Practicing Social Cohesion in the Dark: Diverse Processes and Missing Indicators in Forced Migration Contexts

Despite the large number of studies on social cohesion, research carried out in non-European, refugee and local contexts remains limited. This article provides a structured, focused comparison of social cohesion practices in new refugee hosting settings based on examples from Turkey, Lebanon, and Jordan. The analysis reveals that social cohesion vision is implemented through peacebuilding, community engagement, and resilience building practices that aim at building safe, shared and empowered spaces, respectively. In contrast to national policies with more immediate, tangible and standardized outcomes, local practices involve intercommunal, interactional, and future-oriented processes, which complicates the assessment of social cohesion through an indicator framework. Based on the findings, the article suggests (a) acknowledging context bound manifestations of social cohesion practices, (b) emphasizing social interactions at the local level as a crucial component of social cohesion processes, and (c) advancing research on the causal mechanisms of social cohesion as a long-term policy goal.


INTRODUCTION
Social cohesion is increasingly presented as a remedy to all the socio-economic challenges of contemporary societies where cultural diversity stems from mixed migration flows. There has been particular emphasis on social cohesion both as a social pattern and policy goal in the last decade with ". . . a move away from multiculturalism to a social cohesion agenda" (Gozdecka et al., 2014, p. 56). Understanding the impact of social cohesion processes, however, depends on the soundness of the indicators used to assess its expected outcomes and supporting processes. While there is considerable research with social cohesion assessment tools in European countries, assessment of social cohesion through an indicator framework has been overlooked in non-European, refugee and local contexts. This study aims to fill this gap by examining the possible applications and underlying challenges of research on social cohesion indicators in forced migration contexts. To this end, this study carries out a structured, focused comparison of social cohesion practices through a detailed examination of programs implemented in Turkey, Lebanon, and Jordan. These countries encountered sudden and massive flows of refugees as a result of the civil war in Syria, leading to an emergency situation in which multiple global, international, regional, and domestic actors have engaged to provide humanitarian aid and relief programs since 2011.
Initially, all three countries were caught unprepared without previously established legal and policy frameworks to meet the unprecedented refugee inflows complicated by protracted displacements, uncertainty about the future, and increasing vulnerability of the arriving populations over time. In order to tackle the challenges associated with the Syrian refugee influx, these countries had to reconcile the initial humanitarian aid rhetoric and the emerging refugee development rhetoric.
Against this background, this study seeks answers to the following questions: which practices are at work for promoting social cohesion during intense forced migration in Turkey, Lebanon, and Jordan, and why? How can we assess the impact of these practices through a context-based indicator framework? In addressing these questions, this study has a two-fold purpose. First, it aims to reinforce the local dimension of refugee integration debates through the examination of diverse social cohesion practices implemented at the community level in three forced migration contexts. Second, it aims to propose basic assessment principles necessary to evaluate and enhance the impact of social cohesion programs designed for refugee and host communities, which are currently practiced in the dark against no specific quality thresholds.
The paper proceeds as follows: the second section provides a short overview of the existing studies on social cohesion indicators and points out the lack of research in fragile refugee settings. The third section, provides information about the process of case selection, the number and characteristics of the programs reviewed, and the criteria used to select them. The fourth section, presents a structured and focused comparison of social cohesion programs implemented in Turkey, Lebanon, and Jordan, based upon an analysis of the "program objectives, " "program design, " "involved actors, " and "implementation mechanisms" used on the ground. The fifth section discusses the possibility of assessing social cohesion programs through a context-based indicator framework. The article concludes with some final remarks on the general principles proposed in this study to build social cohesion indicators in forced migration contexts, points out the limitations of this study, and provides suggestions for future research.

MEASURING THE IMMEASURABLE? BUILDING SOCIAL COHESION INDICATORS IN FRAGILE REFUGEE SETTINGS
Since late 2013, social cohesion has turned into the dominant paradigm promoted by United Nations (UN) and European Union (EU) bodies and agencies in forced migration contexts receiving Syrian refugees continuing to flee the protracted conflict. The UN vision matured toward developing a social cohesion framework involving opportunities "for dialogue and co-operation to promote peaceful relations between refugees and host populations" as well as "to encourage peaceful intercommunal relations among refugees" (UNDP, 2014, p. 18). Within this framework, social cohesion is perceived as "absence of latent conflict whether caused by racial, economic or political reasons, among others; and the presence of strong social bonds, as noted by the existence of trust, reciprocity, associations crosscutting social divisions and the presence of institutions of conflict management" (UNHCR-UNDP, 2015, p. 13). The EU has also played a leading role in the region mainly through the Regional Trust Fund (2014), which, "addresses longer term economic, educational and social needs of Syrian refugees in neighboring countries such as Jordan, Lebanon, Turkey, and Iraq, as well as helping overstretched host communities and their administrations" (European Commission, 2017, p. 1). Even though there are no strict demarcation lines, while the UN pays slightly more attention to the rebuilding the social fabric by mitigating tensions and expanding interactions between refugee and host communities, the EU prioritizes the institutional framework needed to promote refugee protection in host countries. Both the UN and the EU perspectives converge on seeking durable solutions for refugees, which require longer term social and policy processes translatable into a common vision of social cohesion with a particular emphasis on community level practices. Yet, despite presenting social cohesion through a set of values and principles to be followed, the UN and the EU provide only minimal policy directions and evaluation criteria for assessing the impact of social cohesion programs. So, even though social cohesion is at the very heart of the current migrant and refugee debates, there seems to be a gap between principles and norms, on the one hand, and performance measures and quality indicators, on the other hand.
Existing social cohesion measures and indicators have been developed almost exclusively in the European context or other traditional refugee and migrant receiving countries. The systematic study of social cohesion indicators dates back to early 2000s, when social cohesion was incorporated in the quality of life assessments as one of the main pillars of the European System of Social indicators (Noll, 2002). Over the years, consecutive studies have refined the conceptualization and operationalization of social cohesion with reference to its constitutive elements and causal mechanisms in order to facilitate further empirical work, in particular within and across European countries (Duhaime et al., 2004;Chan et al., 2006;Vergolini, 2011;Dimeglio et al., 2013;Klein, 2013). More recent studies have advanced the state of the art through sophisticated measurements of social cohesion derived from multilevel/multidimensional models (Andrews et al., 2014;Delhey and Dragolov, 2016;Koopmans and Schaeffer, 2016;Bottoni, 2018;Goubin, 2018;Valli et al., 2019). There is ongoing debate in the literature about standardizing measurements in order to compare social cohesion outcomes across countries or as part of particularistic research agendas that reflect different geographies (regional vs. universal/global, country assessments, etc.), domains (trust, participation, diversity, identification, equality, etc.), or levels of analysis (individual vs. collective/social, community vs. country, etc.) (see Schiefer and van der Noll, 2017).
Research on how to evaluate the impact of social cohesion processes in terms of access, quality, and overall effectiveness in new refugee contexts remains limited. While many programs have been promoted as part of social cohesion as a policy objective, there have been very few attempts to build social cohesion indicators in recent refugee receiving countries. The existing indices of social cohesion used in settlement countries such as Australia, New Zealand, Canada, or the UK are hardly applicable in new forced migration contexts, where refugees have no permanent legal status that can be linked to policy development in the political, social, or economic spheres. Due to the form of temporary protection granted to forcibly displaced population in Turkey, Lebanon, and Jordan, local practices of social cohesion have been prioritized over national policies oriented toward more conventional forms of refugee integration. This situation, by default, makes local actors and practices highly significant in the analysis of social cohesion processes and related outcomes. Even though national actors play a decisive role in terms of political decisions concerning refugee related issues (countries under investigation have centralized international protection regimes), local actors including local government representatives as well as non-state actors play a critical role in de facto facilitating of social cohesion processes on the ground. While the legal and institutional framework for social cohesion processes are shaped at the national level, social cohesion practices constitute local affairs. Based on this premise, this study elaborates on the "local turn" in migration studies which emphasizes a need to shift the focus of the analysis toward the interaction among local actors and contexts (Caponia and Jones-Correa, 2017;Zapata-Barrero et al., 2017;Caponio, 2019;Geuijen et al., 2020).
The dilemma of the new refugee hosting countries is implementing durable solutions without adopting an integrationist policy discourse. The emerging social cohesion vision lacks a standardized indicator framework that explicitly lays out refugee integration policies and expected outcomes applicable across country cases. The challenge, here, is to assess the impact of social cohesion programs implemented under conditions of uncertainty in the political sphere, unemployment in the economic sphere, and antagonism toward the forcibly displaced in the social sphere. To complicate matters further, social cohesion programs are implemented in areas with inherent socio-economic challenges and relatively unprepared local government infrastructure, often referred to as "poverty pockets." Within these fragile urban enclaves, refugees are often perceived as burden rather than opportunity, which adds to their vulnerability. Moreover, with not many tools at their disposal, local governments often resort to improvised strategies or pragmatic responses to deal with the challenges at hand, which can be interpreted as a situation promoting diversity and innovation, but at the same time, as a situation leading to complexity and policy divergence between national and local government. Under such circumstances, it is far more appropriate to examine the practices that shape the micro-levels of society, that is, programs aimed at fostering social cohesion at the community level, rather than investigate policy discourses at the national level.
Against this background, building social cohesion indicators in the non-European, refugee and local contexts resembles an equation with many knowns, which leads to confusion in terms of benchmarking the impact of social cohesion processes, and also in terms of designing related programs and policies. So far, the majority of the existing studies evaluating social cohesion programs in Turkey, Lebanon, and Jordan are based on qualitative research with data retrieved from in-depth or semistructured interviews with key informants from households, communities, or municipalities and a few survey results. Social cohesion indicators are measured at the individual rather than the aggregate level, making them indirect or proxy indicators of social cohesion. In addition, most of the indicators are projectbased assessments built to monitor or measure the impact of individual social cohesion programs rather than overall social cohesion processes and outcomes. For instance, the World Food Program conducted online surveys of refugee and host communities to monitor social cohesion (WFP, 2018) as part of the monitoring framework of the ESSN program in Turkey; the Gender and Adolescence: global Evidence (GAGE) initiative has evaluated the effects of the projects Better Together and FURSA as part of its study of social cohesion among conflict-affected refugee and host community adolescents in Lebanon (Gercama et al., 2018); or the monitoring study of the Jordan Emergency Services and Social Resilience Project (JESSRP) conducted by REACH in Jordan (REACH, 2016). Overall, there is a lack of dialogue between scholarly research and practice aiming to construct social cohesion indicators in new refugee hosting countries. Considering this gap in the literature, the rest of the paper analyzes and classifies initiatives/programs implemented in Turkey, Lebanon, and Jordan which make reference to social cohesion between 2011 and 2020.

METHODOLOGY AND CASE SELECTION
This section, provides information about country case selection, the methodology used in the analysis, and the characteristics of the programs reviewed as well as criteria for their selection. Turkey, Lebanon, and Jordan have been selected as countries hosting the largest number of displaced people since the outbreak of conflict in Syria in 2011. Except for being in the same region, these three countries, fall "under the category of refugee rentier states, given the fact that they received external economic aid that was dependent on their status as hosting forcibly displaced populations" (Tsourapas, 2019, p. 476). The Syrian population in Turkey, Lebanon and Jordan has been granted temporary protection, which defines particularly the residential status of the displaced people in these three countries. According to the UNHCR definition, "temporary protection" refers to arrangements developed to offer protection of a temporary nature, either until the situation in the country of origin improves and allows for a safe and dignified return, or until individual refugee or complementary protection status determination can be carried out" (UNHCR, 2018b, p. 30). In addition, all the three countries have not granted refugee status to Syrians mainly due to national security concerns as witnessed in the case of Lebanon and Jordan or due to limitations put by the international law set in the United Nations Geneva Convention (1951) and its 1967 Protocol as witnessed in the case of Turkey. The Law on Foreigners and International Protection [Article 61, (LFIP, 2013)] in Turkey limits the refugee status to Europeans only, and, as a consequence Syrian population remains under temporary protection until the safe return to their country of origin is possible.
Considering the temporary arrangement of the displaced population, Turkey, Lebanon, and Jordan have provided solutions of a local nature especially through community based social cohesion programs supported by international funds. The analysis presented here is based upon a structured, focused comparison of social cohesion programs, which aims to identify the differences in the implementation of social cohesion programs across and within country cases by taking into account various actor constellations. Structured, focused comparison is a methodological perspective that facilitates policy analysis through in-depth data derived from small-n country comparisons (Kachuyevski, 2014); gives a more accurate picture of the relations among sub-state actors alongside state actors (Kachuyevski, 2018); and, compliments large-n analysis in comparative politics (Stegmaier and Marcinkiewicz, 2019). The structured, focused comparison carried out in this study has a three-fold purpose: (a) to shift the focus of research from principles to practices through detailed analysis of social cohesion programs designed for refugee-host communities at the local level (b) to organize the plethora of social cohesion programs according to distinct classification criteria used to identify different types of social cohesion practices across and within country cases (c) to link ongoing practices to possible evaluation criteria that can be used to assess the impact of ongoing social cohesion processes in refugee hosting settings.
The first step of structured, focused comparison is to develop criteria in order to standardize data collection by classifying selected programs under inductively constructed categories, and, increase the comparability of the selected cases in a more systematic fashion. Four classification criteria have been identified for the purpose of this study: "program objectives, " "program design, " "involved actors, " and "implementation mechanisms." These criteria have been constructed by the authors after reviewing a considerable number of social cohesion oriented programs implemented in Turkey, Lebanon, and Jordan since 2011. Particularly important for the purpose of this study are the programs introduced within the framework of the Jordan Compact, the Lebanon Compact, and the EU-Turkey Statement, all signed in 2016 in respective countries. The selected social cohesion programs included in the analysis provide an illustrative rather than an exhaustive list of all programs supporting social cohesion processes in the countries under investigation. Based on these classification criteria three different types of social cohesion programs have been identified across three country cases, namely (a) peacebuilding (b) community engagement, and (c) resilience building practices. The programs included in the analysis best identify with each social cohesion practice listed above. Some of the programs combine characteristics of two or more of the above categories, representing an integrated approach to social cohesion. On a final note, the analysis in this paper intentionally excludes the discussion of national policy strategies and political responses and focuses on the characteristics of the programs of social cohesion as implemented at the local level, in order to develop general principles that can be used to evaluate the impact of social cohesion processes.

SOCIAL COHESION AT WORK: FROM NATIONAL POLICY STRATEGIES TO COMMUNITY BASED PRACTICES
The analysis of the social cohesion programs in Turkey, Lebanon, and Jordan suggests that they can be classified into three types of practices: (a) peacebuilding (b) community engagement, and (c) resilience building (summarized in Table 1).

Peacebuilding: Opening Up Safe Spaces
Peacebuilding processes are a crucial aspect of broader social cohesion building efforts in conflict-affected countries. Under such circumstances, social cohesion can be perceived as antidote to violence, conflict, and all various forms of social tension arising in host countries, and as a process of stabilization, reconciliation, and peacebuilding through rapid response mechanisms in emergency situations. Scholars studied the conditions under which often externally sponsored peacebuilding initiatives effect social cohesion processes positively and without reinforcing existing social divisions (Cox et al., 2014). Lack of actual or potential cases of extreme violence have rendered peacebuilding activities less necessary in Turkey compared with conflict prone zones in Jordan and Lebanon, where already existing tensions have been exacerbated since 2011. In Lebanon, the arrival of Syrian refugees has been identified as having a "destabilizing effect" (Jütersonke and Kobayashi, 2015, p. 15) in a country where "social cohesion . . . is. . . about the achievement of stabilization" (Guay, 2015, p. 24). Similarly, Jordan's geopolitical location "makes it extremely fragile to spillovers of regional conflicts and violent extremism on the rise in the region" (UNDP, 2017). For instance, "Strengthening Civil Peace in Lebanon" (2011Lebanon" ( -2013 implemented by UNDP aimed at "developing local level peace strategies in three conflict prone areas in Lebanon" (UNDP, 2013a). In Jordan, the first 2 years of the "Mitigating the Impact of the Syrian Refugee Crisis on Host Community Project" (2013-2015) focused on conflict assessment and peacebuilding through the implementation of a Preventing Violent Extremism (PVE) strategy (UNDP, 2013b). Launched by the UNDP in the early years of the Syrian refugee crisis, these projects aimed at preventing crime, conflict, and different forms of extremism in the society, as well as at enhancing stabilization and community security with an expected impact on social cohesion.
Peacebuilding activities are often tailored toward children, adolescents, and young adults through conflict transformation programs implemented largely through non-governmental organization (NGO) platforms specialized in conflict resolution. Previous research has focused on the role of "psychosocial support-based peace education work with young displaced Syrian refugees" (Simpson, 2018, p. 38) within broader social cohesion projects. In Jordan, UNICEF has collaborated with the Generation for Peace to implement the "Social Cohesion Program" (2014-2016) which aims at "reducing tension and risk of violence" (UNICEF, 2014) through a conflict transformation curriculum applied to children and youth in vulnerable, disadvantaged and conflict-sensitive urban areas. The underlying assumption here is that behavior change among the young population, especially youth leaders, will have positive spillover effects to the overall community contributing to sustainable peacebuilding and social cohesion in the long run. Another project funded by the EU and implemented by Search for Common Ground in collaboration with two other local organizations in Lebanon, is named "Better Together: A Youth-Led Approach to Peaceful Coexistence" (2014-2016). The specific objectives of the project are to "develop nonadversarial relationships" between Syrian and Lebanese youth and "implement peacebuilding activities in their communities" (Search for Common Ground, 2014, p. 6). The overall objective of peacebuilding activities is to reduce tension and maintain communal peace in conflict sensitive areas. Peacebuilding activities make use of rapid response mechanisms that alleviate social tensions in emergency situations, and, eventually, prepare the ground for carrying out more advanced forms of societal interactions through community engagement and resilience building activities. Peacebuilding activities aim at opening up safe spaces where refugees can establish their first contact as well as feel accepted in the host society after a traumatic war and migration experience. To this end, many of the projects make use of a conflict transformation curriculum that engages children, adolescents, and young adults. From a governance perspective, peacebuilding activities are carried out either through stand-alone projects or as the initial stage of broader social cohesion projects funded and implemented by intergovernmental organizations, and international and local NGOs specialized in conflict transformation and peacebuilding without much state involvement. The operation rather than the policy component of these projects is paramount, with short term projects often lacking advanced monitoring mechanisms. The most recent peacebuilding activities reflect the approach that "rejects the traditional binary understanding of peace and war" (Rasheed and Munoz, 2016, p. 174) and instead perceives peacebuilding as a longer term and relational process that aims at eliminating all structural factors that might lead to violence in a society.

Community Engagement: Constructing Shared Spaces
Community development activities are usually discussed under community development projects in migration and refugee studies. They take different forms ranging from sport activities (Nathan et al., 2010;Rosso and McGrath, 2016;Mohammadi, 2019;Robinson et al., 2019) to art activities (Lewis et al., 2018;Nashwan et al., 2019;Lee et al., 2020). Community engagement activities are one important aspect of social cohesion programs, which have been implemented in all countries under investigation, namely, Turkey, Lebanon, and Jordan. Community engagement activities often involve municipalities and community based organizations (CBOs) as strategic local partners closer to different groups at the local level. These interactions aim at promoting mutual understanding, sociocultural interaction, and positive communication between refugee and host communities.
In Turkey, a wide range of community engagement platforms are designed to strengthen social cohesion processes particularly through active involvement of municipalities, local NGOs (referred to as CBOs), and multi-purpose community centers. Support toward municipalities focuses on "expanding sociocultural interaction between refugee and host communities and increasing the presence of refugees in local representation structures" (UNHCR, 2018a, p. 3). The burgeoning NGO sector in the area of relief and support services toward refugees is clear evidence of the growing significance of community engagement activities in Turkey. NGOs in Turkey have evolved over time from organizations offering humanitarian assistance at the local level to serving as mediators who contribute to social cohesion processes through different community engagement activities geared toward creating social spaces at the local level (Sunata and Tosun, 2019). Finally, community centers have been very active in refugee protection services and psycho-social support services, as well as awareness raising and socio-cultural activities. The Turkish Red Crescent Community Centers (TRCCCs) combine humanitarian relief and social protection services with health, education, and professional training services, as well as platforms of interaction between different population groups in the society based on the community engagement and accountability approach (CEA). Basic services extended in these community centers represent an attempt to overcome individuals' trauma and psychological disorders and to extend support to families through an emphasis on restoring and strengthening family ties. Other services contribute to social cohesion processes through community engagement activities that foster interactions and communications among refugee and host groups (TRC., 2018).
In the case of Lebanon and Jordan, the majority of community engagement activities are carried out through grants extended to CBOs. One interesting example is the awareness-raising sessions carried out under the social cohesion framework for mixed groups of Syrian and Lebanese participants, held by AMEL Association International and Premiere Urgence-Aide Medicale Internationale (PU-AMI) in 2015 in Lebanon. These sessions made use of dialogue on health and social issues as a fertile ground to build trust between refugee and host communities. Through these mixed awareness sessions, refugee and host communities not only learned together but could also freely discuss common health-related issues such as reproductive health, family planning, or child-rearing by sharing the "cultural, religious and social perspectives affecting their choices" (International Alert, 2015). According to one report, "the awareness sessions were . . . successful in decreasing misunderstandings between the communities and discovering that all individuals are facing similar challenges, regardless of their nationality" by realizing that "they were sharing the same culture and had common political and social history" (International Alert, 2015). One interesting example is the Badr Nazzal Initiative in Aman, Jordan, which has been defined as "a community led initiative using public spaces to encourage social interaction between host communities and refugees" (USE, 2016). This initiative, ongoing since 2016, aims to bring different communities together by creating spaces of encounter such as local charity bazaars, local parks or social networking programs. This project has been perceived as an opportunity to "promote social cohesion by fostering the emergence of a district-based identity" (Feder, 2018). Previous research has shown that similar interactions in urban areas, such as public park encounters, can stimulate social cohesion and interactions between various ethnic groups (Peters et al., 2010).
In the case of community engagement activities, the real challenge is not the lack but rather the abundance of activities and the multitude of actors involved, often lacking coordination (both horizontally and vertically) and collaboration mechanisms. Against this background, carrying out community engagement activities through community centers that combine the efforts of various actors not only contributes to a more efficient use of resources by avoiding activity overlap. Some community engagement activities take the form of workshops, awareness sessions, and interactions in public urban areas, rather than the regular societal interactions carried out in community centers in the Turkish case. The main advantage of community centers is that they enable more systematic work and bring to fruition the efforts of many actors for more efficient and effective use of resources. Even though designed to achieve the same goal, community engagement activities through community centers are more geared toward promoting or strengthening social cohesion through repeated and systematic societal interactions, while, ad-hoc community engagement activities are more geared toward stimulating or encouraging social cohesion through cursory societal interactions. Hence, from a governance perspective, community engagement activities taking place in community centers are carried out in more systematic fashion contributing to both quantity and quality of societal interactions at the local level.

Resilience Building: Developing Empowered Spaces
Even though it remains limited in terms of institutional implications, resilience is considered to be the new paradigm in migration and refugee governance (Paul and Roos, 2019). The concept of "resilience" is used within the context of a development-displacement nexus to highlight the need to build refugee self-reliance, while also ensuring that host communities not only cope and recover from crisis but also develop the longer-term development prospects needed to move toward lasting peace and prosperity (UNDP, 2015(UNDP, , 2016European Commission, 2016). Research has shown, however, that "the different forms and implications of resilience-building are highly dependent on the local context and in particular on the interests of the actors involved" (Anholt and Sinatti, 2019, p. 10).
The majority of resilience building efforts have been directed toward improving the well-being of refugees through their empowerment in the areas of health, employment, and education. In the area of health, one example is the project "Capacity Building for the Syrian Health Professionals in the Turkish Health System and Supporting Culture-Sensitive Health Services in a Way to Serve their own Populations" implemented by WHO, the Association for Solidarity with Asylum Seekers and Migrants (SGDD-ASAM), and the Provincial Public Health Directorates of the Ministry of Health in Turkey (SGDD-ASAM, 2017). This project empowers Syrian doctors and nurses by integrating them into the Turkish health system and it also empowers Syrian refugees by building a space to overcome their language and cultural barriers as patients (WHO, 2017a,b). In the area of employment, one example is the project "Improving Labor Market Integration of Syrian Refugees and Host Communities in Turkey, " which aims "to enhance the livelihoods and social cohesion of Syrian refugees and host communities in Turkey by promoting labor market integration and inclusive economic growth underpinned by decent work" (ILO, 2018a). In the field of education, one example from Lebanon is the project "Reaching All Children with Education" (RACE II) implemented by the Ministry of Education and Higher Education (MEHE) in collaboration with UNICEF, other UN agencies, and local NGOs. This project aims at improving the situation of refugees in the long run through "improved access to education opportunities, improved quality of education services, (and an) improved education system" (UNICEF, 2019).
Resilience building programs are also oriented toward skills development with the purpose of contributing to labor market inclusion of refugees in the host society, targeting women, adolescents, and youth in particular. Within this framework, the Providing Livelihood Support Project implemented by the International Labor Organization (ILO), SGDD-ASAM and Gaziantep Municipality through the SADA Women's Development and Solidarity Center serves as an example that "aims to empower women and girls from both the refugee and the host community, combines the creation of livelihood resources, protection and social cohesion activities" (ILO, 2018b). A similar example in Lebanon is the program aiming to establish seven Agricultural Technical and Vocational Education and Training (TVET) schools implemented by the Ministry of Agriculture in collaboration with UNICEF and the Food and Agriculture Organization (FAO). The objective of this program is to ensure higher levels of participation from refugee adolescents into the formal education system as well as to overcome "the mismatch between skills acquired through education and the skills required at the labor market" (UNICEF, 2019) by establishing skill development environments for adolescents and the youth.
Although still at an incipient stage, other resilience building activities have been designed to strengthen social protection systems for refugees in the host society. One example is the Emergency Social Safety Net (ESSN) developed as part of the program of the Facility for Refugees in Turkey (FRIT), funded by the European Civil Protection and Humanitarian Aid Operations (DG-ECHO), and managed by the Ministry of Family, Labor and Social Services (MoFLSS), the WFP, and Turkish Red Crescent. ESSN provides "multi-purposeful cash assistance" (FOPT/MoFLSS, 2018), which can be considered as a useful tool that empowers refugees by giving them a choice on how to spend money based on their own needs. Besides, the purpose of the ESSN is "to increase the social cohesion of the SuTP (Syrians under Temporary Protection) by supporting their adaptation to the labor market" (FOPT/MoFLSS, 2018, p. 13). Despite its many benefits, there are doubts about the effectiveness of this type of financial assistance as it can as well-lead to "unregistered, low paid and precarious, irregular work in the labor market for the households" especially when aid is perceived as being "non-wage income type" (FOPT/MoFLSS, 2018). In this context, financial assistance can have unintended consequences that run contrary the purpose of social cohesion processes. Furthermore, the program falls short of developing a system of performance/outcome measurement through indicators that can be used to trace graduation of the Syrians from the social assistance stage to their adaptation to the labor market.
Finally, resilience building activities have also been directed toward strengthening local government capacities in terms of strategic planning, decision-making and implementation of programmes designed for refugees and vulnerable population groups. Local government resilience programmes are particularly important in refugee affected urban areas with protracted crisis and underdeveloped infrastructure. The Jordan Emergency Services and Social Resilience Project (JESSRP) implemented by the government of Jordan with the support of the World Bank best illustrates this case. The objectives of this project are "to provide grants and material to the municipalities, and to improve their management through better communication and accountability procedures, in order to enforce community resilience and social cohesion" (Ababsa, 2017). Resilience building processes oriented toward municipalities take the form of grant and equipment transfer designed to strengthen their institutional infrastructure, as well as, consultation services designed to reinforce their social infrastructure.
The content of resilience building programs consists of a wide range of activities which can be classified under three broad categories: (a) improving the well-being of refugees by strengthening already-existing public service capacities (including municipality services) in the areas of education, health and employment in particular; (b) contributing to refugees' wellbeing by promoting social assistance schemes through investments in social safety nets; (c) providing language, vocational, and skills training expected to contribute to labor market inclusion of refugees (particularly women, adolescents, and youth) in the host society in the medium and long terms; and (d) developing local government capacities in order to design durable solutions for refugees and host communities at the local level. Overall, resilience building activities can be considered as the last chain of social cohesion programs, which are based upon the principle of empowerment through emphasis on skills development and self-reliance of refugee and host communities. The advantage of resilience building activities is that they tend to be both policy informed, that is, based upon existing policy frameworks, and policy relevant, that is, with consequences for policies toward refugees. Being linked to policy processes, resilience building programs can be more easily monitored and evaluated through indicator-based assessments. However, unlike integration policies developed in traditional refugee receiving countries, resilience building processes have an inherent long-term and future-oriented character in terms of intended objectives (enhanced livelihoods, self-reliance, and empowerment), the chain of activities involved in the social and economic spheres, and the degree of institutional and policy transformation required to contribute to sustainable development of refugee-host communities. In addition, the majority of policy transformation is expected to take place through strengthening municipalities' institutional and social infrastructure, which makes resilience building an essentially local endeavor.
Recently, there has been a tendency to follow integrated approaches that combine specifically designed activities into single comprehensive programs that explicitly prioritize social cohesion. Through such approaches resilience building activities have become center stage of social cohesion processes at the community level. The underlying assumption here is that integrated approaches can contribute toward strengthening social cohesion processes through more comprehensive packages that merge resilience building activities with either community engagement or peacebuilding activities.
In the case of Turkey, the TRC community centers do represent an amalgam of community engagement and resilience building activities targeting all vulnerable population groups. Besides TRC Community Centers operating across Turkey with the support of the European Union, there are other community centers such as Multi-Purpose Community Centers operating mainly in the GAP region and funded by UNDP (2019) and community centers operating under Community Centers Project, commissioned by GIZ (2020). All these centers are dedicated to the promotion of both (a) shared/exchange spaces targeting peaceful co-existence and inter-cultural dialogue and (b) empowerment spaces targeting enhanced livelihood and resilience of vulnerable population. In other words, the integrated approach applied in Turkey combines community engagement activities with resilience building activities.
Other projects combine resilience building with stabilization goals. For instance, "Supporting Livelihoods and Social Stability for Syrian Refugees and Host Populations, " FURSA, which operates in Lebanon, Iraq, and Turkey, is "designed to engage youth from refugee and host communities through an integrated approach aiming to increase social cohesion, psychosocial wellbeing and livelihood to ultimately strengthen youth social and economic inclusion in their communities" (Dari and Ortmans, 2017, p. 53). In the case of Lebanon, FORSA Programme Fostering Resilience and Strengthening Abilities (2016-2019), is described as integrated program designed to promote resilience and, at the same time, "reduce inter-communal tensions between Lebanese, Syrians and Palestinians by alleviating competition over livelihoods" (Aid Stream, 2019). Similarly, in Jordan social cohesion programmes have been implemented through an integrated approach combining stabilization with enhanced livelihoods. Cash for work programmes are typical examples that combine peace building efforts safe spaces that mitigate gender-based and inter-communal violence and resilience building efforts through empowerment spaces. Jordan provides ample examples of cash for work programmes such as Women and Girls Oasis Programme, Protecting Reservoirs in Jordan, Solid Waste Management, and several urban regeneration programmes.

CONTEXT BASED INDICATOR FRAMEWORK FOR SOCIAL COHESION: IMPLICATIONS FOR RESEARCHERS AND PRACTITIONERS
Based on the above analysis, social cohesion practices in new refugee receiving countries take the form of processes involving peacebuilding, community engagement, and resilience building activities. Unlike policy measures, which often yield to more immediate or tangible outcomes, these activities have an intercommunal, interactional, and future-oriented character, making them intrinsically more sophisticated and complicating the study of social cohesion through an indicator framework. Social cohesion processes in new forced migration contexts are minimalistic in terms of expected policy and societal transformations in the short term. Minimalist policy expectations are mainly due to the temporary form of protection granted to the Syrian population, which renders discussions of refugee integration (as largely defined) futile. Considering the political and legal limitations, social cohesion programs in the countries under investigation are implemented primarily through community based activities rather than national wide policies. Such activities are not expected to bring immediate tangible results linked to issues such as citizenship, employability, or political participation. Social cohesion activities as discussed in the previous section do not aim to bring cultural transformation or contribute to identity building (i.e., of a sense of national or group identity) amongst refugees either. Social interactions between refugee and host communities, which take the form of dialogue based on inter-cultural communication, awareness raising, and information sharing with minimum acculturation strategies, are of paramount importance, to the degree that they are often equated to social cohesion itself. Social cohesion programs as described above have a future-oriented characterexpected to contribute to longer term social and policy processes both in a situation of prolonged stay in the refugee hosting country as well as return to the country of origin.
Against this background, the question arises as to how to develop reliable indicators to measure the impact of such ongoing and intrinsically complex processes leading to a socially cohesive society. To this end, this section addresses the following questions: How can social cohesion indicators be defined and used in new forced migration contexts? What are some of the main measurement principles that can be used to assess social cohesion processes? Addressing these questions would help researchers and practitioners to better trace the processes and evaluate the outcomes of social cohesion processes at the community level. Building upon the findings of the previous section, an indicator framework of social cohesion should take into account the following criteria.

Acknowledging Context-Bound Manifestations
For the purpose of this study, context will be defined in terms of (a) the background setting in which social cohesion programs operate and (b) the content of the social cohesion program under evaluation. The first category includes factors such as the size/composition of the forcibly displaced population; the social and cultural characteristics of the displaced people; the economic, political, and social situation in the refugee hosting country; the urban or rural settings in which the program operates, as well as other subnational factors. Contextual factors are critical for developing social cohesion indicators, contributing to nuanced analyses of social cohesion processes that take into account national and subnational contexts. Taking into account contextual factors is particularly necessary in new refugee hosting countries, which have responded to challenges of refugee population often through semi-formal or informal measures introduced by local actors. Stated differently, social cohesion indicators in new refugee hosting countries should go beyond the formal responses supported by legal and policy frameworks and emphasize context-bound factors at the local level.
In addition, a context-bound evaluation design necessitates a good knowledge of the content of the programs under investigation. Building upon the previous section, social cohesion programs are carried out through three process types containing (a) peacebuilding activities, (b) community engagement activities, and (c) resilience building activities. Hence, indicators of social cohesion should distinguish between these three processes, take into account their differences in terms of program design, and, finally, build related measures that best capture the intended objectives. Peacebuilding activities can be evaluated through measures that gauge the degree of possible conflict and violence including both intercommunal and genderbased violence as well as measures that gauge the prospects of stabilization at the community level and in the labor market. Community engagement activities can be evaluated through an assessment of different types of interactions, taking into account the frequency and quality of interactions. Resilience indicators can be developed through the assessment of different resilience processes as discussed in the previous section. Finally, integrated social cohesion programs can be assessed through composite indices that offer a summative evaluation of different aspects of social cohesion processes at the societal and policy levels. All these indices need to be developed as process-based evaluations that involve various actor constellations operating at multiple levels of governance and across policy sectors rather than solely outcome-based evaluations like in the case of refugee integration policies with more tangible and immediate results as witnessed in traditional migrant receiving countries.

Emphasizing the Social Over the Policy Component
Social cohesion is above all a social phenomenon assessed most meaningfully through social indicators which are built upon subjective experiences (attitudes, experiences, etc.) rather than objective measures (policy measures). The social aspect of social cohesion processes can be assessed through the evaluation of various socio-psychological, socio-cultural, sociospatial, and socio-economic processes, which are carried out with the purpose of facilitating mutual exchange and interactions between different ethnic, linguistic, and cultural groups at the local level. Particularly useful in this context is the intercultural cities index (ICC) constructed by the Council of Europe, which emphasizes interactions in urban areas by focusing on concepts such intercultural mixing and interaction in the education system, residential neighborhoods, public services, business and labor markets, cultural and civil life, and public spaces [see Wood (2009)]. Interactions are expected to bring about a qualitative change in awareness or attitudes toward other people. The ICC identifies the pertinent interactions for social cohesion, however, does not clarify how the quality and exact impact of interactions may be assessed comprehensively across different forced migration settings.
National level policy strategies aiming to facilitate social cohesion processes usually set general principles without providing a repertoire of practices. Social cohesion practices and initiatives involving various interactions at the local level, however, are more transferrable across and within country contexts than national level policy strategies, which tend to be holistic and, as a consequence, more rigid and less comparable. Social interactions at the local level are also transnational both in terms of content (socio-interactional approach among city dwellers beyond their legal status) and methodology (comparable community-based social cohesion practices across various contexts). International organizations have embraced the philosophy of "localism" through "community building efforts" and "community engagement" when it comes to facilitating and strengthening immigrant integration and social cohesion processes. According to a recent OECD report, it is crucial for social cohesion processes to empower refugees through active local communities "by revitalizing demand for local business, bringing local and migrant families together around public schools and health centers and diversifying the cultural activities for all residents" (OECD., 2018, p. 18). The move toward local specificities fosters social interactions at the community level, supports bottom-up innovation with opportunities for all vulnerable groups, and, allows for variations across time and space as opposed to assessment against predefined policy thresholds or patterns of success. Even though the social aspect is becoming increasingly paramount, social cohesion processes are more and more becoming policy informed or policy relevant particularly through resilience-focused schema. Yet, even in this case, resilience building activities are more a product of local level platforms that require active municipal engagement rather than national level policy strategies.

Evaluating the Causal Mechanisms of Social Cohesion
As discussed in the previous section, social cohesion practices can be classified under peacebuilding, community engagement, and resilience building activities, which can be considered as constituting elements of social cohesion as a policy agenda. Even though constituting elements of social cohesion have stimulated mutual interest among researchers and practitioners, causal mechanisms of social cohesion in refugee receiving settings have not been studied systematically. When such mechanisms are not clearly identified, the boundaries of what explains what becomes fuzzy, making it hard to build causal inferences about the consequences of social cohesion processes and policies. Hence, researchers need to clarify the issues around framing social cohesion and identifying its constituting elements and related causal mechanisms in order to generate an accurate indicator framework. The research and assessment of causal mechanisms becomes particularly important in planning future policy strategies, especially in the context of resilience building or integrated social cohesion packages, which involves a chain of social and economic activities contributing to enhanced livelihoods in the long run. Following this logic, researchers should consider identifying the specific causal mechanisms that support resilience building activities and, as a consequence, strengthen social cohesion processes.

DISCUSSION AND CONCLUSIONS
The massive displacement caused by the prolonged humanitarian crisis in Syria has initiated a major social transformation in surrounding refugee receiving countries since 2013. Since then, social cohesion has been used in many policy statements as a desirable end state and has entered the social policy agenda of Turkey, Lebanon, and Jordan. Despite the willingness to recognize social cohesion as a policy target and social pattern, social cohesion processes are still conducted in the dark, without any quality thresholds that can be used to assess the impact of the implemented programs. As exploratory research, this study seeks to start a new discussion prompted by the need to construct social cohesion indicators in forced migration contexts, specifically in new refugee hosting countries. Existing theoretical and empirical studies on social cohesion indicatorsprimarily undertaken in the European context-do not address challenges such as legal temporariness of refugees, a lack of fully established policy frameworks, persistent regional conflicts and violence, fragility in the economic sphere, and uncertainty in the political sphere as witnessed countries experiencing intense forced migration. Likewise, new assessments that can be used to evaluate social cohesion programs implemented in new refugee hosting countries are largely lacking. Hence, this study introduces a context-based indicator framework, which is expected to serve as a starting point toward assessing the impact of ongoing social cohesion programs in refugee hosting settings.
The comparative analysis of social cohesion practices across Turkey, Lebanon, and Jordan reveals diverse processes contingent upon the responses of local actors rather than a predefined, uniform model of social cohesion. Nevertheless, it is still possible to identify three types of social cohesion practices designed to enhance peacebuilding, community engagement, and resilience building activities, respectively. The most recent trend has been to combine these activities into integrated packages aimed at offering a more comprehensive framework to address the challenges facing refugees and other vulnerable population groups residing in so-called "poverty pockets" within refugee hosting countries. The analysis of social cohesion practices reveals the growing significance of municipalities, community based organizations, and community centers as pivotal actors in strengthening social interactions and supporting community engagement at the local level. The comparative advantage of local actors compared to international donors and central governments is their proximity to refugee-host communities and their ability to sustain repeated and systematic interactions between service providers and receivers. Such interactions are at the heart of social cohesion processes in new refugee hosting countries, to the extent that they can be equated to the concept of social cohesion itself.
Overall, the analysis reveals that social cohesion practices are supported by intercommunal, interactional, and futureoriented processes taking place at the local level, which makes the development of measurable indicators highly challenging. The shift from policies implemented at the national level toward processes taking place at the local level compels researchers to develop more sophisticated indicators that take into account the complexity of relations among multiple actors (relational rather than structural indicators), the possibility of change over time and space (situational rather than standardized indicators), and a spectrum of experiences within and across communities (subjective rather than objective indicators). Unlike the traditional refugee contexts, the paradox of the Syrian refugee context originates from the legal temporariness of the displaced people despite their long term residence in the host country, which further complicates the assessment of social cohesion through an indicator framework. Against this background, this research proposes three criteria that can be used to develop context-based indicators of social cohesion: (a) acknowledging context bound manifestations of social cohesion practices by focusing on reflexes developed at the local level, (b) emphasizing the social over the policy component of social cohesion processes by paying more attention to the quality of social interactions, and, finally, (c) overcoming the endogeneity problem resulting from pronouncing the constituting elements over the causal mechanisms of social cohesion as a desirable end state in refugee hosting countries. These criteria would be relevant to both researchers and practitioners interested in assessing social cohesion processes and outcomes in forced migration contexts.
On a final note, this study includes only a limited number of cases, with implications that cannot be generalized to the entire population of refugee receiving countries. The findings presented here are particularly relevant to new refugee receiving countries faced with massive displacement situations, but they, also hold pertinence to traditional refugee receiving countries. Recent evidence from 20 European countries shows that the policy, economic, and migratory context is not adequate to explain the cross-national divergences in attitudes toward asylum policy (van Hootegem et al., 2020). This study suggests that social cohesion indicators can serve as an alternative explanation to predict divisions in asylum policy across European countries. Considering that this study does not offer specific indicators but rather general principles, future research can advance the research agenda by translating the context-based indicator framework introduced here into qualitative and quantitative measurements to assess the impact of social cohesion processes in refugee receiving countries. Finally, this study draws attention to the need for further studies on the causal mechanisms underlying social cohesion processes in refugee receiving settings, as a key, yet, overlooked dimension of indicator research.

AUTHOR'S NOTE
SO is the Director of the HMPI Research Lab and an Associate Professor in the Department of Political Science and Public Administration at Bilkent University. She conducts research and publishes on migration policy and politics in the European Union, Turkey, and Canada; health and immigration; gender and immigration; irregular immigration; integration and citizenship; media representation of migrants; comparative politics of deliberative democracy; and the Europeanization research agenda. SO has been a Visiting Scholar at the Department of Global Health and Social Medicine and Center for Middle Eastern Studies, Harvard University; post-doctoral researcher at the McGill Institute for the Study of Canada, and doctoral researcher at Center for European Studies, Universite Libre de Bruxelles.
JH is a Post-Doctoral Researcher at the HMPI Research Lab at Bilkent University. She is involved in a research project on the "Development of an Interactive Social Cohesion Governance Model in Turkey" (InScol Project) funded by the Scientific and Technological Research Council of Turkey (TÜBITAK). Her main research interests involve topics in comparative politics and public policy, more specifically collaborative governance, participatory democracy, policy networks, policy deliberation, local governance, health, migration and other public policy issues, with a special focus on developing countries.