Is Voluntary Repatriation the Preferred Durable Solution? The View of Refugees in South Africa

Voluntary repatriation as the preferred durable solution for refugees has attracted much scholarly attention and the existing literature highlights the complexities of the process. This paper attempts to answer the question of when return can be considered as truly voluntary and preferred. To answer this question, the paper considers reasons offered by refugees themselves. The percentage of voluntary repatriation applications by refugees living in South Africa is low in relation to the number of refugees living in the country. However, reasons why refugees might choose to repatriate are still worthy of interrogation if we are to truly establish whether it is a preferred solution. Refugees’ answers illustrate that they are not always completely free to make choices. They may indicate consent, but consent does not necessarily indicate a preference. This paper in-corporates a study which reveals that refugees choosing to repatriate from South Africa are indeed very small in number, despite the vigorous attempts of the United Nations High Commissioner for Refugees (UNHCR) at promoting voluntary repatriation. It further reveals that this can be partly attributed to the fact that in an urban setting such as South Africa with a rights-based framework, refugees are often able to better integrate into their host society without the direct assistance of the UNHCR as they would do in a camp-based setting. As such, assumptions that voluntary repatriation is a preferred durable solution for all refugees, need to be interrogated.


INTRODUCTION
Voluntary repatriation is the voluntary return of refugees or displaced persons to their countries of origin. The number of refugees and asylum seekers seeking to be voluntarily repatriated to their countries of origin from South Africa, a country with a rights-based legal framework and a non-encampment or urban policy, is significantly small. For example, between 2004 and 2014, only 137 Angolan refugees living in South Africa were voluntarily repatriated to Angola (UNHCR, 2015). This is surprising, considering that 281 000 Angolans sought repatriation from refugee camps in neighboring countries including Namibia, Democratic Republic of the Congo, Botswana, and Zambia during 2004 alone. (Human Rights Watch Report, 2005:1). Equally small numbers of Rwandan refugees returned from South Africa (UNHCR, 2015) as opposed to the large numbers that returned from refugee camps elsewhere in Africa (UNHCR, 2003a). In 2003 the United Nations High Commissioner for Refugees (UNHCR) assisted 23 000 Rwandans in the Democratic Republic of the Congo (DRC) and a further 11 000 Rwandans in Tanzania, with repatriation (UNHCR, 2003b). Furthermore, in the last five years fewer than 100 refugees were voluntarily repatriated from South Africa to their countries of origin by the UNHCR (UNHCR, 2020). Despite these statistics, the United Nations High Commissioner for Refugees continues to promote voluntary repatriation as the preferred durable solution and actively campaigns to promote voluntary repatriation in South Africa.
The main purpose of this paper is to interrogate voluntary repatriation as a preferred durable solution for refugees living in an urban setting by considering the reasons offered by refugees themselves for wanting to return home. The paper begins with a brief discussion on the methodology used in this study and then presents the debates by scholars on voluntary repatriation as the preferred solution. The paper also provides an analysis of the Organisation of African Unity's (OAU) Article 5 of the 1969 OAU Convention Governing Specific Aspects of Refugee Problems in Africa (Hereafter the OAU Refugee Convention) (OAU, 1969) as it is the only international legal instrument that has provided for voluntary repatriation. Moreover, this paper discusses the UNHCR's approach and programs regarding repatriation from South Africa. Lastly, and most importantly, the paper analyzes the responses of refugees from data obtained over seven years (from 2008 to 2014) which consists of 254 applications for voluntary repatriation made to the UNHCR in South Africa. In addition to analyzing the responses from refugees, statistical data obtained from the UNHCR's regional office in Southern Africa is also used to illustrate that refugees from an urban environment do not regard voluntary repatriation as the preferred solution. Leading on from this, in order to demonstrate how the UNHCR operationalized voluntary repatriation within South Africa during this period, this paper offers a brief overview of two of the UNHCR's assisted repatriation programs for Mozambicans and Angolans.

Background
This study's desktop research reveals that the assertion by the UNHCR that voluntary repatriation is the preferred durable solution to ending refugee status, has been highly criticized by scholars. The focus of this paper is to ascertain whether voluntary repatriation is in fact considered to be the voluntary and preferred durable solution by refugees themselves. To answer this question, data was drawn from the 254 applications for repatriation that were made and received by a legal implementing partner of the UNHCR in South Africa. In South Africa, the UNHCR does not provide direct legal services to refugees and aslyum seekers; instead it has appointed local organizations and institutions to implement its mandate.The cases studied were all applications made over a period of seven years, from 2008 to the end of 2014, to one of the legal implementing partners tasked with the facilitation of the voluntary repatriation program, for the UNHCR.
It is important to bear in mind that the number of repatriation applications processed by the legal partner were minimal considering that the partner assists over 5 000 refugees per annum. The research was prompted precisely because the number of refugees seeking voluntary repatriation was so low, given the UNHCR's assertion that repatriation is the preferred solution and their targeted strategy of repatriation in South Africa among Angolan and Rwandan refugees.

Research design
Voluntary repatriation applications are conducted by the UNHCR's legal implementing partner. The application requires refugees to complete a single questionnaire (see Annexure 1). All applicants are informed that the data will be shared with the UN-HCR and that the UNHCR uses such data to study trends in South Africa. The results are used to inform its policies and develop its annual operational plan for South Africa. The requisite permission was obtained from the UNHCR to use the data from these applications and ethical clearance was obtained from the legal implementing partners' ethics committee to study the case files.
A simple coding of all the questions was done by focusing on key questions and the nationality of applicants. In particular, the study focused on the questionnaire questions that could provide insight into the reasons for seeking repatriation. The two key questions analyzed were: "What will happen if you return to your home country?" and "What change of circumstances explains that you no longer fear returning home?" As stated, 254 cases were analyzed. This is the sum total of the voluntary repatriation applications made at this legal partner. Applicants comprised of 14 nationalities, most from sub-Saharan countries. The purpose of the case review was to establish who sought repatriation, to determine their reasons for repatriation, and to establish the extent to which their repatriation was truly voluntary.
Additionally, further statistics were sought from the UNHCR's voluntary repatriation officer in Southern Africa. The statistics provided by the UNHCR proved particularly useful because the UNHCR keeps meticulous records of all voluntary repatriations that it facilitates throughout the country.

Limitations
Relying on the case files and the UNHCR statistics for a study of this nature could be considered a limitation, since they exclude refugees who have been assisted by other organizations. This study also does not include instances of spontaneous return by refugees, and neither the South African Government nor the UNHCR are able to provide statistics on spontaneous return by individual refugees. The lack of recorded information on instances of spontaneous return illuminates a gap in research and record-keeping in the South African context and it would be worthwhile for a future study to interrogate these cases further.
Since the empirical element of the research was conducted in an urban setting that utilizes a rights-based framework, these findings cannot address refugees in camp-based settings. Thus, a parallel study that seeks to explore why refugees in camps seek repatriation might also be worthwhile. The empirical review was drawn purely from application documents (see Annexure1), and thus follow-up questions could not be posed to applicants. A qualitative study on reasons behind repatriation can add further insight into whether repatriation is a preferred durable solution.
When assessing applications of this nature, it is also difficult to assess the truth-value of reasoning, as some applicants might have understandably tailored their answers with the end goal of wanting to go home.
Furthermore, the sample size used for this paper has implications on the ability to respond fully whether voluntary repatriation is a preferred durable solution. An additional study is required to ascertain the validity of such claim. Despite this, the small number of voluntary repatriation cases over the seven-year period (254) against the total number of refugee and asylum seekers who were provided with legal assistance (35 000) by the legal partner for this same period suggests that voluntary repatriation is not a preferred durable solution for refugees living in an urban-based setting, utilizing a rights-based framework.

DEBATING VOLUNTARY REPATRIATION
Repatriation has gained much scholarly attention for a variety of reasons. Much of the debate between scholars is centered on the UNHCR's policy of voluntary repatriation as its preferred durable solution for refugees despite the complexities associated with return, the lack of definition of voluntary repatriation, and the absent consideration of refugees' preferences. This section briefly considers these debates.
The UNHCR believes that there are three durable solutions: resettlement, local integration, and voluntary repatriation. The UNHCR asserts that this last option -voluntary repatriation -is its preferred option, as it believes that it is the only solution that re-establishes the bond between the refugees and their countries of origin, and between the citizen and the state (UNHCR, 2013a). This assumption is based on the UNHCR's understanding that refugees ultimately want to return home.
As refugees can never be forced or coerced into returning home, the UNHCR Handbook on Voluntary Repatriation states that the voluntariness of repatriation involves a refugee making an informed decision about the conditions of the country of origin and choosing freely to leave the host country (UNHCR, 1996). Thus, the refugee's decision must be informed and uncoerced. However, critics are quick to point out that voluntary repatriation might not be completely voluntary because of the complexities associated with return, that are "unforeseeable, neglected or ignored" (Bradley, 2008: 285-304). Bradley reiterated this position in 2019 when she stated, "When refugees do return, voluntarily or otherwise, they commonly face impoverishment and violence, sometimes resulting in repeated displacement" (2019: 154-173). Scholars even say that, depending on the circumstances, a refugee's decision to return to their country of origin is often mandatory, forced, or coerced and therefore involuntary rather than voluntary (Chimni, 2004: 55-73). For example, Chimni argues that a refugee's decision to repatriate may not be spurred by a desire to return home or a belief that the country of origin's conditions have become safe; instead, the need to repatriate is the only viable option available because life in the asylum state is unbearable. Under these circumstances, the decision to return home cannot be considered voluntary (Chimni, 2004).
Scholars also question the voluntariness of voluntary repatriation due to the relationship between voluntary repatriation and the processes of cessation, as outlined in Article 1C of the 1951 Refugee Convention (Hathaway, 2005: 917-953). According to Hathaway, while voluntary repatriation and cessation are two distinct processes, both processes return refugees to their countries of origin (Hathaway, 2005: 917-953). Cessation is invoked when, due to changes in circumstances in the home country, refugees no longer require international protection and cannot, therefore, refuse to avail themselves of the protection of their country (Siddiqui, 2011). Cessation allows for the withdrawal of refugee status by a host state and raises the question of whether repatriation facilitated after invoking a cessation clause can be considered voluntary (Stein, 1994). According to Hathaway, this sort of repatriation is mandatory rather than voluntary (Hathaway, 2005). Furthermore, if voluntary repatriation precedes the invocation of cessation, it raises the obvious question of whether the requisite level of safety in the country of origin for voluntary repatriation differs from that required for cessation. How is it possible for a country to be safe for voluntary repatriation, but not safe enough for cessation? Hathaway makes a legal argument that repatriation should not be possible where circumstances have not changed (Hathaway, 2005).
Assessing role players also indicates whether voluntary repatriation is truly voluntary. While host states, countries of origin, and the UNHCR may have roles in the voluntary repatriation process, refugees themselves are the most important role players in determining whether the process is truly voluntary (Stein, 1994). Goodwin-Gill and McAdam note that the promotion of voluntary repatriation anticipates varying degrees of encouragement by outside bodies. The motivation for the promotion of voluntary repatriation by the UNHCR raises various questions (Goodwin-Gill and McAdam, 2007:270-291). For this reason, Goodwin-Gill and McAdam assert that the UNHCR should not promote voluntary repatriation as the best solution; instead, the UNHCR should promote refugees' agency to choose for themselves (Goodwin-Gill and McAdam, 2007). Hathaway agrees and adds that the UNHCR's approach of "go and see" or "look and see" visits are an acceptable means of promoting repatriation, as it is carried out in secured conditions and without prejudice to the refugees' continued right to remain in their host countries (Hathaway, 2005).
The promotion of voluntary repatriation as the preferred solution also begs the question: preferred by whom -the host country, the refugee, or the country of origin? Harrell-Bond (1989) contends that assumptions around refugee conceptions of "home", possible confusion regarding nostalgia for the country of origin, and other factors are responsible for the push toward voluntary repatriation as the preferred solution (Harrel-Bond, 1989:41-70). Chimni debunks the underlying assumption that all refugees want to go home and argues that the UNHCR's acceptance of voluntary repatriation as the preferred solution must be understood within the context of developed countries that are attempting to contain and manage the global refugee problem (Chimni, 2004).
Scholars have also positioned their attention on the analysis of 'just return, ' shifting the focus to reintegration and the sustainability of voluntary repatriation (Black and Gent, 2004:1-23). Bradley thus introduces a model of 'just return' , which she admits cannot be neatly applied to all circumstances of return; still, Bradley's model is an essential demonstration of the country of origin's responsibility towards its displaced citizens (2008; 2019). She states that "just return is best understood as the restoration of a normal relationship of rights and duties between returnees and the state, such that returnees and their non-displaced co-nationals are rendered equal as citizens" (Bradley, 2008:1).
There is clearly no consensus regarding voluntary repatriation. However, it is simultaneously apparent that this method has a major impact on the lives of refugees. Considering these debates and the impact that voluntary repatriation has on refugees, the paper seeks to address two questions. Firstly, when can the return of refugees to their countries of origin be considered truly voluntary? Secondly, is voluntary repatriation the preferred solution for refugees themselves?

VOLUNTARY REPATRIATION AND THE OAU REFUGEE CONVENTION
Article 5 of the OAU Refugee Convention outlines the requirements for voluntary repatriation and explains how the process should be implemented (OAU, 1969) . Most importantly, Article 5(1) of the OAU Refugee Convention underlines the voluntary nature of repatriation, stating that "no one should be repatriated against their will" (OAU, 1969). Thus, consent from the refugee is a necessary element. Even though the UNHCR promotes and facilitates repatriation, the UNHCR's role should not influence the refugee's decision to return to his or her country of origin. Similarly, an invitation from a country of origin for return should not impact this decision.
Alternatively, when the governments of host countries fail to protect refugees, as evidenced by the xenophobic attacks in South Africa (Maharaj, 2018), the decision to return cannot be deemed voluntary. Voluntary repatriation can only be considered such if alternative durable solutions are available. Still, a refugee's consent is not always an accurate measure of voluntariness, particularly if the decision to repatriate is made under duress which can manifest in many ways. The questionable nature of voluntary repatriation was illustrated when the Australian government paid Afghan refugees sums of money to repatriate (Bialczyk, 2008: 1-23;Gerver, 2017:631-645).
Article 5(2) of the OAU Refugee Convention (OAU, 1969) requires 'cooperation' between the governments of the country of asylum and the country of origin to facilitate safe return. An important element of safe return is the formal recognition of the refugee as a national of his or her country of origin. On a practical level, this requires the issuance of proper documentation such as passports and corresponding cooperation of embassies. Although refugees may approach their respective governments to request return, the UNHCR is mandated to oversee the safe return of refu-gees when agreements are signed. In an urban setting like South Africa, it will require the refugees to either approach the embassies for a passport or seek the assistance of the UNHCR. While both processes demand agency from the refugees, it is not an indication of the voluntariness of their return or that it is a preferred solution.
Because there is an acknowledgment that refugees may have been displaced from their original homes and will require resettlement assistance, Article 5(3) of the OAU Refugee Convention (OAU, 1969) places an obligation on the country of origin to facilitate refugees' resettlement and grant them the full rights and privileges of nationals of the country. Despite the weight this provision places on the country of origin, there is no guidance regarding what is expected of states and their resettlement duties.
Refugees should not be penalized for voluntarily returning to their countries of origin, as outlined in Article 5(4) of the OAU Refugee Convention (OAU, 1969). This may be an important consideration in the refugees' decision to return voluntarily and should be provided in a concrete form -such as Amnesty Declarations -to safeguard against risk of persecution upon repatriation. Furthermore, refugees are entitled to assurances that new circumstances in their country of origin will enable them to return to lead a peaceful life. Such reassurances include effective governance and functioning administrative systems which are essential in cases where refugees have fled because of war or occupation.
Article 5(5) of the OAU Refugee Convention (OAU, 1969) recognizes that refugees being repatriated will need assistance in many different areas. When the UNHCR is present in a country, it is the first agency approached to facilitate repatriation and, given the terms of its statute, the UNHCR is obligated to facilitate return (Statute of UNHCR 1950: Art 8(c)).
Rutinwa argues that the OAU Refugee Convention's Article 5 is "much more about elaborating the principles and the modalities of effecting voluntary repatriation than a prescription of it as the only solution" (1999:1-29) and that it does not state that voluntary repatriation is a preferred solution. Article 5 does list many important requirements that must be met before repatriation can be considered or implemented (OAU, 1969) but most importantly, it underlines the voluntary nature of repatriation, stating that "no one should be repatriated against their will" (OAU, 1969: Article 5).

THE UNHCR'S ROLE IN VOLUNTARY REPATRIATION
The UNHCR is the main proponent that voluntary repatriation is the preferred durable solution. The UNHCR continues to promote voluntary repatriation in the recently adopted Global Compact for Refugees (UNHCR, 2018). Although return to one's country of origin may always be possible, repatriation implies facilitation of return and the UNHCR is mandated to facilitate voluntary repatriation in terms of its Statute (Statute of the UNHCR 1950: Art 8(c)). In a United Nations (UN) resolution on the 'Question of Refugees' adopted in 1946, the General Assembly stated: "the main task concerning displaced persons is to encourage and assist in every way possible their […] return to their country of origin" (UN, 1946). The UNHCR's ability to facilitate return, however, is limited and subject to the willingness of the states to cooperate. In the UNHCR Statute, the General Assembly specifically called upon states to cooperate with the High Commissioner's office in the performance of its functions, but most importantly, the UNHCR is urged to respect the international law principles of sovereignty when facilitating return (Chetail, 2004:1-23).
The UNHCR sets out four prerequisites for its involvement in the return of refugees to their countries of origin. First, if and when there is a fundamental change to the circumstances from which they fled; second, if the decision to return is truly voluntary and not based on lack of assistance in the host country; third, if tripartite agreements are signed between the host country, the UNHCR, and the country of origin; and finally, if the refugee is assured safety and dignity upon return (UNHCR, 1992).
The UNHCR does not hold the fundamental changes expected at the conclusion of a cessation agreement to be at the same threshold as those expected when assisting with voluntary repatriation (UNHCR, 2003). According to the UNHCR Handbook, voluntary repatriation takes place at a lower threshold; the basis of voluntary repatriation must merely be the general improvement in the situation of the country of origin, unlike cessation (UNHCR, 1992). Thus, there is an assumption by the UNHCR that return can take place in safety, even though the change in the country of origin is at a lower threshold than expected for cessation. With regard to the 'with dignity' statement, the UNHCR concedes that the concept of dignity is less self-evident than that of safety.
Even though the UNHCR has adopted a checklist approach to the complex concept of dignity, the concept not only seeks full restoration of rights, but it also includes exemptions for past violations. To the concept of safety, the UNHCR has developed a tripartite approach including the legal, physical, and material safety of those requesting repatriation. According to the UNHCR, physical safety entails a return to a secure environment in which there is protection, guaranteed freedom of movement, and access to land. Legal safety envisages equality with citizens while material safety implies access to humanitarian assistance. It is, however, difficult to envisage such conditions of safety and dignity in the absence of a fundamental change.
The UNHCR's involvement in voluntary repatriation ensures that there is a difference between unassisted return by the refugee and facilitated voluntary repatriation as envisaged by the UNHCR. Despite the justifiable criticisms of the UNHCR's facilitation of voluntary repatriation, the reality is that many refugees would not be able to return home without this assistance.

Introduction
This section explains the nature of repatriation from South Africa and who is seeking repatriation in South Africa before analyzing the case studies of voluntary repatriation from this country. All three analyses attempt to answer whether the UNHCR has satisfied its own safety and dignity standards, the extent of consent obtained, the voluntariness of return and whether voluntary repatriation is deemed the preferred solution by refugees.

The nature of repatriation from South Africa
In South Africa, applications for repatriation are made at the initiative of individual refugees through a UNHCR implementing partner. Regardless of whether a formal voluntary repatriation agreement exists between the UNHCR, the country of origin and South Africa, the implementing partner facilitates the completion of the application and the UNHCR voluntary repatriation officer will consider the repatriation request. Under the circumstances, the application form can be a form of consent, because there is no coercion from the UNHCR. The entire process starts at the request of the refugee.
The current application form, however, lacks direct questions about a refugee's reasons for leaving South Africa. Furthermore, there are no questions to directly determine the applicants' reasons for return to their countries of origin. Accordingly, applicants' voluntary and individual requests for voluntary repatriation cannot be the only measure of the voluntariness of their decision to return to their countries of origin. As previously noted, the UNHCR Handbook on voluntary repatriation assures return if the return is conducted with the consent of the returnee and with safety and dignity.
Instead, applicants are asked the following two questions: "What will happen if you return to your home country?" and "What change of circumstances explains that you no longer fear return?" These questions are meant to establish the applicant's safety upon return and, therefore, help satisfy the UNHCR's safety requirement. For refugees leaving South Africa, this is a legal necessity because Article 5 of the OAU Refugee Convention (OAU, 1969) demands the safe return of refugees. The UNHCR's approach to physical safety entails a return to a safe environment which requires freedom of movement, protection from attacks, and access to land (UN-HCR, 1996).
The application process of minors requires the consent of both parents; according to the implementing partner, every effort is made to ensure that families are not arbitrarily separated. This is a key component of dignity. Also, the UNHCR assists with acquiring national identity documents including passports and travel documents. The issuance of such documents by the country of origin is a clear in-dication of a refugee's acceptance as a national of his or her country of origin, and this compliance satisfies the requirements of the assurance of both dignity and safety (legal safety in this case). Furthermore, refugees repatriating from South Africa are received by the UNHCR field officers upon arrival in the country of origin to ensure a smooth transition. By offering such material and logistical assistance, the UNHCR can be deemed to be facilitating and promoting voluntary repatriation.

Who seeks repatriation?
The answer to this question does not lie in legal or policy frameworks, but in understanding the complex relationships between refugees and their countries of origin as well as refugees' attititudes towards living in exile. Migration theorist Egon Kunz has explored these questions in his research.
Kuntz categorizes refugees into three groups based on their relationships with the population of their home country (Kunz, 1981). The first group of refugees are alienated from their home country by some sort of event that is out of their control. The second group constitutes refugees who alienate themselves from their home country and the third group, usually the majoirty, holds a firm conviction that the events that caused them to seek asylum, are shared by most of their compatriots (Kunz, 1981).
The first group identified by Kunz, "events alienated refugees", consists of refugees such as social minorities or religious groups, who were marginalized and whose marginality came to the fore due to an event. These refugees are less likely to seek repatriation because they felt alienated from their homeland and faced discrimination (Kunz, 1981). The second group comprises "self-alienated persons" who have no wish to identify themselves with their fellow citizens, and thereby are the least likely to repatriate (Kunz, 1981). Kunz claims that it is the social relationship which may be a determining factor in their choice to repatriate (Kunz, 1981). The third group retains a strong bond with their home countries. Similarly, Rogge (1994) suggests that these refugees are also most likely to repatriate when the cause of their exile is removed.
The third group are typically refugees who fled because of war, external aggression, or events seriously disturbing the public order, situations they perceived to be intolerable. An analysis of Burundian and Congolese repatriates from South Africa bears some evidence of this theory (see Table 1 above). Of the 254 case files reviewed, all the Burundian and Congolese refugees who were repatriated were refugees who fled war, yet, they also retained social relationships with people in their countries of origin. This is evident from their responses which included: "there is stability in my country now, the war is over […] there is no more conflict between Hutus and Tutsis"; and "the war has ended". Close communication with family and friends provided reassurance of a change in conditions, and it is these bonds, the longing for their family, and their attitudes towards living in exile that drew them home. These bonds were particularly encapsulated by one applicant who stated, "I contacted my family and they told me it is safe now".
Kunz's theory demonstrates that it is a false assumption that the preferred solution for all refugees is voluntary repatriation. While there are refugees who retain a bond with their country of origin and are therefore likely to appeal to the UNHCR to facilitate their repatriation even if the threat of harm remains, this number is often small (Kunz, 1981). There are also refugees who are alienated from their home country and are less likely to want to return home. Irrespective of its applicability, Kunz's theory neglects to consider the dynamism of the life of urban refugees. Factors such as the number of years spent away from their countries of origin, the new bonds created in the country of asylum, and refugees' established sense of self-reliance may also contribute to whether refugees decide to repatriate. The next section explores these factors through the responses given by refugees in the case files studied.

Refugees' reasons for repatriating from South Africa
Data from the case files reveals that refugees' reasons for return are nuanced and individualistic. The answers provided confirm that even though refugees are not completely free to make choices as they are constrained by many factors, they are not 'passive role-players' in the process of repatriation. The responses demonstrate that when given a choice, the refugees are critical in their decision to repatriate. The data reveals that safety is as important a consideration for the refugees as it is for the UNHCR. As noted above, applicants sighted safety as a reason for wanting to return, noting that "the war has ended" and the 'political situation in the country is no longer so bad".
The two questions posed to applicants: "What will happen if you return to your home country?" and "What change of circumstances explains that you no longer fear returning home?" have been instrumental in gathering the refugees' reasons for return. For a better understanding of the refugees' reasons, the researcher grouped answers to the above two questions into push and pull factors. Although these broad categories exist, the findings indicate that the reasons for return remain intensely personal. The answer lies largely in the social relationship with the society back home (pull factor) and the refugees' attitude toward living in exile (push factor).
Their attitude toward living in exile can be gathered from the information below. These applicants were clearly unhappy in South Africa. For example, 167 of the 254 applicants asserted that they lived better lives in their home countries and, despite their best efforts, are victims of downward social mobility in South Africa. A refugee woman remarked, "in my country, I was a scientist, but here in South Africa I am braiding hair". Similarly, a refugee man remarked that in his country of origin, he was a respected teacher, but in South Africa, he is a car guard.
The data also indicates that all the applicants struggled to provide for themselves. Despite the right to work for refugees and asylum seekers in South Africa, the applicants have cited various obstacles to finding employment or decent employment and identified these obstacles as a major push factor. Eighty percent of male appli-cants also noted that they were only able to access low-skilled jobs and, therefore, struggled to support their families even when employed. Refugee applicants also cited constant xenophobia and were often referred to as job stealers. These interactions also factored into their belief that things would be better at home. It is apparent that this group had failed to benefit sufficiently from the rights-based framework in South Africa. Not one of the applicants were well employed or financially stable. One applicant revealed, "Life is hard here. I have no job and cannot find accommodation". Applicants also held the view that the South African government and the asylum determination system were directly responsible for their primary reason for repatriation, their inability to receive proper services at the Department of Home Affairs, and the fact that the refugee documents are not enabling documents. One Burundian male noted that he felt "worse than a second-class human being".
Moreover, longing for family was a pull factor for many women, a factor that was further emphasized through these women's pattern of maintaining strong family bonds. An astonishing 100 percent of the applicants had contacted family or friends in their countries of origin before seeking the assistance of the UNHCR to facilitate their repatriation. In each case, the UNHCR established contact with the relatives before facilitating repatriation. The application reveals the contact details of family members who had been contacted in the home countries. The UNHCR thus establishes contact with family or relatives listed on the application forms.
Thirty-two female applicants who were either abandoned by their husbands, victims of domestic violence, or single women with children, cited loneliness and the longing for family as reasons for seeking repatriation. In contrast, an inability to adjust to life in South Africa was not their primary reason for wanting to repatriate. The major pull factor for the women was evidently the need to be with family. At the time of application for voluntary repatriation, the applicants had already been in contact with family back home and had been assured of support upon return. They therefore felt they had a guarantee that they would not be destitute upon return. One applicant stated, "I want to go home to Burundi, my father has returned from the refugee camp and he will look after me". Another revealed, "My family told me to come home. My house and land are waiting for me". Another significant pull factor was the perceived security situation in the country of origin. Except for the Angolans, none of the other nationals had the assurance of safe return under the auspices of a voluntary repatriation agreement between their government, South Africa, and the UNHCR, yet they requested assistance with repatriation. Applicants assured themselves of safety upon return through contact with family or determined themselves that it was safe through engaging with the radio and news reports. For example, 53 of the 85 Burundians were single males, all of whom cited the end of the war as the reason for their return. This may well be the case for women as well, except that the majority of women also cited longing for family as a primary reason.
Though no one in the study stated love or attachment to their land as a reason for return, they did express wanting to be with their families. Some of these refugees chose to relocate internally in their country of origin with the help of family support. In seven cases, women and children chose to return because they wanted to be reunited with their families while their husbands, who still felt unsafe, chose to remain in exile in South Africa. For the women in these cases, the desire to be with family was stronger than safety considerations. Four of the applicants -one from Rwanda and three from Burundi -had fled during times of war and presumed family to be dead when they left. They were among those who wanted to return and join long-lost family members who had resettled in different areas. One can conclude that "home" for them was not considered a physical space but rather was associated with specific people. More precisely, it appears that for many of the applicants, return to family was the overwhelming reason for return. The applicants were all required to sign a declaration stating that they were returning to their home countries of their own free will. The declaration may indicate consent to return but it does not necessarily mean that it is their preferred choice of a durable solution; it could well be the only option available to them.

EXAMPLES OF REPATRIATION FROM SOUTH AFRICA: AN OVERVIEW OF UNHCR'S POLICY APPLICATION IN SOUTH AFRICA
In 1993, South Africa signed a Memorandum of Understanding (MOU) for the UN-HCR's involvement in Mozambique to facilitate the repatriation of Mozambican refugees. (Handmaker and Ndessomin, 2011: 145). This occurred before the introduction of refugee legislation or the ratification of international refugee conventions. The only formal voluntary repatriation agreement signed by South Africa since then was in 2003 for Angolan refugees. The UNHCR and Rwanda are currently considering a repatriation agreement, but South Africa has given no indication of its intentions.

Mozambican repatriation
The Mozambican refugees who were repatriated by the 1993 agreement between the UNHCR, South Africa, and Mozambique represent the only large-scale voluntary repatriation exercise that took place from South Africa (Polzer, 2007). These Mozambican refugees had no formal legal status in South Africa because the apartheid government failed to register them as refugees in South Africa. Accordingly, they were unable to assert their right to remain in South Africa. The UNHCR, the apartheid government, and the Mozambican government nevertheless signed a voluntary repatriation agreement. Handmaker and Ndessomin contend that false assumptions were made about the refugees' desire to return and that the MOU was signed on the basis that repatriation was the preferred solution (Handmaker and Ndessomin, 2011). It is evident that it was preferred by the signing parties, but there is no evidence that the UNHCR undertook any study or research to determine whether it was preferred by the Mozambican refugees themselves. Since this Mozambican exodus, the UNHCR has not been able to undertake a mass repatriation program of refugees from South Africa.

Angolan repatriation
The only formal voluntary repatriation agreement signed in South Africa since the introduction of refugee legislation pertained to Angolan refugees in 2003 (Handmaker andNdessomin, 2011). In addition, in May 2013 a cessation agreement was reached between the same parties. However, neither the voluntary repatriation agreement nor the cessation agreement has led to significant repatriation to Angola, as evidenced by the small number of Angolans (137) who have been repatriated since 2003 (UNHCR, 2015). Spontaneous return by individual Angolan refugees has not been recorded. There is no sign of any mass departures of Angolans during this same period.
This is the case even though the UNHCR has been actively promoting voluntary repatriation for Angolan refugees following the tripartite agreement in 2003 (Handmaker and Ndessomin, 2011). Information sessions were held for large numbers of Angolans in Cape Town and Johannesburg (Handmaker and Ndessomin, 2011). Though no governmental bodies conducted an independent study to assess the readiness of Angolan refugees to repatriate, the voluntary repatriation agreement was signed (Handmaker and Ndessomin, 2011). The African Centre for Migration undertook a study in Johannesburg which revealed that few Angolans were willing to return because they felt well-integrated in the South African economy and society. (Landau and Jacobsen, 2004). This can be contrasted with the refugees in the analyzed case files who sought repatriation because they failed to integrate.
The manner of Angolan repatriation from South Africa demonstrates the UNHCR's interpretation of voluntary repatriation as a stand-alone program, distinct from cessation. Even though the South African government signed a tripartite Voluntary Repatriation Agreement together with Angola and the UNHCR in 2003, it was only nine years later (in January 2012) that it recommended that refugee status cease through the invocation of the cessation clause for Angolans (Redden, 2006;UNHCR, 2012). This approach highlights the UNHCR's willingness to facilitate voluntary repatriation in the absence of a fundamental change in the country of origin. It is also indicative that the UNHCR regards cessation and voluntary repatriation as two distinct programs. Given their continued refugee status (from 2003 when the Voluntary Repatriation Agreement was signed to the invocation of cessation in 2012), Angolans did not face the withdrawal of their rights and remained beneficiaries of the rights to which they are entitled under the Refugees Act 130 of 1998 (RSA, 1998) and the South African Constitution (RSA, 1996). Because refugees in South Africa are not confined to refugee camps, they have freedom of movement and access to various socio-economic rights that have allowed them the opportunity for meaningful integration. These rights enable them to become self-reliant, as found by Handmaker and Ndessomin (2011) and Landau and Jacobsen (2004). The examples given above indicate that voluntary repatriation is a highly specific program facilitated by the UNHCR. As stated above, the example of Angolan repatriation happened both before and after cessation in South Africa. Hathaway argues that the voluntary repatriation that took place after cessation cannot be considered voluntary; at best, Hathaway considers this assisted return (Hathaway, 2005).
Considering Hathaway's argument, can it be said that this amounted to assisted return? Only 34 Angolan refugees of the 5 700 present in South Africa chose to repatriate after the signing of the cessation agreement in 2012. It is apparent from these statistics that no forced or mandatory return was instituted for Angolans after the signing of the cessation agreement. Because the UNHCR followed the same processes for voluntary repatriation after cessation in 2013, the UNHCR continued to refer to the Angolans' return as the voluntary repatriation program because only the refugees who requested repatriation from the UNHCR were assisted (UNHCR, 2015).
Given that the South African government granted Angolans the right to remain for an additional two years, it cannot be deduced that the Angolan refugees who returned after the cessation agreement were coerced into doing so. Although the UNHCR concluded the repatriation program at that time, it did not coerce Angolans to return by withholding assistance for return if they chose to stay in South Africa for a further two years on their tempoary residence permits. Thus far, the repatriation exercise facilitated by the UNHCR from South Africa to Angola cannot be considered anything other than voluntary.
It can also be safely inferred from the small number of refugees who repatriated, that the durable solution of choice for Angolans in South Africa (where they were able to integrate and become economically self-reliant) is not voluntary repatriation. However, it remains to be seen whether the South African government will choose to forcibly repatriate Angolan refugees who chose to stay in South Africa after their temporary residents permits expire (extended twice since the first issuance) and whether the UNHCR will be involved in this operation.

Promoting return amongst Rwandans
Of the 254 cases analyzed in this study, only 14 Rwandans (seven of whom were children) chose to repatriate with the assistance of the UNHCR over the seven-year period under study. None of the voluntary repatriation applicants expected to return to their original homes. Only one man returned, and the rest were women with children who had contact with their families in Rwanda and were expecting to return to them. In these cases, the social relationships with those back home were strong and provided the biggest motivation for wanting to return.
The UNHCR has recommended the invocation of the cessation clause for Rwandans (UNHCR, 2019) and has vigorously promoted voluntary repatriation despite refugees' claims that the circumstances that forced them to flee continue to prevail (UNHCR, 2011). Although the UNHCR has openly promoted cessation for Rwandans, the South African government has not indicated that it is willing to enter into a voluntary repatriation or cessation agreement with the UNHCR or Rwanda (Okoth-Obbo, 2015). While evidence indicates that Rwandan refugees who repatriate will either seek asylum once again or face persecution, like those who returned from Belgium, Uganda and Malawi, the UNHCR's strategy to promote voluntary repatriation and cessation is yet another example of the UNHCR overstepping (Whitaker, 2002). By doing so, the UNHCR should be concerned with what Bradley (2008: 1) has termed "just return". An approach of just return would consider the totality of repatriation including whether Rwandans would have an opportunity to restore their relationship as citizens with their state and whether they would face persecution.

CONCLUSION
Refugees choosing to repatriate from South Africa are indeed few in number. This is evidenced by the fact that the implementing partner only received 254 repatriation applications in a seven-year period, despite seeing 35 000 clients during that same period. Despite the UNHCR's vigorous attempts to promote voluntary repatriation for Angolans and Rwandans, numbers have remained small. This can partly be attributed to the fact that in an urban setting such as South Africa with a rights-based framework, refugees are not as dependent on the UNHCR for their daily survival. The few refugees who have chosen to repatriate from South Africa in this study came to seek the UNHCR's assistance to facilitate their repatriation of their own volition and, in many cases, this was due to a longing for family and perceived security in their home countries. In other instances, the refugees who returned were unable to forge a meaningful life in South Africa where they could enjoy financial autonomy or live without the fear of violence. For them, it appears that repatriation was the only solution, even if repatriation was not the preferred solution and or did not guarantee a better life. Given the small percentage of refugees who want to repatriate from South Africa, it is evident that this is not the preferred durable solution for many refugees. The refugee is the best person to decide if and when the time is ripe for return. The small number of refugees who chose voluntary repatriation from South Africa indicates that the refugees who were assessed within this study and who reside in an urban setting with a rights-based framework do not consider voluntary repatriation to be the preferred solution.