Integration in Afghanistan : a systematic review

(1)

Integration in Afghanistan: A Systematic Review

Marcus Tudehope

Masters Thesis Political Science Political Economy Track

Graduate School of Social Sciences

Thesis Supervisor: Darshan Vigneswaran Student Number: 12100544 Date of Submission: August 23rd₂₀₁₉

(2)

Note to Assessors:

This thesis was completed as part of an individual research project coinciding with a research internship with UN-Habitat Afghanistan. The original intention was to conduct a study of the adequacy of the ‘durable solutions framework’ on facilitating integration of displaced Afghans by collecting primary data on the living conditions of households who had participated in different interventions informed by the framework. Two missions to Afghanistan were planned, one in February 2019 to develop and pilot the data collection tool and a subsequent mission in April 2019 to collect data over a period of three weeks. The February 2019 mission was completed as planned, however in April 2019, deteriorating security conditions resulted in the UN Department of Safety and Security declaring that non-essential staff were not permitted in country. Thus data collection was not possible and at a late stage in the preparation of the thesis it was necessary to change focus and develop a new research question. Following a preliminary review of the literature it was determined that there was scope to conduct a Systematic Review, as this would address a gap in the scholarship and had potential to make a substantive contribution to the discourse. This methodology was applied as comprehensively as possible given the time and resource constraints. Constraints were acknowledged and accounted for; the form of narrative synthesis ‘meta-ethnography’ was applied, as it accommodated the heterogeneity of the evidence base and required the review to be comprehensive but not necessarily exhaustive. Limitations arising from this methodology and mitigation strategies adopted are discussed in the methodology in Chapter 4.

(3)

Abstract:

Purpose: Integration has emerged as the most widely used general concept for describing the objective of post immigration/post-displacement policies. Despite being the subject of substantial amounts of scholarship there remains considerable variance in understandings of what the term signifies. As one of the world’s largest sources of displacement and the site of the world’s largest assisted repatriation operation, Afghanistan currently faces immense challenges of integration. However, the lack of conceptual clarity complicates measurement and undermines the collection of accurate data on which to base integration policy. This study seeks to develop integration theory and practice through addressing the question ‘what does integration look like in the Afghan context?’ and propose a series of locally specific indicators. Methodology/Approach: A conceptual framework is established through a review of integration theory and a synthesis of a number of the most influential conceptualizations. That framework then provides the structure for a systematic review of the evidence base to identify, evaluate and integrate the findings of relevant studies addressing integration in Afghanistan. The meta-ethnography method of narrative synthesis is used, complimented by the results of an in-depth analysis of three key empirical studies in the evidence base.

Findings: The meta-ethnography resulted in a comprehensive overview of the evidence base, eight new indicators were found to be appropriate to add to the original ten in the theoretical framework giving explicit consideration to subsistence strategies, tenure security, shelter, access to state services, food security, a supportive policy environment, access to basic services and official documents.

Theoretical & Practical implications: The paper contributes to the integration theory by proposing an updated conceptual framework grounded in the lived experience of displaced Afghans. The synthesized conceptual framework presents a tool for the design and comparison of interventions. The meta-ethnography synthesis increases the cohesion of the evidence base, producing a broad and comprehensive view that can provide valuable direction to future research.

(4)

1. Introduction

Integration has emerged as the most widely used general concept for describing the objective of post immigration/post-displacement policies (Bommes et al 2009 p44). Despite being the subject of substantial amounts of scholarship there remains considerable variance in understandings of what the term signifies. Integration remains ‘a chaotic concept: a word used by many but understood differently by most (Robinson p118). As one of the world’s largest sources of displacement and the site of the world’s largest assisted repatriation operation, Afghanistan currently faces immense challenges of integration (IOM 2019). However, the lack of conceptual clarity complicates measurement of integration and undermines the collection of accurate data on which to base policy (Schmeidl et al 2010 p108). Meanings vary between locations, evolve over time and depend on the interests, values and perspectives of the people concerned (Castles 2002 p112). This thesis seeks to increase conceptual clarity by addressing the question of ‘what does integration looks like in the Afghan context; what indicators are

appropriate to measure integration in contemporary protracted displacement situations (2004-present)?’.

Integration in Afghanistan: The Body of Knowledge

The accommodation of refugees returning from abroad and internally displaced people (IDPs) has been one of Afghanistan’s most pervasive challenges over the previous four decades and remains highly salient today (Schmeidl 2009 p33). Reflective of the ‘chaotic’ nature of the concept, the body of knowledge on integration in Afghanistan is characterized by methodological heterogeneity, differing theoretical frameworks and conceptions of integration and a lack of common research objectives. Diverse methodologies and conceptual foci are not necessarily problematic in themselves. However, there is a risk that a lack of systematic research and lack of coordination between different knowledge brokers contribute to a situation where there is an insufficient body of social scientific knowledge for evidence based policy development (Castles et al 2002 p24).

A Systematic Review

Systematic Review (SR) is defined as a means to make a vast and diverse literature more legible, through a process that is scientific, minimizes potential for bias and is rigorous and transparent (Hagen Zanker et al 2013 p2). SR seeks to increase the cohesion of a body of knowledge by identifying, evaluating and synthesizing the findings of all relevant, high quality

(6)

individual studies addressing a research question (Siddaway 2014 p2). Petticrew et al. (2008) argue that there is scope to conduct SR in the social sciences in situations where a): there is uncertainty about the effectiveness of a policy and where there has been some previous research conducted, b) when it is known that there is a wide range of research on a subject but where questions remain unanswered, c) when a general overall picture of the evidence on a topic is needed to direct future research efforts and d) when an accurate picture of past research is needed to facilitate the development of new methodologies (p368). This thesis makes the case that the topic of integration in Afghanistan satisfies the above criteria. SR has the potential to increase the cohesion of the evidence base, contribute to the discourse and development of integration theory and provide valuable direction for further research.

This study employs the ‘meta-ethnography’ method of synthesis, defined as an interpretative approach to narrative synthesis that conceptualizes key themes and extends the findings of individual studies (Snilstveit et al 2012 p421). The traditional SR methodology has been adapted to include a deeper analysis of a number of key primary studies to complement the meta-ethnography. Three studies were selected based on their involving primary research with the displaced, a large sample size and a consideration of multiple domains of integration. Studies were then analyzed for their framing of the issues of integration, conceptual foundations, methods used and claims generated. The results of the analysis are then considered in conjunction with the results of meta-ethnography to culminate in proposing a set of locally specific indicators to measure integration in the Afghan context.

1.1.

Research Question

The proposed SR will focus on the following research question:

‘What does integration looks like in the Afghan context; what indicators are appropriate to measure integration in contemporary protracted displacement situations (2004-present)’?

In investigating the research question the SR has the following core objectives: • Comprehensively map the evidence base regarding the research question

• Synthesize findings so as to improve the coherence of the evidence base and contribute to evidence based decision making

(7)

• Contribute to integration theory through the development of a conceptual model specific the Afghan context

1.2.

Utility of the Research

Increasing the conceptual clarity around integration in Afghanistan has the potential to have meaningful implications for theory and practice.

The question of when displacement ends, or more specifically when IDPs [and returnees] are no longer considered vulnerable as a result of their displacement, is particularly relevant in the Afghan context and has significant consequences for the targeting of appropriate assistance and the development of advocacy strategies. (Schmeidl et al 2010 p21)

The SR will assist with contextualizing findings of individual studies into different elements of integration and elucidating information on their generalizability and practical application (Cho et al 2009 p432). The results will be useful for scholars and practitioners seeking to build on the evidence base and to identify which elements are the most reliable on which to base policy and practice.

SR differs from a traditional literature review in that it adopts a replicable, scientific and transparent process, providing an ‘audit trail’ of the reviewer’s decisions, procedures and conclusions (Tranfield et al 2003 p210). Rather than focusing on results of individual studies, SR aim to assess the best evidence available, collate findings and present them in a way that’s accessible and relevant (ibid). Whilst the methodology has been applied extensively to the social sciences, to date no SR regarding integration in Afghanistan has been conducted. Results of this study will fill a gap in the literature whilst contributing to improving the coherence of the evidence base. By focusing on contemporary Afghanistan, the research will apply integration theoretical frameworks to one of the most severe displacement contexts in the world. The results will assist with the development of theory and may be useful for consultation and comparison in other protracted displacement situations regionally and further afield in the Global South.

(8)

1.3.

Structure of the Thesis

This thesis will commence by contextualizing the research with a brief examination of the issue of displacement and the situation in contemporary Afghanistan. Subsequently a traditional literature review serves to outline the theoretical foundations of the study, examining salient themes in integration theory. The review of the literature will culminate in an operational definition of integration and conceptual framework to provide structure to the SR. The SR methodology is then discussed, including the rationale for its selection in the current study, limitations of the method and mitigation strategies employed in the research. Following this, findings of the review are presented, through a meta-ethnography method of narrative synthesis and a more in-depth analysis of three empirical studies. The thesis concludes with a return to the original research objectives including a discussion of the implications of findings, and recommendations for further research.

2. Study Context: Displacement in Afghanistan

The following chapter will contextualize the study, commencing with a discussion of the concept of displacement and clarification of key terminology. The Afghan context is then explored through an examination of the country’s history of displacement and the current situation.

2.1.

Displacement

Displacement refers to the forced movement of people from their locality, environment and/or occupational activities (UNESCO 2017). The issue was first recognized as one of global significance with the stipulation in the 1948 Universal Declaration of Human Rights that all individuals have the right to move in order to claim asylum (Article 14). The rights of the displaced were expanded and geographical limitations removed in the 1951 Refugee Convention and subsequent 1967 Protocol relating to the Status of Refugees (UNHCR 2010 p2). During the 1970’s there occurred a considerable shift in the location and direction of displaced population flows, concentrated in South America and Sub-Saharan Africa, with a substantial increase in flows from the Global South to the Global North (McNicoll 2005 p195). Subsequent decades saw gradual reform of the institutional framework to better respond to changing population flows (ibid). During the 1990s the issue of internal displacement, whereby the displaced have not crossed a national border arrived on the international agenda as an issue of growing concern (IDMC 2013 p1).

(9)

The displaced are typically categorized by the direction and extent of their movements into three classifications1_:

• Refugees: Those displaced that cross a national border for reasons of a well-founded fear of persecution, claim asylum and are recognized as a refugee by UNHCR and the host country.

• Returnees2_{Those that were refugees and have returned to their country of origin.}

• Internally Displaced People (IDPs): Those that are displaced but have not crossed an international border.

This study will use the broad designation ‘the displaced’ to refer to all three categories. Individual categories will be stipulated where differences are relevant to the discussion.

2.2.

Displacement in the Afghan Context

Over the past four decades, displacement has been one of the most pervasive issues confronting Afghanistan. A widely cited study conducted by the International Committee for the Red Cross in 2009 found that 76% of the Afghan population report being displaced at some point during their lifetime (p6). Of that 76%, 41% were internally displaced at least once, 42% were displaced as refugees and 17% were both internally and externally displaced (ibid). Following the fall of the Taliban regime in 2001, UNHCR engaged in the largest assisted repatriation operation in the world to date, repatriating approximately four million Afghans, predominantly from Iran and Pakistan (UNHCR 2005). Afghanistan held the position as the most significant source of forced migration in the world until 2014 when it was overtaken by Syria (UNHCR 2015 p7).

A 2010 study conducted by the Brookings Institute identified six major phases of displacement since 1978. They are briefly summarized below:

Table 1: Phases of Afghan Displacement

1_{The displaced are also frequently further grouped into sub categories based on the type of event that was the}

catalyst for their movement, however for the purposes of this study three groupings are sufficient.

2_{Note that in the Afghan context this moniker is generally applied to anyone who has returned from an extended}

(10)

Displacement Phase

Main Drivers of Displacement Nature of Population

Movement Phase 1:

1978-1988

‘Saur revolution’ coup d’etat led by the People’s Democratic Party of Afghanistan against the rule of President Mohammed Daoud Khan, supported by the Soviet Union, declaration of jihad against the government that was seen as a threat to Islamic values, precursor to the 1979 Soviet intervention, initially 80,000 troops sent in 1979 Predominantly external displacement, refugee outflows Phase 2: 1989-1995

Soviet conflict, civil conflict between Mujahedeen factions following the withdrawal of the Soviet forces in 1989 and overthrow of the Communist government in 1992, widespread destruction and war crimes ultimately contributing to the rise of the Taliban in 1994.

Both internal and external displacement and refugee returns following the Soviet withdrawal

Phase 3: 1996-2001

Displacement under the Taliban regime, clashes between the Taliban and Northern Alliance, significant drought induced displacement in 2000

New internal and external displacement

Phase 4: 2001-2002

Displacement in anticipation of the US led intervention following September 11 2001, anti-Pashtun violence following the fall of the Taliban, and fleeing arterial bombardments of US led coalition forces

New internal and external displacement

Phase 5: 2002-2004

Following the fall of the Taliban, UNHCR undertook the largest assisted repatriation operation in the world to date, simultaneously the period saw significant amounts of IDPs return voluntarily

Large scale returns of refugees and IDPs

Phase 6: 2004-Present

Protracted conflict, clashes between US led coalition forces and non-state actors, growing

New internal and external displacement,

(11)

strength of Taliban and other actors such as Islamic State Khorasan (ISK) active in Afghanistan’s East and Pakistan. Significant drought induced displacement

significant urbanization of displacement, significant contraction of international protection space, trend of increasingly internalized

displacement

(Reprinted from Brookings Bern Project on Internal Displacement, Tribal Liaison Office 2010)

The drawdown of the US-led coalition troop presence since 2014 coupled with the resurgence of the Taliban and to a lesser extent ISK has led to a recent spike in conflict induced displacement (IOM 2019). This coupled with some of the most severe droughts in decades have culminated in a situation where it is estimated that approximately 200,000 Afghans are newly displaced every year (USAID 2018). The following table summarizes the current displacement situation of various target populations according to the International Organization for Migration’s (IOM) Displacement Tracking Matrix as of December 2018.

Table 2: Afghanistan Displacement as of December 2018

Target Population Number of Individuals

Returnees from Abroad period 2012-2018 3,225,380 IDPs currently in host communities 3,529,971 Out-migrants fled abroad period 2012-2018 2,291,584 Out-migrants fled to Europe period

2012-2018

271,248

Returnees & IDPs currently living in tents/temporary shelter/open air

266,943

(Reprinted from IOM Displacement Tracking Matrix 2019)

Shifting geo-political relations in the region in recent years has led to a rapid contraction of protection space for Afghans in neighbouring countries (UNHCR 2015). In addition, Afghans were the second largest group after Syrians to arrive in Europe in 2015-16, with official figures of approximately 220,000 (IOM 2016). However, whilst Syrians are considered prima facie

(12)

refugees, Afghans must individually prove a well-founded fear of persecution. Indications from some of the most significant destination countries such as Germany and Austria suggest that as many as 60% of Afghan asylum claims will ultimately be denied (Potzel 2016).

Displacement of Afghans continues to exert a destabilizing influence regionally and further afield (Roehrs 2015). Displacement situations have a proclivity to become protracted, with the displaced either unable or unwilling to return to their places of origin and lacking viable options for resettlement elsewhere (UN-Habitat 2016 p5). Whilst in recent years ‘donor fatigue’ has contributed to pervasive contractions in the provision of support to the displaced (Roehrs 2015).

With the severity and persistence of the issue of displacement established, the discussion will now turn to a review of integration theory in the literature and identify an operational definition to be applied throughout the study.

3. Literature Review:

With the study contextualized, the following chapter will review the relevant literature, establish the study’s theoretical framework and position the research within the wider integration discourse. As the most widely used general concept for describing the objective of post-immigration/post displacement policies, integration has been the subject of considerable amounts of scholarship (Robinson 1998 p118). As a result of academic debates, three broad conceptualizations of integration have developed in the literature. The first strand of literature explains integration as an acculturation strategy, involving a process of mutual adaption between the displaced and hosts. The second strand of literature explains integration as a multi-dimensional process involving spatial, economic, political, legal and psychological factors among others. The final strand of literature explains integration as progress in a series of interdependent ‘domains’. The three strands of literature are summarized and evaluated for their contributions to the discourse. The culmination is an operational definition of integration informed by all three strands to structure the subsequent SR and analysis of the Afghan evidence base.

(13)

3.1. Integration as an Acculturation Strategy

Acculturation is defined as phenomena that result from individuals from different cultures coming into contact and the subsequent changes that occur among both groups (Berry 1988 p7). Conceptualizing integration as an acculturation process recognizes that there are two parties involved; new arrivals with their own characteristics and different levels of effort towards adaption; and the receiving society with different levels of acceptance and other factors that are conducive to integration (Penninx 2005 p142). Interactions between the two groups determine the outcome of the integration process, however the two parties are fundamentally unequal (ibid).

Berry (1988) situates integration among four distinct acculturation strategies based on the nature of interactions between a dominant and a non-dominant group (p2). Central to these interactions are the concepts of cultural maintenance and participation. Cultural maintenance refers to the extent to which cultural identity is seen as valuable and efforts are made for its preservation, whilst participation refers to the extent to which new arrivals become involved with and seek to emulate a dominant group (ibid p10). Four distinct acculturation strategies determined by the intersection of various levels of cultural maintenance and participation are outlined in Figure 1.

Figure 1: Model of Acculturation Strategies (reprinted from Berry 1988)

(14)

identity and participating in the dominant culture, the situation is described as integration, representing the optimal acculturation strategy (ibid p9). Such a process can only occur through mutual adaption of migrants, host communities and the State (ibid).

The emphasis on the two-way process of mutual adaption is a key contribution of conceptualizing integration as an acculturation strategy. The above model operates on the assumption that the non-dominant group has the freedom to choose their acculturation strategy, whilst in reality the dominant group may constrain choices or enforces a certain strategy (Penninx 2005 p142). This conceptualization also does not consider what ‘society’; at what ideological or spatial level new arrivals are integrating into (Phillmore 2005 p145). For example, if a new arrival only becomes an accepted part of an excluded sub-culture, it is debatable whether this could be considered integration or indeed a desirable outcome. Nevertheless, this conceptualization provides a useful theoretical starting point from which to move closer to an operational definition of integration.

3.2. Integration as a Multi-Dimensional Process

Critiques of conceptualizations of integration that are too broad to be operationalized are common in the discourse (see Kuhlman 1991, Nee et al 2001, Beversluis et al 2016, Grzymala-Kazlowska et al 2018). The potential for the term to be deployed rhetorically to mean whatever a dominant group wants it to mean is also recognized (see Kuhlman 1991, Penninx 2005). Multi-dimensional conceptualizations of integration attempt to address these issues, seeking to increase clarity whilst encapsulating the breadth influencing factors. There arose a consensus in the discourse that conceptualizations of integration must acknowledge-among others-the spatial, economic, social, political, legal and psychological contributing factors (Castles et al 2002 p118). An influential multi-dimensional definition of integration is provided by Kuhlman (1991):

If refugees [or returnees or IDPs] are able to participate in an economy in ways commensurate with their skills and compatible with their cultural values, attain a standard of living which satisfies culturally determined minimum requirements, socio-cultural change they undergo permits them to maintain their identity and adjust psychologically to their new situation, if economic opportunities or standards of living for the host community have not deteriorated as a result of the influx, if friction between the host population and the refugees is not worse than within the host community itself, if refugees do not encounter

(15)

more discrimination than exists between groups within the host community then integration has been achieved. (p8)

The above raises questions regarding the relative weighting of the different dimensions and how they interact with one another. However, the breadth and specificity of the multi-dimensional conceptualization makes significant contributions towards an operational definition of integration.

3.3. Integration as a Framework of Interdependent Domains

Attempting to elucidate relationships between different dimensions of integration led to the development of a number of conceptual frameworks (see Ager et al 2004, Kuhlman 1991, Penninx 2005, Philmore 2011). Among the most influential was Ager & Strang’s ‘Indicators of Integration’ framework, the aim of which is to ‘describe key outcomes in normative

understandings of integration, accommodating the diversity of assumptions and values in

different contexts whilst maintaining conceptual coherence’ (Ager et al 2004 p144). To this

end a large number of conceptualizations were reviewed and common themes examined to identify 10 distinct but interrelated indicators of integration (ibid p167). Indicators were then organized into a framework in one of the following four ‘domains’ (ibid).

• Markers and Means: referring to key areas of attainment in the public arena that are widely recognised as essential for successful integration. The moniker markers and means refers to the fact that achievement in these areas is at once an indication of successful integration as well as a potential means to support integration (ibid p169). Indicators under this domain include employment, housing, education and health. • Social Connection: referring to processes of individual and group connection that are

regarded as essential to integration. Social connection indicators include ‘social bonds’, referring to connections within a community (defined by for example, ethnic, national or religious identity), ‘social bridges’ referring to connections between new arrivals and the host community, including means by which new arrivals can participate in the host society, and ‘social links’, referring to linkages between new arrivals and structures of the state, such as government services.

• Facilitators: referring to key areas of cultural competence that remove barriers to integration and facilitate achievement in other domains. Facilitator indicators include language and cultural knowledge and safety and security.

(16)

• Foundation: referring to the expectations and obligations of citizenship; what new arrivals have a right to expect from the state and other community members and what is expected of them. The foundation indicator is labelled rights and citizenship (ibid p177)

Figure 2: Indicators of Integration Framework (Reprinted from Ager et al 2008)

The framework does not suggest a hierarchy of domains, or that integration should happen in a particular order, rather the domains are seen as inextricably linked, with shortcomings in one having implications in others (Ager et al 2008 p176). Furthermore, the authors note that the framework seeks to reflect normative understandings of integration, and that locally relevant indicators should be negotiated for each domain (ibid).

The contribution of the framework of interdependent domains is to synthesize a large number of conceptualizations of integration and to encapsulate ‘functional’ (employment, health, education etc.) and social/psychosocial elements (relationships, sense of safety etc.). This could also be described as encapsulating both the objective and the subjective elements of integration into a single framework. Critics of the framework have suggested that insufficient consideration is given to the role of the adaption of the host community and host institutions such as is outlined in section 3.1 above (Platts-Fowler et al 2015 p417). However, this conceptualization has the potential to provide an approach to measuring integration that addresses some of the challenges around complexity and multi-dimensionality and can facilitate the comparison of different interventions.

(17)

3.4. Operational Definition of Integration

The three strands in the literature considered above; integration as an acculturation strategy, integration as a multi-dimensional process and integration as a framework of interdependent domains, offer complementary perspectives. A synthesis has the potential to address some of the shortcomings evident when the conceptualizations are considered in isolation. The below definition employs a framework of interdependent domains as a conceptual foundation, with explicit focus on principles of mutual adaption. Additionally, as the focus of the research is Afghanistan, it is necessary to consider that the displaced and host communities are likely to have comparable living conditions. Thus as in Kuhlman’s definition above, explicit focus is given to integration not having a detrimental impact on the living conditions of host

communities.

For the purposes of this study, integration is defined as:

A process of mutual adaption between host communities and new arrivals, whereby new arrivals, without having an adverse effect on the standard of living of host communities are able to:

• Achieve public outcomes within the domains of employment, housing, education and health which are equivalent to those achieved within the wider host communities

• Are socially connected with members of a (national, cultural, ethnic, religious or other) community with which they identify, with members of other communities and with the functions of the state

• Have sufficient linguistic competence and cultural knowledge and a sufficient sense of security and stability to confidently engage in that society

The above definition as well as the four domains (markers and means, social connections, facilitators and foundation) and the ten indicators of integration identified by Ager & Strang (employment, housing, education, health, social bridges, social bonds, social links, language and cultural knowledge, safety & security and rights & citizenship) will be used to structure the subsequent SR of the Afghan evidence base. The discussion will then return to the research questions of ‘what does integration look like in the Afghan context; what are a series of locally

(18)

of SR.

4. Methodology: Systematic Review

With the study’s conceptual framework established, the SR methodology will now be discussed. The following chapter will define the approach and briefly outline its origins and the evolution of its application to the social sciences. The rationale for SR’s selection in the current study will be discussed and the research question outlined. Following this each step of the SR process is explained in detail.

Systematic Review

SR is defined as a means to synthesize all the existing evidence on a specific question using transparent methods to give the best possible, generalizable statements about what is known (Waddington et al 2012 p359). SR employs predefined, explicit, systematic methods that aim to minimize bias, to provide more reliable conclusions on which to base decisions (ibid). SR differs from a traditional literature review in that it adopts a replicable, scientific and transparent process, providing an ‘audit trail’ of the reviewer’s decisions, procedures and conclusions (Tranfield et al 2003 p210). Rather than focusing on results of individual studies, SR aims to assess the best evidence available, collate quality findings and present them in a manner accessible and relevant (ibid).

Origins of the Method

SR originates from the ‘evidence based movement’ that became prominent in the natural sciences in the 1980s (Cornish 2015 p263). During this period in the health sciences in particular, the pace of knowledge production was very high (Tranfield et al 2003 p208). The process by which knowledge was produced attracted considerable criticism among scholars and practitioners alike however, who argued that research tended to be ad hoc, piecemeal and poorly conducted (Peckham 1991 p367). A consensus developed that the amount of information being produced was unmanageable and that it was unlikely that practitioners or scholars possessed the time, resources or expertise to locate, appraise and incorporate all evidence into decisions (Chandler et al 2017 p5). The SR methodology was developed in response to this challenge, the 1990s saw the formation of several organizations with the aim

(19)

of establishing agreed and formalized procedures for conducting SR (Tranfield et al 2003 p208).

As SR gradually gained mainstream acceptance in the natural sciences the method began to be increasingly applied to the social sciences. This required a number of methodological adaptions to accommodate increased disciplinary and methodological diversity and the increased prevalence of qualitative evidence (Walker et al 2013 p1). Following in the footsteps of the natural sciences, the 2000’s saw the formation of a number of organizations aiming to mainstream SR in the social sciences, such as the International Initiative for Impact Evaluation, International Development Coordination Group and the Campbell Collaboration. The latter organization describes its objective as ‘to help people make well-informed decisions by preparing, maintaining and disseminating SR in education, crime and justice, social welfare and international development’ (Campbell Collaboration 2019). SR has now come to be seen as a key tool for evidence based policy making in the social sciences, considered by some to offer the most comprehensive statement about what works (Hagen-Zanker et al 2013 p5).

Systematic reviews have been in use in one form or another in the social sciences for many decades, and are increasingly being used to support practice and policy, and to direct new research efforts. (Petticrew et al 2008 p389).

4.1.

Rationale for the Selection of SR in the Current Study

Petticrew et al. (2008) argue that there is scope to conduct a SR in the social sciences in situations where a): there is uncertainty about the effectiveness of a policy and where there has been some previous research conducted, b) when it is known that there is a wide range of research on a subject but where questions remain unanswered, c) when a general overall picture of the evidence on a topic is needed to direct future research efforts and d) when an accurate picture of past research is needed to facilitate the development of new methodologies (p368). The following will discuss each of the above conditions individually in relation to the Afghan context and consider the appropriateness of the SR methodology.

Uncertainty about a particular policy & existing research

Integration has been the subject of considerable scholarship, yet there remains significant variance in understandings and usage of the term (Robinson 1998 p118). Conceptual ambiguity

(20)

undermines the collection of accurate data on which to base policy, as well as hampering the measurement of outcomes to gauge the effectiveness of different approaches (Schmeidl et al 2010 p108). In the context of the immense integration challenges confronting Afghanistan there remains considerable uncertainty regarding the optimal policy approach (See UN-Habitat 2016, Brookings Bern Project 2010, Amnesty International 2016).

Wide range of research conducted, remaining unanswered questions

As demonstrated in Chapter 3, integration theory has developed considerably over the past several decades. Questions of how best to define and measure integration in the Afghan context remain however. Whilst the question has been taken up in a number of sources of grey literature (See Brookings Bern Project 2010, UN-Habitat 2016) there remains scope to investigate questions around the applicability of theoretical frameworks to the Afghan context and locally specific indicators to measure integration.

Overall picture of the evidence on a topic to direct future research efforts

A preliminary review of the literature found that the knowledge base regarding integration in Afghanistan involves an array of knowledge brokers across different disciplines including academia, international organizations, donor governments and the Government of Afghanistan. As such, there is a significant amount of methodological heterogeneity, differing theoretical frameworks and conceptions of integration and a lack of common research objectives. Whilst these conditions are not necessarily problematic in themselves, they can contribute to a situation where it is difficult for scholars and practitioners to remain abreast of developments, and to effectively direct research efforts to build on previous achievements and avoid duplication. (Petticrew et al 2008 p389). Increasing conceptual clarity and considering the Afghan evidence base as a whole has the potential to identify the extent to which different elements of integration have received scholarly and practical attention and which areas potentially warrant further enquiry.

Accurate picture of past research to facilitate the development of new methodologies As above the breadth of contributors to the knowledge base can complicate obtaining an accurate picture of past research. SR through identifying, critically evaluating and integrating the findings of previous studies can shed light on the performance of previous methodologies. This as well as increasing the conceptual clarity around integration has the potential to inform the development of improved methods for its investigation/measurement.

(21)

In light of the above, it is here argued that the context of integration in Afghanistan satisfies the four criteria and that there is sufficient scope to conduct SR. To date no SR regarding integration in Afghanistan has been conducted. Results of this study will fill a gap in the literature whilst contributing to improving the coherence of the evidence base. The SR will assist with contextualizing findings of individual studies and elucidating information on their generalizability and practical application (Cho et al 2009 p432). Furthermore, the study will provide insights into the applicability of integration theory in one of the most severe displacement contexts in the world. The discussion will now turn to an elaboration of each step of the SR methodology.

4.2.

Systematic Review Protocol

The study commences with the development of a SR protocol. The SR protocol describes the rationale and planned methods of the SR (PRISMA 2015). The development of the protocol at the outset is important to make the review process as well-defined, systematic and unbiased as possible while maintaining a practical perspective (Campbell Collaboration 2001). The protocol contains the following:

1. Research question and objectives of the SR

2. Criteria for inclusion and exclusion of studies in the review 3. Search strategy for the identification of relevant studies 4. List of databases, journals and websites to be used

5. Explanation of how data will be synthesized, analyzed and presented

Figure 3: Systematic Review Methodology Flowchart

1. Research Question & Objectives

(22)

intervention and outcome to be considered. This study will focus on the following research question:

‘What does integration looks like in the Afghan context; what indicators are appropriate to measure integration in contemporary protracted displacement situations (2004-present)’?

In investigating the research question the SR has the following core objectives: • Comprehensively map the evidence base regarding the research question

• Synthesize findings so as to improve the coherence of the knowledge base and contribute to evidence based decision making

• Identify a set of indicators appropriate to measure integration in the Afghan context

2. Criteria for inclusion and exclusion of studies in the review

Establishing inclusion and exclusion criteria at the outset enables the identification of relevant material for the review, it also assists with transparency and ensures screening is conducted in a consistent manner (Hagen-Zanker et al 2013 p7). Inclusion and exclusion criteria are outlined below:

Inclusion Criteria:

• Population: Afghans that have experienced displacement as defined in Chapter 2

• Interventions or exposures: Any interventions targeting one or more of the ten indicators of integration outlined in the operational definition in Chapter 3, including:

• Employment, housing, education, health • Social bridges, social bonds, social links

• Language and cultural knowledge, safety and security • Rights and citizenship

• Outcome of Interest: Studies that provide some measure (qualitative or quantitative) for the above outcomes will be included

• Setting and Time Period: Afghanistan and abroad during the study period; ‘Phase 6’ of Afghan displacement (2004-present)

(23)

Exclusion Criteria:

• Resources outside the date range (pre 2004)

• Resources relating to interventions in the above areas that do not explicitly target displaced Afghans

• Studies in languages other than English3

3. Search Strategy for the identification of relevant studies

Given the nature of the evidence base, a search strategy was developed involving three distinct components; academic literature search, grey literature capture and snowballing. Searching was an iterative process, with additional relevant terms and sources of evidence becoming apparent as the search progressed. Each of the three tracks is elaborated below:

Academic Literature Search: Academic literature involves searching peer reviewed journals for relevant studies using a pre-determined ‘search string’ to identify relevant material. Search terms were identified through a preliminary literature review. A number of search strings and Boolean operators were piloted and the following was found to be optimal:

“Displacement” OR “Forced migration” OR “Displaced person” OR “Internally displaced person” OR “IDP” OR “Returnee” OR “Refugee” OR “Asylum seeker” OR “Integration” OR “Durable solutions” OR “Durable solutions framework” OR “Resettlement” OR “Local integration” OR “Repatriation” OR “Voluntary repatriation” AND “Afghanistan” OR “Afghans”

Placing the Boolean operator ‘OR’ rather than ‘AND’ before the term ‘integration’ and the various monikers for durable solutions serves to make the search string deliberately over inclusive for the initial round of literature capture.

Grey Literature Capture: Grey literature includes all those sources of evidence outside of academic peer reviewed channels. This includes donor reports, working papers, policy documents and evaluation reports. Grey literature was located through directly searching the websites of relevant institutions and knowledge brokers. This search is further supplemented

3_{The effect of this exclusion criteria on the number of studies is negligible. The vast majority of resources are in}

English, in cases where resources are published in local languages Dari and/or Pashtu this is typically a translation of a document originally produced in English

(24)

by using Google and GoogleScholar with the above search strings and Boolean operators. Using the GoogleScholar ‘cited by’ function showed where identified studies had been cited enabling the identification of further relevant literature.

Snowballing: Snowballing refers to seeking advice on relevant publications from key experts, which are then reviewed, including a review of the references of each publication. Both the selection of key experts and their recommendations are not objective, however this search method was included to get a sense of what literature has been influential in the field, which may be outside of traditional channels. Furthermore, snowballing resulted in the researcher gaining access to a number of unpublished resources. Given the topic’s specificity, the technique of searching reference lists of both academic and grey resources proved useful.

Six key experts with direct experience with the displaced in Afghanistan were identified including:

• Two Housing Land & Property Rights Expert with a UN agency with a development focus • One shelter expert specializing in assisting the displaced with an international NGO

• One protection expert from a UN agency with a humanitarian focus

• One programme manager from a UN agency with a development focus working with IDPs and returnees

• One programme officer from a UN agency with a humanitarian focus

Each expert was asked to identify 5-10 influential pieces of literature (academic or grey).

The process of the three search tracks is summarized in Table 3 below:

Table 3: Summary of Search Tracks Summary of Search Methods

1. Academic Literature Search • Anthropology Plus • Anthro Source • JSTOR

• International Encyclopedia of Human Geography • Scopus

• Social Science Citations Index • Web of Science Aggregator

(25)

• GoogleScholar Aggregator 2. Grey

Literature Search

• British Library of Development Studies • 3ie Database of Impact Evaluation • Social Science Research Network

• Afghanistan Research & Evaluation Unit • Afghanistan Analysis Unit

• UK DFID publications library • USAID publications library • UNilibrary

3. Snowballing Recommendations received from six experts for grey literature; sources not mentioned in the above list included:

• Durable Solutions Unit

• Regional Durable Solutions Secretariat • Durable Solutions Platform

• Samuel Hall Consulting • Save the Children

• Norwegian Refugee Council

• International Organization for Migration

4. Synthesis of Findings:

There are a number of methods of synthesizing the results of a SR, both qualitative and quantitative. Quantitative forms of synthesis whereby findings are pooled and analyzed statistically require a level of methodological consistency in the evidence base that is lacking in this case. The study thus utilizes a form of narrative synthesis; whereby text is used to summarize, explain and further investigate findings (Popay et al 2006 p92). Within narrative synthesis there are a number of different methods, distinguished by the extent to which each aims generate or test theories and the measures each employs to be as systematic and transparent as possible (Dixon-Woods et al 2005 p45).

This study employs the ‘meta-ethnography’ method, defined as an interpretative approach to narrative synthesis that conceptualizes key themes and extends the findings of individual studies (Snilstveit et al 2012 p421). The objective of meta-ethnography is to develop interpretive explanations and understanding from multiple cases of a given study phenomenon

(26)

by utilizing research which is grounded in the experience of participants (Burns et al 2010 p201).

Meta-ethnography has the advantage of being able to accommodate methodological heterogeneity and variations in context (ibid). The synthesis process involves determining how studies are related, identifying and synthesizing the most salient themes using a combination of the following methods:

• Reciprocal Translation Synthesis: when concepts in multiple studies are consistent and can be synthesized into overarching themes

• Refutational Synthesis: when concepts in multiple studies refute each other or otherwise advance inconsistent perspectives

4.3.

In-depth examination of key primary studies

Although not part of a traditional SR, the decision was taken to include a deeper analysis of a number of key primary studies to complement the results of the meta-ethnography. Three studies were selected using purposive sampling based on their involving primary research with the displaced, having a large sample size and considering multiple domains of integration. Studies were then analyzed for their framing of the problem, conceptual foundations, methods used and claims generated. The results of the analysis are then considered in conjunction with the results of meta-ethnography to contribute to proposing a set of locally specific indicators to measure integration in the Afghan context.

5. Methodological considerations & mitigation strategies

The following outlines methodological issues with the application of SR to the research question and discusses mitigation strategies that were employed. The most significant issues included methodological heterogeneity, complex ‘information architecture’, subjectivity in the retrieval process and the non-exhaustive nature of the review. These issues and mitigation strategies are discussed below. Finally, the status of SR as a practitioner tool and its limitations and relationship with other forms of research are discussed.

Methodological heterogeneity and complex information architecture: Studies addressing the research question are not written in uniform fashion. Furthermore, an initial screening indicated that there are few studies that attempt to focus on integration holistically, with various

(27)

elements examined in different literature. For example, one study may be concerned with shelter, another with livelihoods, another with conflict with host communities etc. with relevant data contained in a range of different forms. Thus it was necessary for inclusion criteria to be broad enough to encapsulate a range of foci and sources of information, including working papers, conceptual studies and evaluation reports among others, all of which contained some contribution to the research question. This complex information architecture presents challenges in the retrieval process, as resources are frequently located outside of large, searchable databases and a comprehensive search is highly time and resource intensive.

Challenges stemming from the information architecture are mitigated through the application of the three interrelated search tracks; academic literature search, grey literature capture and snowballing. The latter in particular facilitated the identification of a number of valuable sources that were not otherwise captured.

Subjectivity in the review process: There is an element of subjectivity inherent in grey literature capture and engaging experts in snowballing. There is a risk that the reviewer be directed towards the most well-known resources or those that focus on positive outcomes at the expense of others, not resulting in a representative sample and lessening the usefulness of the review (Hagen-Zanker et al 2013 p8). Data extraction and analysis also has a degree of subjectivity. There is also a risk that analysis is conducted in an inconsistent and non-transparent way. The risk of bias and inconsistency is minimized through strict and consistent application of the pre-defined inclusion and exclusion criteria in Table 4. The detailed discussion of the methodology above aims to maximize rigor and replicability in the review process.

Non-exhaustive nature of the review process: Whilst the SR is comprehensive, it is unlikely to be exhaustive, given time and resource constraints and the disparate location of different elements of the knowledge base. The risk that significant resources are omitted is mitigated by the application of the three distinct search approaches and the deliberately broad inclusion criteria. Furthermore, whilst spread across disparate locations, the overall size of the evidence base was found to be relatively small (39 studies included in final review). Thus whilst not all-inclusive, the results of the review are comprehensive enough to provide sufficient grounds for a highly informed discussion of the research conducted into integration in contemporary Afghanistan.

(28)

Limitations of SR/position as a practitioner tool in relation to other research: SR enables a research question to be investigated, for the results of existing studies to be synthesized and consistencies and inconsistencies to be examined (Petticrew et al 2008 p130). Although a form of primary research (with the literature as opposed to research subjects), SR is not intended to replace other forms of primary research. Rather SR has the potential to be an efficient method of identifying where research is currently lacking and to provide direction to future efforts (ibid p147).

SR can provide information on the applicability of integration theory, locally specific indicators and provide an overview of the most significant findings in the evidence base. However, decisions on which specific interventions are optimal to facilitate integration will involve other considerations, such as data on the costs, cultural appropriateness and sustainability of interventions (among other factors). SRs can thus contribute to evidence based decision making, but are not decisions in themselves and should not be read as such (Petticrew et al 2008 p369). With the SR methodology elucidated and the rationale of its application to the current study established, the following section will present the study’s findings.

6. Presentation of Findings:

The following chapter will present the findings of the SR. Initially the results of each stage of the process are outlined and displayed in a flowchart. A meta-ethnographic method of synthesis is then employed including reciprocal translation synthesis and refutational synthesis. Following this a number of key empirical pieces of evidence are examined in detail, to account for questions asked and methods used and to contribute to addressing the research question

‘What does integration looks like in the Afghan context; what indicators are appropriate to measure integration’. A full presentation of review results is contained in the Data Extraction

Matrix in Annex 1.

6.1.

Systematic Review Process:

Initial searches were conducted from April – June 2019, and yielded a total of 1198 resources in both peer reviewed and non-academic sources. Following the removal of duplicates an initial screening was conducted on the title and abstract of 956 resources. 790 resources were

(29)

excluded based on the criteria outlined in Chapter 4, giving a total of 166 resources that were reviewed on full text. Following the full text screening 39 studies were ultimately included in the review. This process is represented in Figure 4.

Figure 4: Systematic Review Flowchart

6.2.

Meta-ethnography

The following presents the results of the meta-ethnography narrative synthesis of the findings of the 39 included studies. The synthesis is structured using the operational definition of integration outlined in Chapter 3. Reciprocal translation and refutational synthesis are used to examine the most salient themes regarding the lived experience of the displaced in each of the dimensions described by the indicators.

6.2.1. Reciprocal Translation Synthesis (RTS):

RTS refers to the identification of concepts in the evidence base that are consistent across multiple studies and can be synthesized into overarching themes. Viewed through the lens of the operational definition in Chapter 3, the following synthesizes the most salient themes regarding the experience of displaced Afghans in each of the dimensions of integration considered. Following each identified theme is the reference number of the corresponding studies as contained in the Data Capture Matrix in Annex 1.

(30)

Employment

The availability of employment opportunities is one of the chief ‘pull factors’4_{influencing the} location of settlement for the displaced (1, 3, 10, 16, 24, 26, 27, 33). The displaced are predominantly engaged in insecure, seasonal employment in sectors such as agriculture and construction. The majority of the displaced rely on a single breadwinner in the household, rendering households highly vulnerable to external shocks (1, 8, 9). Household ability to buy food and basic items typically items changes from day to day, with the seasonal nature of employment rendering winter a particularly difficult time (1, 8, 9, 27).

Lacking in basic literacy and numeracy is a significant impediment to the displaced securing gainful employment (8, 9, 27). Drawn to urban areas by the presence of job opportunities in the informal economy, the displaced often find themselves in competition for jobs with the urban poor (1, 27). An influx of the displaced can saturate wage labour markets, pushing wages down and creating tensions with host communities (1, 27).

Unemployment among the displaced is significantly higher than the urban poor (8, 9, 14, 17). Inability to access employment generates a cycle of other needs and negative coping strategies, such as child marriage, debt and child labour (25, 27) Government attempts to resettle the displaced often involved sites located far from existing urban centres and employment opportunities (26, 27, 28). Resettlement plans are typically accompanied by plans to create local drivers of employment however this has proved unfeasible in the current context (26, 28). Interventions to stimulate livelihood generation are often poorly linked to the needs of local labour markets (26, 28)

A key coping mechanism regarding employment for displaced households is mobility, which facilitates seasonal employment in different areas of the country or abroad (18, 29, 35, 39). Households spread members between countries and over wider geographical areas to diversify livelihoods and reduce risks (29, 33, 35).

Housing

Housing is frequently cited by the displaced as one of the most significant concerns (1, 4, 6, 8, 27, 28). Large numbers of the displaced live in tents, temporary shelters or self-constructed

(31)

shacks (1, 26, 27). Dwellings of this type are often affected by overcrowding, offer insufficient protection from the elements and are vulnerable to weather events (26, 27). A number of deaths from exposure among children in Kabul informal settlements garnered significant media attention highlighting housing issues of the displaced (8, 9, 28). Furthermore, lack of tenure security results in pervasive threats of eviction, as the displaced in urban areas often have no option but to occupy land without formal land titles (1, 8, 9, 27).

Housing issues are also closely correlated with access to basic services such as piped water, improved latrines, drainage and electricity (27, 33). In some cases, international organizations have been prevented by authorities from constructing improvements to shelter, as this is seen as acting as a pull factor that will attract more of the displaced to already overcrowded areas (26, 27). A number of interventions to provide housing and tenure security to the displaced have failed due to being in inappropriate locations, lacking access to employment and basic services (21, 27, 28).

Education

Despite significant gains in recent years in terms of levels of enrolment and literacy, access to education still remains a significant challenge for the displaced. The displaced tend to have lower levels of formal schooling and higher levels of illiteracy than the urban poor (8, 9, 26). The formal education system is difficult to access, as transport costs from locations of settlement are often prohibitive (2, 26, 38). The displaced also frequently lack official identity documents and issues of language hamper accessing education services5_{, in such cases} community based schooling and emergency schooling in locations of displacement is vital (8, 9, 13).

Availability of education is also cited as a strong pull factor influencing the location of settlement (8, 9, 13). Returnees from neighbouring countries frequently lament the inferior quality of education available in Afghanistan (17, 27). Education has significant positive implications in other areas such as levels of sanitation and hygiene and improved resilience to shocks (27)

5_{Afghanistan has two official nationwide languages and seven regional languages, see Language and Cultural}

(32)

Health

It is common for the displaced to experience issues in accessing health facilities. Facilities are often located considerable distances from settlement areas and lacking in reproductive health services (17, 27). Cultural issues around women being seen by male doctors further complicate access to adequate care (8, 9, 27). Whilst users who do access health services complain that the quality of care is very low, and that they are prescribed the same medication regardless of the ailment (17, 24)

There are grave needs regarding sanitation and hygiene, with both open defecation and the use of unimproved pit latrines in proximity to water sources common (1, 17, 27). Further health issues stem from food insecurity. It is common for displaced households to have low levels of dietary diversity and a lack of key nutrients (1, 2, 10, 27).

Social Bridges (connections between the displaced and the host community)

The quality of relations with host communities is a determining factor in the ability of the displaced to secure employment and access services (8, 9, 26, 27). However, as the population of urban areas has exceeded the capacity of government to plan for and provide basic services, the displaced are often viewed as an unwelcome presence. An influx can result in increased competition for basic services, downward pressure on local wage labour markets and conflicts over land (27, 28). Returnees who have been abroad for extended periods and who may have different accents or linguistic abilities are often viewed as foreigners and/or a security threat (8, 9, 26).

Social Bonds (connections within a community)

The presence of social and kinship networks is among the most significant pull factors influencing the location of settlement for the displaced (1, 27, 28). Networks act as social and economic safety nets and are an instrumental part of coping strategies (34, 35). Transnational social networks sustain families through remittances and family obligations and are a vital contribution to household welfare (29, 35, 37, 39). Lack of social networks or loss of them on account of moving can be a serious source of shock that undermines coping mechanisms and household resilience (17, 26).

Social Links (linkages between new arrivals and government services)

Protracted displacement is strongly correlated to weaknesses in the state-citizen relationship. Levels of trust in the Government among displaced communities is virtually non-existent (9,

(33)

20, 27). Whilst there are large gaps in awareness among government officials of key pieces of policy pertaining to the displaced and the Government’s obligations (1, 8, 9). Furthermore, in many urban areas the Government has a de-facto policy of encouraging the displaced to return to their place of origin, by restricting assistance and prohibiting permanent development interventions (24, 27, 28).

Policies that outline the Government’s obligations towards the displaced have been criticized for being prepared by international organizations with insufficient consideration of local political will and capacity (24, 26, 27). There are pervasive shortcomings in the Government’s capacity to respond to the needs of the displaced at the local level (8, 9, 21, 26). In the case of services such as justice, this seriously undermines perceptions of State legitimacy, as the Taliban and other non-state actors present themselves as an alternative choice for justice services (20, 29).

Language and Cultural Knowledge

Afghanistan has two official national languages (Dari and Pashto) and seven regional languages. A significant proportion of the population is mono-lingual and displacement events can see groups fleeing to areas where language becomes an issue. Language barriers can be a significant impediment to finding employment, as well as accessing basic services such as health and education (10, 14).

Language is a central element of cultural identity, with the Dari language associated with the Tajik ethnic group and Pashto with Pashtuns. Historical inter-cultural tensions render it possible for the displaced to be long term residents in a location yet still be considered outsiders by the host community. Religious differences and tensions between Shia and Sunni Muslims are also a factor (22, 30).

Safety & Security

Insecurity is the most cited push factor, inducing households to flee and preventing households from returning to their place of origin. Insecurity dictates the scope and direction of movement, with the relative safety of cities exerting a considerable pull factor (8, 27, 32, 33). Security at the local level however is seen more to be derived from integration into social networks and the local community rather than something provided by the State (29, 33). Security is a critical

Integration in Afghanistan : a systematic review