it. Infrastructure surrounding the camp has been collected additionally.

Plotting of maps and indices used

Maps were used in several instances in the study to plot the geographies such as selection of districts, Tamil diaspora across world and place of origin of respondents. A database management technology called DevInfo has been used to customize the study data and plot maps (DevInfo, 2005).

Infrastructure index

Composite indicators are widely used to assess the state of development and disparities among various regions/groups. Normalization technique is used in social science and in various studies when the indicators are of multiple dimensions and regions are the comparable units (Nanda et al, 2007; Pasquale et al, 2010). For positive indicators that would have direct relation with infrastructure availability, the procedure given below was used for normalisation. (V i – V min) / (V max – V min)

For negative indicators that would have inverse relation with infrastructure availability, the following procedure was used for normalization.

(V max – V i) / (V max – V min) where, ‘V’ is the selected infrastructure variable of interest ‘i' is the index value of the selected infrastructure variable ‘max’ is the maximum value among the range of values for a variable ‘min’ is the minimum value among the range of values for a variable

After normalization, the results of different indicators are summed to arrive the “Infrastructure Index”. From the summed up index value, the poor status of camp infrastructure is defined when the index data value is close to ‘0’ and good infrastructure is defined when the index data value is close to ‘1’. Thus, all five camps are rated using this technique based on different infrastructure available in and around the camp.

Social network index and refugee adjustment index

While different technique is used to construct relative index at the regional level, a different approach needs to be applied for constructing index for unit such as household or individual. One of the strongest methodologies applied calculating household level index is articulated in National Family Health Survey (IIPS and ORC Macro, 2000).

An attempt is made to understand social network of respondents with their social connections. Hence, connections at three levels have been taken into account while constructing the index – connections within refugee circle, connections in origin, connections with the origin, mutual contact with one another and connection established with natives after coming to India. Considering that these variables have its own importance in explaining the social network among refugees, all the ten variables are given equal

6 weight. The index value for social network would range from 0 to 10 where ‘0’ would mean absence of any social network while ‘10’ would mean strong social network. Using same method, an index called “Refugee Adjustment Index” (RAI) has been developed. This is a composite measure that would differentiate the refugees by their level of adjustment on social, cultural and economic variables.

Statistical techniques applied

In the present study univariate, bivariate, and multivariate techniques have been applied. Univariate analysis shows only the percentage and frequency distribution; on the other hand bivariate analysis provides only the association between two variables under investigation and Chi Square and Fisher’s Exact Test provides the statistical significance of association between the two variables. Along with the univariate analysis, mean and standard deviation have been calculated wherever required and appropriate. Multivariate technique like binary logistic regression has been employed to find association between dependent variables and its explanatory variables mentioned in research hypothesis.

2. Global Tamil diaspora and profile of selected camps and study respondents

People who leave their traditional ethnic homelands and live in other parts of the world are called as “Diaspora” (IOM, 2004). Recent statistics published by the United Nations (2013) show that 232 million people live abroad in other countries. Tamil population is originally Dravidian which mainly lives in the southern part of India in the state of Tamil Nadu. Its mother tongue is Tamil. The total population in India which speaks Tamil is 60,793,814 (ORGI, 2001). Migration of Tamil population during the colonial period was mainly because of cheap labour. Movement of the Tamil population also took place during the post-colonial period. Recent data on world Tamil population reveal that significant numbers of Tamil people are found in 23 countries, including India. Next to India, a vast majority of Tamil population has settled in Sri Lanka counting above 31 lakh, followed by Malaysia with around 19 lakh.

Sri Lankan Tamil population has a long history way back to the 2nd century BC (Mahadevan, 2002). Tamil people are classified under two different groups in Sri Lanka. The first group is called Sri Lankan Tamils, Ceylon Tamils or Eelam Tamils. The second group is called Indian Tamils, Hill country Tamils or Up-country Tamils. They are the Tamils of Indian origin sent to Sri Lanka during the 19th century to work as bonded labourers. These immigrants are considered as ‘indentured labourers’ who were bound by a contract and it was almost impossible for them to breach it. Though the country has seen an increase in Tamil population in the early 20th century, the proportion started declining after 1946. The recent census of Sri Lanka (2012) shows that 15 per cent of Sri Lankan population is composed of Tamils.

Profile of refugee camps

The camp in Krishnagiri district is large in size, followed by the camps in Dharmapuri and Villupuram districts. On an average, each family has 3 to 4 members. Almost all the selected camps have government hospitals in their vicinity, i.e., within five kilometres. Except

Villupuram camp, other camps have private hospitals in their vicinity. All the camps are connected to tap water for drinking along with storage tank. Majority of households do not have bathroom and toilet facilities within the premises but the government and NGOs have provided them. Salem camp has better facilities with more bathrooms and toilets than other camps. The assessment of camps based on greeneries and cleanliness shows that Salem camp is very good in terms of environment while Thiruvanamalai camp is the worst. The cleanliness in Villupuram camp and the greeneries in Dharmapuri and Krishagiri camps found average as compared to other camps. Except Villupuram camp, other camps are situated adjacent to metallic road with regular public transport. In all the five districts, post offices are located close to the bus stand and hence they have the same distance as the bus stand.

Out of the 17 variables collected on infrastructure, seven variables dropped while 10 variables given in table 3.5 were used to arrive at the indicator. All the variables used in the formation of index were standardized with units such as kilometre, per person, minutes, binary coding and five point scales. Further, five of the selected variables denoted negative relations with infrastructure while the rest five showed positive relation. The “negative relation” denotes that the lesser the data value, the better is the camp in terms of infrastructure availability. This is vice versa in the case of variables with “positive relation”. The ranking of camps based on the infrastructure index shows that the camp in Salem district has the finest infrastructure compared with other selected camps followed by the camps in Dharmapuri, Krishnagiri and Thiruvanamalai districts.

Profile of respondents

The background information collected from the refugee respondents include place of origin, year of arrival in India, demographic characteristics such as age, sex, family size, education, language proficiency, marital status, religion, ethnicity and occupation. Nearly two-fifths of the respondents arrived during the first phase (1983-87) and fourth phase (2006 onwards). More than half of the respondents arrived during the second phase (1989-91), while only seven per cent arrived during third phase (1996-05). Almost all the respondents who came in the first phase and one-tenth of them who came in the second phase repatriated to Sri Lanka. This is after the signing of peace accord between India and Sri Lanka during 1987. Average family size of the surveyed refugee households is 4, ranging from the smallest of single member households to a largest of 9 members.

The classification of respondents by gender shows that three-fifths of them are male while the remaining are female. Average age of the respondents is 39 years while the youngest respondent is 18 years, while the oldest respondent is 72 years. One-tenth of the respondents are illiterates. Over one-third of them have completed up to primary level of education. Another one-third of them have secondary education. Respondents who have completed higher secondary and are either doing their graduation or completed it constitute only 10 per cent. All of them can speak Tamil and a vast majority can read and write this language. Only 30 per cent of them can speak Sinhala, while 12 per cent can read and 10 per cent can write this language.

Only above one-tenth of the respondents are unmarried at the time of survey. Almost three-fourths of them are married. One-tenth of them are either widowed or deserted. It was also found that single member households mainly constitute of widowed or deserted respondents. The operation manual of Government of Tamil Nadu (2007) states that the refugees are allowed to work between 6 am and 6 pm so as to earn additional income for the betterment of their lives. Refugees mainly work in the informal sector. Almost one- fourth of them work as painters, moulders and carpenters. One-tenth of them have their own business such as textile or grocery shop. Those respondents who do not work are housewives, elderly people, those incapable to work and students. Only two-fifths of the men are fulltime employed, while the rest are employed part time (out of 116 working men). In the case of women who work, an equal proportion is employed as fulltime and as part-time workers (50 per cent each out of 40 working women).

3. Social connection and network

“Social connections” are most common among human beings. In particular, irrespective of type of migration, the migrants try and build social connections in the host so as to have good adjustment and integration with their living environment. This relationship does not only restrict to the host, but it also extends to their origin to have lasting network with their home. On the other hand, social network is defined as the frequency of contacts with social connections (Cheung and Phillimore, 2013). Thus, social network and social connection go hand in hand. Studies on diasporas have shown the linkages between social network and social relationship influencing migrants’ behaviour and their improvement in the host. The associations and informal networks growing out of this unity can be used as a resource to solve the problems faced by the refugees in their new country of settlement (Wahlbeck, 1996; Bousquet, 1991; Gold, 1992; Steen, 1992).

One of the objectives of the study is to examine the socio-economic relationship between individuals within the refugee population itself, and the socio-economic relationships between the refugees and the surrounding host population. This section also tests the hypotheses “Refugees with longer duration of stay in the host country have better social network” and “Refugees with more number of movements (as refugees) have better social network”. So as to investigate the network within refugees, information was collected from the respondents on three broad network dimensions, viz., (i) Do you have any of your friends/ relatives/ family members living in Sri Lanka (as refugees)?; (ii) Do any of your friends/ relatives/ family members from Sri Lanka living in Tamil Nadu?; and (iii) Do you have any of your friends/ relatives/ family members living in Tamil Nadu who are Indian citizens? Within each dimension, additional questions were asked about the place they live, frequency of contact and mode of contact. All the respondents reported any of the three aforementioned network dimensions. Almost all the respondents reportedly have network in Sri Lanka (96 per cent). Two-thirds of the respondents reported knowing friends/relatives/family members from Sri Lanka living in Tamil Nadu as refugees (including the same camp and other camps). Only 44 per cent of the respondents reported knowing local Indian citizens.

Relations from Sri Lanka living in Tamil Nadu by background characteristics

Current age and education are the characteristics that are found to have statistically significant relationship in deciding network in Tamil Nadu. In case of age, the younger generation seems to be well connected with refugee network in Tamil Nadu as compared with their counterpart of older generation. In case of sex, both men and women have equal social relationship with other refugees in Tamil Nadu. Education has a varying result as less educated refugees have more social relationship within refugees. Those who arrived during the first phase and recent phases to India are more likely to have social relationship with other refugees in Tamil Nadu.

Relations who belong to and are citizens of India by background characteristics

In case of refugees having network with the local population and who are citizens of India, significant relationship is found in case of characteristics such as age, education, household income, language proficiency, phase of arrival and number of moves. Education has direct relationship with social network with local population. Younger refugees have significantly higher proportion of social relations from India as compared with older refugees. It could be assumed that because of their connection through education, they are involved in employment and other activities. Variation in education supports the earlier assumption. Higher the education, the better is the social relationship with the local population. In the case of phase of arrival in India, almost three-fourths of the refugee respondents who arrived in the first phase have local network as compared with 33 per cent to 38 per cent of those who arrived during later phases (1989 onwards). Also, more than one move as refugees have better social network with local population as compared with those who have had come to India for the first time. These results highlight that the strength of refugee lies when there is long duration of acquaintance in the host country and also when the movement is more frequent.

Strength of social relationship

Out of the three different relationships, an attempt is made to look into how many respondents have all the three types of relations, two of them, one of them and none of the three relations. Result shows that one-third of the respondents have all the three types of relations. Further, 43 per cent of them have two of the three relations, while 23 per cent have only one of the three relations. There were no respondents who did not have any of the three relations. Age, education, type of employment, household income, language proficiency, phase of arrival and number of movements as refugees are the characteristics that are found to have statistically significant relationship in deciding the three types of relations among refugees.

As compared with older refugees, younger ones in the age group of 18-29 years have all the three types of relations, and the result is statistically significant. Education shows marked difference in deciding the number of relations. While only one-fifth of the refugees with low education have all the three types of relations, nearly half of the refugees who have attained education of 12 years or more have all the three relations. More than half of the respondents in higher income group have all the three types of relations as compared with the lower income groups. Ability to speak, read and write Tamil also has significant result where nearly two-fifths of the respondents with this ability have all the three types of

10 relations. Of those respondents who arrived during the first phase to India, above three- fifths have all the three types of relations as compared with those who arrived during later phases.

Communication with social relations

Communication with social relations is another vital element that determines the intensity of social status among refugees in the host community. In the study, all the respondents were asked about the means of communication they have with their relations; i.e., (i) whether they contact the people in their connection; and (ii) whether the people in their connection contact them. More than three-fifths of the respondents reported having mutual communication between them and their relations in Sri Lanka, and between them and their relations within refugees. Only one-fourth of them reported having mutual communication between them and their relations in India.

Mutual communication with connections within refugee relations in Tamil Nadu

Age, phase of arrival in India and number of moves as refugees are the characteristics that are found to have statistically significant relationship in deciding mutual communication among refugee relations. As compared with older refugees, younger refugees, particularly in the age group of 18-29 years, have mutual communication with refugee relations. Further, those refugees who arrived during the first phase to India have mutual communications as compared with those arriving later. Above four-fifths of the refugees who arrived in the first phase reported to have mutual communication with refugee relations. Of those who have moved more than once as refugees, 71 per cent reported having mutual communication as compared with those who moved only once and who constituted 57 per cent.

Mutual communications with relations in Sri Lanka

Education, marital status, household income and phase of arrival in India seem to determine mutual communication with the relations in Sri Lanka. As compared with respondents with lower education, more than three-fourths of the respondents with higher education above 12 years or more have mutual communication with relations in Sri Lanka. Over two-fifths of the respondents with full time employment and nearly half of the respondents with income above Rs.6001 have mutual communication with relations in Sri Lanka.

Mutual communication with relations in India

In case of mutual communication of refugees with relations in India, the determining characteristics seems to be age, education, household size, type of employment, household income, proficiency in Tamil and proficiency in Sinhala which are statistically significant in explaining the pattern. A proper trend cannot be found in the case of household size as refugees in small families and large families seem to have better communication mutually with their relations in India. This trend is, however, difficult to explain.

4. Family union and mental health status

Migrants often tend to leave their families and friends when they transit to other destination. Those with economic motives transit away from families to have better economic conditions which acts as a pull factor. On the other hand, migrants who flee from the original place due to war and conflict leave their families and loved ones apart. The economic migrants and forced migrants have variably different stances that separate them away from their loved ones. While the former have possibilities of meeting and joining back their families, the later often do not know the existence and survival status of their loved ones which would invariably aggravate poor psychosocial and mental health conditions. Gindling and Poggio (2010) mention that the context of within a forced migration, the stress and challenges in which refugee families are required to leave loved ones behind with the desperate hope that they may be reunited one day. This predictably results in different outcomes from those of normal migrant families.

From the sociological perspective, families are considered as the basic building blocks that underpin our fundamental social structure and the most durable basis for imparting social values, customs, traditions, beliefs and languages between generations (Bogenschneider, 2002). Families have been perceived as the central element and most important aspect of peoples’ lives and as most intimate social environment where the children begin the vital process of socialising and learn how to survive and thrive in the world (DeFrain et al., 2008). This section tries to assess the association between reunion and psychosocial well-being of refugees using qualitative data. In addition, an analysis has been carried out to test the hypothesis that refugees with better social network would have good mental health. Binary logistic regression has been used to test this hypothesis.

Disconnection with loved ones

Separation with family and loved ones can be pervasively seen among forced migrants. The study enquired from the respondents, “Have you ever lost contact with any of your family members/ relatives/ friends due to conflict?” followed by the questions “Did you ever tried to contact them? How did you contact them?” Following this question, an open ended question was asked to understand their experience in getting through contact with their loved ones. Reasons were also explored in case any one cannot connect back to his loved ones. Almost two-fourths of the respondents lost connection with their loved ones due to conflict while they moved away from their country. Some of the anecdotes given below show the concern that the refugees have with regard to their family members and loved ones with whom they have lost their contact.

“I lost my daughters. I wrote many letters. But no response from them. I don’t know whether they are alive or not. I should not say and think like that though (Male, age 59)”.

“My husband tried to contact them several times. We never got them back. After my husband’s demise, I never thought of contacting anyone. Now I am alone and old. I will end my life like this (Female, age 65)”.

Attempt to re-establish connections with lost contacts

Of all the refugees who lost contact with their loved ones, nearly three-fourths of them tried to re-establish connection by trying to contact them. Irrespective of background characteristics, the result shows that almost all the refugees attempted to re-establish connections with their loves ones with whom they lost contact due to war and conflict. The qualitative discussion with respondents shows the reasons for their inability to contact their loved ones.

“We don’t have any contact. We lost touch with them. I lost my brother. Did not try to contact them (Female, age 41)”.

“We did not contact them as we separated long back and do not have their contact details (Male, age 22)”.

“… don’t know their address and contact details. We lost all our contacts (Female, age 53)”.

Mental health status

The most commonly used definition of good health is given by WHO which covers all the aspects of health such as physical, mental and social well-being. Physical health could be measured directly. The latter two aspects of health are often neglected given the difficulties in measuring them. Literature is abundant that shows positive association between forced migrants and poor health status, both physical and mental (Mariam, 2009; Salma et al, 2001; Verma, 2004). This study collected both mental and social health proxies which would affect the overall health status of the respondents. As part of the quantitative data, three common mental and social health issues namely sad/ depressed/ fear, uncertain future and refugee feeling were assessed using a four point scale which consists of “to a great extent”, “somewhat”, “rarely” and “not at all”. The respondents were asked “In the recent past, to what extent did you feel sad, depressed, fear or none of these?”, “In the recent past, to what extent did you feel that your future is uncertain?” and “In the recent past, to what extent did you feel you are a refugee?”

Result shows that nearly three-fourths of the respondents reported that have felt to a great extent that their future is uncertain. Further, two-fifth of the refugees reported to have had feeling to a great extent that they are refugees. Only around one-fifth of refugees have felt to a great extent sad, depressed and fear in the recent past. Some of the anecdotes given below indicates that deprivation from decent employment, external problems and lack of citizenship in India emerge as some of the reasons that make respondents feel as refugees.

“Though I am B. Sc. qualified, I don’t have any work. I could not get a job because I am a refugee. After completing my studies, I am still unemployed (Female, age 23)”.

“I have grown here. We don’t like going there. We want to save like our parents did for us. We would like to live as normal citizens here. I feel as a refugee when there are some issues that affect me and my family (Male, age 29)”.

“I didn’t get a job because I am a Sri Lankan refugee. I was teacher… I had to resign because of low salary (Male, age 21)”.

The bivariate analysis of respondents by feeling sad, depression and fear by background characteristics shows a statistically significant association with age of the respondent, education, marital status, household members, language proficiency in Tamil, language proficiency in Sinhala, phase of arrival in India and number of moves as refugees. The variables that have significant association with the feeling of uncertain future among refugees are sex, age, type of employment and household monthly income. Age, education, marital status, language proficiency in Tamil, phase of arrival and number of movement as refugees are the characteristics that have significant relation with refugee feeling.

Association of mental health with other determinants

Feeling uncertain about the future and refugee feeling show a strong association with those who reported to have felt sad, depressed and fear in the recent past. As compared with those respondents who did not feel uncertain about their future, those who felt so are almost three times likely to be sadness, depression, fear in the recent past. Similarly, those who reported to have had refugee feeling in the past are three times more likely to have encountered sadness, depression, fear. Other background variables that are found to have significant association with mental health are age and income. As compared with younger refugees, the older ones are less likely to have had a mental health problem. Similarly, those with higher income are less likely to have had a mental health problem.

A strong association is found among those who have felt sadness, depression, fear with the thoughts about uncertain future. Result shows that as compared with those who were sad, depressed and had fear are four times likely to have thoughts about uncertain future. Social network among refugees has no association with all three mental and social health variables, namely, sadness/depression/fear, uncertain future and refugee feeling. These findings helps us to reject the hypothesis that refugees with better social network would have good mental health.

5. Adjustment with the host population

Adjustment is seen as a complex and multi-faceted process. It is one of the important processes that the refugees undergo in the host communities. Barry (1981) points out that the studies on their adjustment process fall into two groups of which the first group of studies focus on their mental health, occupational adjustment, language, residence and community patterns, culture and identity problems, health, reactions to stress, relationships with natives, problems of women and children, and various measures of their progress and adjustment. Tamil Nadu is a friendly border to Sri Lankan refugees as they have their ethnic kins who have similar characteristics such as ethnicity, language, culture, etc.

It is often not possible for immigrants to transfer their occupational status and skills from the country of origin to the host country. This is further critical in the case of forced migrants and refugees. For the latter, getting a job in the host country is challenging due to their identity crisis. Thus, employment among refugees serves not just as a source of income

14 but it is the critical factor that leads them to move into the mainstream of society. Besides occupational adjustment, there are other indicators that refugees who tend to adjust in the host country are related to socio-cultural dimensions (Montgomery, 1996). The study by Ali et al. (2006) indicates that the cultural adjustment process is affected by different factors such as language, tolerance of host society, satisfaction, different occupational opportunities, social support, length of residence and finally the degree of orientation toward their origin as the most effective factor of cultural adjustment. The current study investigates two different adjustment patterns among the refugees, viz., occupational adjustment and socio-cultural adjustment.

Occupational adjustment

As a part of occupational adjustment, various questions were put before respondents such as primary occupation in Sri Lanka, primary occupation in Tamil Nadu, average monthly income, mode of searching job, how they learnt the work (in case of different work in Sri Lanka and Tamil Nadu). While 14 per cent of the respondents were agriculturist in Sri Lanka who had their own land, there were none in Tamil Nadu who own land and are involved in agriculture. While six per cent respondents were into fishing and allied activities in Sri Lanka, none of them was involved in this occupation in Tamil Nadu. Result shows that local network acts as a strong mode through which the respondents get employment. Of the 83 respondents who were working both in Sri Lanka and Tamil Nadu, almost 93 per cent of them were doing different job in Tamil Nadu as compared with their job in Sri Lanka. Of those, two-thirds of them started doing new work without formal training while one-third reported some assistance either in the form of training or through friends. The below anecdote shows the difficulties faced by refugees in getting into new employment.

“I am a fisherman. Painting was new to me. I hardly painted even my house in Sri Lanka. When I came here I did not have anything to do. I could not do fishing here. I almost had no livelihood. The only income we had was the fund from the government which is hardly enough to meet our requirements. So, I started searching a job outside the camp area in our locality. I could not go to distant places where I can involve in fishing because we have restrictions in this camp. We have to report weekly to get the incentives. So I had to learn painting (Male, age 38)”.

“It was difficult as the work is risky. I fell down from a height of 25 feet and was hospitalised. I was in coma for several months and recovered slowly (Male, age 55)”.

“My earlier work was in a chemical factory. I was exposed to chemicals. I found it tough to work. It created problem in my eyes. I always had burning sensation due to formalin. I worked there for three years and was paid only 50 rupees per day. After that I worked as a coolie. We did not have enough money when we came here. Due to poverty, I started looking for jobs. I learned all my work by doing the work. Initially some of my colleagues taught me how to work and later I learned (Female, age 30)”.

“We don’t have work here. I am now working in a different district. I cannot work there continuously because every week I have to come here for the roll call, if not, my ration will be

15 cut. My family will also be in trouble. My employer doesn’t like my coming here every week as this affects his work (Male, age 63)”.

Socio-cultural adjustment

To investigate the socio-cultural adjustment, respondents were asked whether they attend festivals and marriages in their locality. Further, the respondents were assessed on their language accent, whether purely Tamil accent, purely Sinhala accent or mixture of both. Over three-fourths of the respondents attend marriages in their locality while 70 per cent attend festivals. The assessment of language shows nearly two-thirds of the respondents predominantly have Tamil accent, while one-fourth of them have Sinhala accent. Just above one-tenth of the respondents have a mixture of both Tamil and Sinhala accent. Among the different background characteristics, sex, age, type of employment, proficiency in Sinhala and phase of arrival in India have a significant relation with the refugees who attend festivals in their locality. Characteristics such as sex, type of employment, proficiency in Sinhala and phase of arrival have statistically significant relationship with refugees who attend marriages in their locality. Sex, age, marital status, household size, proficiency in Tamil and phase of arrival are the characteristics of refugees which show statistically significant relation with those who have pure Tamil accent.

Refugee adjustment index

Refugee Adjustment Index (RAI) computed in this study is a composite measure of four variables of socio-cultural and occupational adjustment among the refugees. The variables are attend any festivals in the locality, attend any marriage in the locality, accent used while speaking Tamil and Household monthly income (continuous variable). The indicator value ranges from 1 to 22 where the value close to ‘1’ is relatively poor adjustment while the value close to ‘22’ denotes better adjustment. Results of mean refugee adjustment index shows that age, education and marital status have marked difference in representing adjustment among the refugees.

Association of mental health with other determinants

The background variables such as age and sex are significantly associated with adjustment of refugees. As compared with males, the adjustment among female respondents is low. Also, the refugees in older age group are likely to have low adjustment levels as compared with the younger ones in the age group 18-29 years. As compared with unmarried refugees, the adjustment level of widowed/deserted refugees is low. Refugee feeling is significantly associated with the adjustment levels of respondents. As compared with the respondents who do not have the refugee feeling, those with the latter are less likely to have adjustment in the host country. The respondents have better adjustment if their family members are married to the Indian citizens. The adjustment levels of the respondents are two times higher when their family members are married to native persons. Strength of social network gives an edge to adjustment levels of respondents in both the models. This helps in answering the hypothesis that refugees with strong social network have better adjustment.

Pathway of integration through marriage in host community

Integration within host country population has been recommended by UNHCR as one of the durable solutions for refugees. There are three types of integration pointed out in the pool of refugee literature, viz., economic integration, social integration and legal integration. The study came across an interesting finding which could be ascribed to pathway towards integration. That is the marriages of refugees with the local population who are citizens of India. There are few studies that talk about inter-marriage or marriage with host population as a social integration process that the refugees experience in the host country (Ager and Strang, 2004; Crisp, 2004). Finding shows that 15 per cent of respondents reported that their family members married native Tamilians who are citizens of India. The qualitative findings from those respondents reveal their expectation of becoming citizens of India wherein marrying Indian citizens has been perceived as one of the solutions for their refugee status.

“Our children are here. They are married to the local people here… So I want to live here (Female, age 53)”.

“My son is getting married to an Indian girl. I feel that he will be united with the local people here. I feel that if not my generation, my son's generation will have a relief from the name "refugee" (Female, age 48)”.

“… My brother-in-law married an Indian girl. We hope that he will become an Indian citizen (Female, age 35)”.

“We can manage here as all of us have grown here. Our daughter is also married here and she is happy (Female, age 46)”.

6. Intention to return

Return migration is defined as ‘movement of emigrants back to their homeland to resettle’. It must be distinguished from circular migration and re-emigration (Gmelch, 1980). In case of forced migrants, return is mainly upon the establishment of peace in the place of origin. The end of civil war in Sri Lanka has resulted in the initiation of talks on repatriation and resettlement of refugees and IDPs in their homeland. Many studies show that most of the forced migrants intend to return back to their country of origin. They also highlight different mechanisms adopted by the forced migrants to return back to their native place. Often governments initiate dialogue to undertake measures for resettlement and ensure peace that does not bring back the resolved crisis.

Intention of refugees to return

Three-fifths of the respondents wanted to settle in India, while the remaining preferred to return to Sri Lanka. Among different background characteristics, education, marital status, income, proficiency in Tamil, duration of stay, number of movement as refugees, and feeling of sadness/depression/fear have a significant relation with the refugees’ intention to return to Sri Lanka or stay back in India.

Association of social network, health and adjustment on intention to return

Binomial logistic regression has been used to investigate the factors that determine the decision to return. The dependent variable is dichotomous where ‘0’ stands for stay back in India, while ‘1’ stands for return to Sri Lanka. Mutual communication with social connections in Sri Lanka, poor mental health, possession of land in Sri Lanka and number of movements as refugees determine the refugees’ intention to return. Psycho-social and mental health plays an important role in deciding the return to Sri Lanka. Respondents who recently had symptoms like sadness/depression/fear are thrice more likely to return to Sri Lanka as compared with those without these symptoms. Though the odds of returning to Sri Lanka are high among those with feeling of uncertain future, the result is not statistically significant.

Refugees with low adjustment in host population are twice more likely to return to Sri Lanka as compared with those with high adjustment. The result is statistically significant. The findings also support in answering the hypothesis that the refugees with better adjustment with the host population have lower intention to return. The phase of arrival explains the duration of stay of refugees in India and its impact on the decision to return to Sri Lanka. As compared with the refugees who arrived in India for a longer duration (first phase), those arriving during the second phase and later are more likely to return back to Sri Lanka. Further, the most recently arrived refugees (1996 onwards) are more likely to return to Sri Lanka as compared with those who arrived during the first phase.

Reasons behind intention to return to Sri Lanka

The respondents reported almost the same opinion about their returning to Sri Lanka. While eagerness to see family members and relatives are reported by most of them, there is also expectation that they will have better lifestyle and their children will have better education and employment there.

Reasons behind intention to stay back in India

The refugees who intended to remain in India were asked the reasons why they want to stay back. While Lee’s theory (1966) explains the push and pull factors of migration, in this study the reasons that lead to refugees’ decision to stay back in India are termed as “halt factors” which could be applied to the context of refugees and other forced migrants. The key reasons stated by refugees for staying back in India are analysed from the qualitative data. They are broadly categorised as safety and insecurity, absence of resources in the country of origin, social support and expectation of citizenship and uncertainty in the country of origin.

7. Summary, conclusion and recommendations

This study was conducted when the time was ripe to effectively think about returning back to Sri Lanka as conflict that made them flee was halted and the condition was under restoration. Thus, the findings presented in this study could be strongly attributed to the decisions of refugees about their intention to return to Sri Lanka or stay back in India as refugees. Also, uncertainty about future remains low among the recently arrived refugees as

18 compared with the protracted counterparts. These result suggests the readiness of the recently arrived refugees to return back to their origin. Hence, the repatriation measure should take into account these preferences and gradually initiate the process of sending back to the origin.

Some of the research questions investigated in the study suggests strong association between duration of stay and social network; social network and adjustment; and adjustment with host population and lower intention to return. It is interesting to note the role of social network on adjustment which is one of the processes that indicates the pathway of assimilation. Though association was found between few variables, there are other research questions where the association could not be established such as more number of movements (as refugees) and social network; social network and mental health; and duration of stay in the host country and adjustment.

Extensive measures have been taken by Sri Lankan Government and UNHCR in the resettlement of refugees and IDPs. The most challenging task is the resettlement of IDPs rather than of refugees. Over time, it could be possible that the uncertainty would wean-off and the situation will improve, bringing in conducive environment for both of them to resettle in their own land. While a good number of refugees want to stay in India, this becomes a policy concern as their repatriation cannot be forced. Given this, the question arises whether the country would be able to continue supporting and providing aid to the refugees who stay back? Also, what is the solution in the long term? Important alternatives are given below.

1. There is a strong desire among refugees for settling in India. Providing citizenship to the most desired population could be a durable solution for them. 2. There is a huge population which does not know about its parental origin as it born in India or has spent its whole life growing in India. Indian foreign policy should be reviewed to accommodate this population to provide it citizenship. 3. There are refugees who have close links with India and Indian citizens. Providing citizenship to them could be a viable solution. 4. There are discussions at international level about shifting in refugee protection by UNHCR and integrating refugees to national development plans so as to develop livelihood opportunities and promote self-reliance among them, capitalizing on their skills and utilize their potential to contribute to the local economy. The Government can think about these matters when the time is ripe for necessary action. 5. For the refugees who want to live in India, the Government can continue providing aid to them and can periodically assess their interest in returning to Sri Lanka. Thus, it can gradually initiate the repatriation process as and when the people are interested to return.

While this study is expected to fill the gap in the pool of literature on refugees, it also has some limitations like any other studies. The foremost limitation is low sample coverage which was due to the sensitiveness nature of the study subject and denial of permission to get connected to this population. Though the results do not seem to have affected with low sample size, it still gives an impression that the findings lacks accuracy due to standard error of sample size. With 95% confidence level, the standard error would be 4.2 for an indicator

19 value of 10 per cent with the number of samples studies. For the same 10 per cent indicator value, with 400 samples, the standard error would have reduced to 2.9 in which case the accuracy level is high as compared with the earlier sample size.

The study does not provide comparison with other population such as general migrant population, IDPs or other forced migrants. It is limited only to the aspects of refugees. For instance, the household monthly income for a refugee driver or painter could be the same as of the person from general population who work in the same profile. On the other hand, the refugees would have had decent employability in the host country while the IDPs could lack such a decent employability being in the country of origin.

Another limitation is the study does neither focus on the political aspects surrounding the issue nor talk particularly on the militant group. There are various other issues which haunts the refugees in the host country such as gender and violence, physical health, sexual abuse, alcohol use among male refugees and its associated factors and consequences, etc. The study prevents such a broader investigation onto diverse aspects/issues and limits the researcher only to the issues highlighted in the objectives.

Some of the limitations highlighted in the previous section could add onto the scope of future research. However, there are certain aspects of the current study has scope for future research. For instance, studying the repatriates in the origin, about their adjustment in the initial days of return and later, reunion with social connections who would have moved away as IDPs, mental health status, enquiry into the decision that they made to return, etc. Studying on these issues could also lead to completing the limitations that the current research has. Also, the studies in the country of origin can be done immediately as well as after few years of settlement.

------x------x------x------

References

Ager, A. and Strang, A. 2004. Indicators of integration: final report, Queen Margaret University College, Edinburg. Ali A. Moghaddas, Amir M. Amiri and Ali Rahimi. 2006. “The process of adjustment/ acculturation of first and second generation of migrants: The case of nomad Qashqaee Turks immigrants in Shiraz-Iran”, Paper presented at European Population Conference, Liverpool, United Kingdom. Barry N. Stein. 1981. “The refugee experience: Defining the parameters of a field of study”, International Migration Review, Vol. 15(1). Beiser, M. and Hou, F. 2006. Ethnic identity, resettlement stress and depressive affect among Southeast Asian refugees in Canada, Social Sciences and Medicine, Vol. 63. Black, R. and Gent, S. 2006. “Sustainable return in post-conflict context”, International Migration, Vol. 44 (3). Bogenschneider, K. 2002. Family Policy Matters: How Policymaking Affects Families and What Professionals Can Do. Mahwah, Lawrence Erlbaum Associates.

Bordieu, P., & Wacquant, L. 1992. Invitation to Reflexive Sociology, Chicago: Chicago University Press. Crisp, J. 2004. “The local integration and local settlement of refugees: a conceptual and historical analysis”, Director of Policy and Research, Global Commission on International Migration, Geneva, Switzerland. DeFrain, J., G. Brand and D. Swanson, 2008, Creating a Strong Family: Why are Families so Important?, NebGuides, Lincoln, University of Nebraska. 2009. Department of Census & Statistics. 2012. A2 : Population by ethnic group according to districts, 2012, Department of Census & Statistics, Sri Lanka. Faist, T. 1997. The Crucial Meso-Level. International Migration, Immobility and Development, T. Hammer, G. Brochmann, K. Tamas and T. Faist (Eds), Oxford, Berg Publishers. Faist, T. 2000. The Volume and Dynamics of International Migration and Transnational Social Spaces, Oxford, Clarendon Press. Fogel, A. 1993. Developing through relationships: Origins of communication, self, and culture. Chicago: University of Chicago Press. Fussell, E. 2012. Space, Time, and Volition: Dimensions of Migration Theory, in M.R. Rosenblum and D.J. Tichenor (eds.), The Oxford Handbook of the Politics of International Migration, Oxford University Press. Gindling, G. H. & Poggio, S. Z. 2010. The Effect of Family Separation and Reunification on the Educational Success of Immigrant Children in The United States. IZA discussion paper 4887. April, 2010. Gmelch, G. 1980. “Return migration”, Annual Review of Anthropology, Vol 9. Government of Tamil Nadu. 2007. An Operation Manual for the Officials Working in Sri Lankan Tamil Refugee Camps, Department of Rehabilitation, Tamil Nadu. International Institute for Population Sciences (IIPS) and Macro International. 2000. National Family Health Survey (NFHS-3), 1998–99, India, Mumbai. Jou, Y., and Fukada, H. 1995. Effects of social support on adjustment of Chinese students in Japan, Journal of Social Psychology, 135. Keller, S.L. 1975. Uprooting and Social Change: The Role of Refugees in Development, Delhi: Manohar Book Service. Koser, K. 1997. “Social Networks and the Asylum Cycle: The Case of Iranians in the Netherlands”, International Migration Review, 31(3). Lee, E. 1966. “A Theory of Migration”, Demography 3(1). Mahadevan, I. 2002. Aryan or Dravidian or Neither? – A Study of Recent Attempts to Decipher the Indus Script (1995–2000), Electronic Journal of Vedic Studies, 8(1). Mariam, G. 2009. Sri Lankan Tamil diaspora: Contextualizing pre-migration and post- migration traumatic events and psychological distress. Faculty of Social Work, University of Toronto. Massey, D. S. 1990. "Social Structure, Household Strategies, and the Cumulative Causation of Migration." Population Index, Vol. 56(1). Massey, D., J. Arango, et al. 1994. “Theories of International Migration: A Review and Appraisal”, Population and Development Review, Vol. 19(3).

Montgomery. R. 1996. “Components of Refugee Adaptation”, International Migration Review, Vol. 30(3). Mui, A.C. and Kang, S.Y. 2006. Acculturation, stress and depression among Asian immigrant elders, Social Work, Vol. 51(3). Office of Registrar General. 2001. Migration tables D1, D2 & D3, Census of India, 2001. Salma, Petter et al. 2001. “No less vulnerable: the internally displaced in humanitarian emergencies”, The Lancet, Vol. 357 (9266). Sukumaran N.G., Saradhi V.P., Shivalingam R.R. and Bijoy C.R. 1992. “Repatriation of Sri Lankan Refugees”, Economic and Political Weekly, Coimbatore Human Rights Protection Forum. United States Committee for Refugees and Immigrants. 2009. Warehoused Refugee Populations, World Refugee Survey: India, (see: http://www.refugees.org/resources/uscri_reports/archived-world-refugee- surveys/2009-wrs-country-updates/india.html). Verma, M.K., 2003. “Development induced displacement: A socio-economic study of thermal power projects”, Man in India, Vol. 84 (3&4).