REPATRIATION of REFUGEES from ARAB CONFLICTS: Conditions, Costs and Scenarios for Reconstruction

REPATRIATION OF REFUGEES FROM ARAB CONFLICTS: Conditions, Costs and Scenarios for Reconstruction

Ibrahim Elbadawi

Belal Fallah Jala Youssef Maryse Louis Roger Albinyana Samir Makdisi Semih Tumen

FEMISE Euromed Report 2019 FEMISE Euromed Report 2019

Repatriation of Refugees from Arab Conflicts: Conditions, Costs and Scenarios for Reconstruction

Ibrahim Elbadawi

Belal Fallah Jala Youssef Maryse Louis Roger Albinyana Samir Makdisi Semih Tumen First published in 2019 FEMISE and ERF

This report received financial support from the European Union through the FEMISE project on “Support to Economic Research, studies and dia- logues of the Euro-Mediterranean Partnership”. Any views expressed in this report are the sole responsibility of the authors. About the Authors

Ibrahim Elbadawi is managing director of the ERF. Earlier, he served as director of research at the Dubai Economic Council, which he joined in March 2009. Before that, he was lead economist at the Development Research Group of the World Bank, which he joined in 1989. He has published widely on macroeconomic and development policy, democratic transitions and the economics of civil wars and post-conflict transitions. He was the thematic leader of the Natural Resource Man- agement and Economic Diversification theme at the ERF for five years and is a member of the advisory board of the Arab Planning Institute. He holds a PhD in economics and statistics from North Carolina State and Northwestern Universities.

Belal Fellah earned his Ph.D. in 2008 in Regional Economics from the University of Saskatchewan- Canada. He currently works as a director of Research at Palestine Economic Policy Research Institute-MAS. His research interest and expertise cover regional and development economics, labor economics, and policy-program evaluation. During the past years, he has published a number of papers at international refereed journals including Journal of Development Economics, Jour- nal of Economic Geography, Papers in Regional Science, and Annals of Regional Science. He currently focuses his research on the Palestinian labor market, emphasizing the economic cost of conflict, commuting to the Israeli labor market, and linkages between the labor market and public sector

Jala Youssef is a Consultant in the World Bank, Cairo Office at the Macroeconomics, Trade and Investment Global Practice. She is also a part-time Assistant Lecturer at the French Section of the Faculty of Economics and Political Science, Cairo University. Prior to joining the World Bank in 2018, she worked as a researcher at the Economic Re- search Forum (ERF). She previously worked for the Egyptian Govern- ment at the Ministry of International Cooperation and the Egyptian Competition Authority. She holds a B.A. in Economics from the Fac- ulty of Economics and Political Science, Cairo University and Paris 1 Panthéon-Sorbonne University, and an M.A. in Applied Economics from Paris 1 Panthéon-Sorbonne University. Maryse Louis PhD, is the General Manager of the FEMISE association, (Marseille, France), which she joined since its inception in 1999. Dr. Louis is also a member of staff of the Economic Research Forum (ERF, Cairo, Egypt), coordinator of FEMISE. Maryse holds a PhD in Economics from Aix Marseille University, France. Her research interest covers issues of development with particular focus on migration and refugees, inequalities and women empowerment in the Med region. She also gained expertise on issues related to knowledge transfer and innovation

Roger Albinyana is director of the Mediterranean Regional Policies and Human Development department at the European Institute of the Mediterranean. He is also associate professor at the Faculty of Eco- nomics of the University of Barcelona where he teaches international economic policy. He holds a MA in Economic History from the Uni- versity of Barcelona and the Autonomous University of Barcelona, a BA in Economics from Pompeu Fabra University (Barcelona) and a BA in Business Administration and Management from the Open Univer- sity of Catalonia. From 2013 to 2016, he served as secretary general for Foreign and EU Affairs of the Government of Catalonia. Prior to that, from 2010 to 2013 he served as advisor on private sector development at the Secretariat of Union for the Mediterranean

Samir Makdisi is Professor Emeritus of Economics and Founding Director of the Institute of Financial Economics, the American Univer- sity of Beirut (AUB). He is a former Minister of Economy and Trade, Republic of Lebanon, 1992, Deputy President of AUB, 1992-1998 and chair of the Board of Trustees, Economic Research Forum (Cai- ro),1993-2001 and since 2012 an elected member of the Board; and member of the Board of Directors of the Euro-Mediterranean Forum of Economic Institutes (FEMISE, Marseille); he has served as advisor to national governments and to various regional and international economic and financial organizations.

Semih Tumen is an Associate Professor of Economics at TED Uni- versity. He is also affiliated with the ERF and IZA as a Research Fel- low. He received his PhD in Economics from the University of Chicago in 2012, MSc in Econometrics and Mathematical Economics from the London School of Economics and Political Science in 2006, and BSc in Economics from the Middle East Technical University in 2000. He served as the Director General of Economic Research at the Central Bank of the Republic of Turkey. His research mainly concentrates on Labor Economics with a special focus on Migration and Refugees. Contents

Introduction ...... 1

Chapter One: Who are the Arab refugees? ...... 5 1.1 Introduction and objectives ...... 5 1.2 Refugees by country of conflict ...... 6 1.3 Syrian refugees across the main host countries ...... 8 1.4 Conclusions ...... 28

Chapter Two: Enabling conditions, challenges, and processes ...... 31 2.1 Introduction ...... 31 2.2 Early repatriates ...... 32 2.3 Conditions for return ...... 37 2.4 Other considerations ...... 47 2.5 What does this mean for Arab repatriation? ...... 51 2.6 Conclusions ...... 58

Chapter Three: Repatriation: scenarios for conflict resolution and reconstruction ...... 65 3.1 Repatriation of refugees and their rights...... 67 3.2 3.2 Reconstruction: definitions, scenarios, and impact on refugees ...... 69 3.3 Review of the literature on political settlements ...... 74 3.4 Typology of political settlement scenarios ...... 81 3.5 Applying a political settlement typology to historic conflicts in some MENA countries ...... 83 3.6 The nexus between political settlements and returned refugees in South Med conflict-affected countries ...... 90 3.7 Conclusions ...... 97

Chapter Four: Repatriation: potential contributions to post-conflict economic development ...... 107 4.1 Introduction ...... 107 4.2 The economic situation in the conflict countries ...... 108 4.3 Results for Arab conflicts ...... 112 4.4 Post-conflict growth potential ...... 116 4.5 Review of the existing results and country experiences ...... 122 4.6 Concluding remarks and policy perspective ...... 129

Conclusions: Repatriation of refugees from the Arab conflicts: prospects and potential political and economic implications ...... 137

Introduction | 1

Introduction

The prospects for early retaken. This would help to support patriation of refugees who have the most positive outcomes – pri- fled conflicts in Arab countries marily for the refugees, but also in recent years do not yet look for other stakeholders, such as promising. The conflict in Yemen host communities and those left is at a stalemate; Libya is wedged behind in the conflict countries. in a power struggle between two With these considerations in military/political factions; Iraq mind, this year’s ERF-FEMISE is struggling to recover from de- Euromed Report discusses the is- cades of instability; and Syria re- sue of repatriation in general and mains a country at war. as a potential solution to the refu- Nevertheless, not only have gee crisis in South Med countries. discussions about repatriation This is accomplished through started at both national and in- undertaking a rounded approach ternational levels, but there is that begins with an understand- also a steady, though still limited, ing of the characteristics of refu- stream of refugees in neighbor- gees and repatriates, as well as ing countries trickling back to their decision-making processes. their war-ravaged homes. With This is followed by a thorough the doors of naturalization and analysis of scenarios for political resettlement all but closed and settlements and reconstruction, the socio-economic situation in and what they mean for repa- host countries weakening, the triation, with a focus on the role refugees have found themselves that the international community caught in very difficult circum- can play to ensure a sustainable stances. return. In particular, the Report While mass repatriation at attempts to answer the following this stage remains premature for questions: all four war-torn countries, the • What are the main character- current situation dictates that we istics of refugees from these recognize and unpack the issue countries? of repatriation in all its dimen- · What are the factors that affect sions, so that if and when the time refugees’ decisions to return or comes, informed actions can be not to their home countries? 2 | Introduction

· Under what conditions would disproportionately accommodate repatriation be considered a Syrian refugees, namely Turkey, sustainable solution for refu- Jordan, Lebanon, and Germany. gees’ problems in these countries? The aim of this chapter is to · What are the implications of shed light on the difficulties that different scenarios of conflict the refugees face in their host resolution and reconstruction countries, especially in terms of for repatriation? economic outcomes related to · What role can the international education, poverty, and the la- community play to ensure a bor market, and social outcomes sustainable return of refugees? in terms of acceptance by and integration into host communities (or lack thereof). It also high- From a policy perspective, the lights the diverging experience Report aims to contribute to de- of refugees depending on their bates on post-conflict repatriation country of asylum and whether and reconstruction in the South they live in camps or not. This Med region, highlighting some chapter sets the stage for the rest possible scenarios, their implica- of the report by highlighting the tions, and the potential role of the critical stage that the refugee cri- international community. There sis has reached, contextualizing is a particular focus on Syria the experience of refugees, and since its conflict has led to the stressing the urgency of finding a emergence of millions of refugees sustainable solution. who, at the time of writing, make up the majority of the total refu- Chapter 2 analyzes the vari- gee population in the region. ous dimensions of refugees’ decision-making processes about The Report also contributes whether or not to repatriate, with to the growing literature on the a particular focus on households enabling conditions, challenges, headed by women. It begins and processes of repatriation, with an analysis of the factors exploring it in the context of the that pushed (or pulled) early re- Arab refugee crisis, and thereby patriates to return to Iraq, Syria, filling a gap in knowledge on its and Yemen. It then undertakes economic and political impacts. a literature review of the vari- The Report is organized into ous security, political, economic, four chapters and a final set of and social considerations behind conclusions. refugees’ decisions about repa- Chapter 1 explores the demo- triation, and how the information graphic and socio-economic char- that guides these decisions are acteristics of refugees from South collected and processed. Med countries, with a particular In addition, the chapter briefly focus on Syrian refugees. It also discusses the economic and po- touches on the responses of host litical conditions for sustainable communities and the challenges repatriation for both the return- that they face. The analysis in this ees and for the home country. Fi- chapter includes countries that nally, the conditions for return – Introduction | 3

as reported by the Arab refugees mates, the chapter quantifies the themselves – is discussed, with a extent of economic destruction in focus on Syrian refugees. The aim the four conflict countries. of the chapter is to highlight the Post-conflict growth scenar- role of refugees as independent ios are discussed based on an decision-makers despite their extended version of the World limited agency, and to pinpoint Bank’s Long-Term Growth Mod- the factors that are their priori- el. These include an optimistic ties when considering whether, scenario that leads to quick eco- when, and how to return home. nomic recovery and a pessimistic Chapter 3 follows up with a scenario that drags out the cost of discussion of political settlements conflict for decades. Finally, the and reconstruction plans in the economic consequences of repa- region. This is accomplished triation for both the home and through a thorough understand- host countries are elaborated, ing of the literature on recon- along with a detailed analysis of struction and repatriation, a de- repatriation experiences from se- tailed mapping of stakeholders’ lected countries. goals and responses, and a brief The conclusions link together discussion of past experiences of the main findings of each chap- political settlement in the Middle ter, bringing out the Report’s co- East and North Africa (MENA) ordinated theme in addressing region. the major issues of repatriation. This is followed by a discus- In addition, they provide policy sion of potential scenarios of insights into how to address this political settlements and the ac- issue so as to ensure a sustainable companying reconstruction plans and dignified return of the refu- that would facilitate the process gees to their home countries. of sustainable repatriation, dis- tinguishing between refugees in neighboring countries and those in Western countries. The aim of the chapter is to give an informed, References realistic picture of the most favorable settlement and reconstruc- Crisp, Jeff, and Katy Long. 2016. tion plans that would ensure the ‘Safe and Voluntary Refugee sustainable return of refugees. Repatriation: From Principle Chapter 4 analyzes the eco- to Practice.’ J. on Migration & nomic costs of conflict, post-con- Hum. Sec. 4: 141. flict growth scenarios drawing Elbadawi, Hanan. 2018. ‘Syrian on the experiences of other South Refugees in Lebanon: Poten- Med countries, and the potential tial Forced Return?’ Atlantic contribution of repatriation to Council. 2018. http://www. political transition, post-conflict atlanticcouncil.org/blogs/ development, economic growth, menasource/syrian-refugees- and reconstruction, again with a in-lebanon-potential-forced- particular focus on Syria. Using return. satellite data and GIS-based esti- 4 | Introduction

Marks. 2018. ‘Pushing Syrian Refugees to Return.’ Carn- egie Endowment for Inter- national Peace. 2018. http:// carnegieendowment.org/ sada/75684. UNHCR. 2007. ‘Chapter 7: Solu- tions for Refugees’ in ‘The 10 Point-Plan in Action.’ UN- HCR. 2007. https://www.unhcr.org/publications/manuals/5846d10e7/10- point-plan-action-2016-update- chapter-7-solutions-refugees. html UNHCR. 1995. ‘Handbook on Voluntary Repatriation: Inter- national Protection.’ UNHCR. 1995. https://www.unhcr.org/publications/legal/3bfe68d32/ handbook-voluntary-repatriation-international-protection.html. CHAPTER 1

Who are the Arab refugees?

Belal Fallah

1.1 Introduction and objec- conflict in the other countries tives has mainly led to internal displacement. While the economic and social cost of displacement In 2011, a series of pro- is non-negligible, this chapter democracy uprisings erupted focuses on refugees, providing in several Arab countries, in- the background for a discus- cluding Bahrain, Egypt, Libya, sion of refugee repatriation in Morocco, and Tunisia. The Chapter 2. outcome for Libya and Syria Another reason for the fo- turned deadly, setting the cus on Syria relates to data stage for civil wars and paving availability: most published the way for terrorist groups, reports, studies, and data sets mainly ISIS, to control parts are limited to Syrian refu- of these countries. ISIS also gees. Therefore, to the extent expanded its territorial con- that data allow, we will shed trol into Iraq, capturing popu- light on non-Syrian refugees lated cities, such as Mousol. covering basic demographic Yemen has also fared badly: indicators. Then a more thor- political competition between ough discussion will focus on the Houthi movement and the Syrian refugees in major host government has turned into countries, including Turkey, bloody civil war. The outcome Jordan, Lebanon, and Western of these conflicts, in which Europe. In this context, this hundreds of thousands have chapter also explores the ex- been killed and millions have tent to which Syrian refugees become refugees, is devas- have integrated in host coun- tating. Syria has suffered the tries, focusing on the percep- most. tions of both the natives and This chapter explores the the Syrian refugees. salient demographic and so- The analysis provides an cio-economic characteristics insight into the economic con- of refugees from conflicts in ditions of refugees, as well these countries. The discussion as helping to understand the disproportionately highlights challenges that they face. The the case of Syrian refugees, results are vital to inform poli- mainly because of scale: they cy-makers about issues related number over six million, while to providing education and 5 6 | Who are the Arab refugees?

healthcare, which are key de- ern Syria, namely Idleb, Hama, terminants of household wel- Deir ez-Zor, Aleppo, Raqqa, fare and future labor market Homs, Al-Hassakeh, and Lat- outcomes. They can also help takia governorates. A total of with the preparation of poli- 6.2 million Syrians remained cies to improve socio-econom- internally displaced,1 out of ic conditions and lessen social which 5.3 million still live in and economic costs await- poor shelters and 4.2 million ing the end of conflict in the need shelter support.2 refugees’ home countries and In addition, about 6.4 mil- eventually their return home. lion Syrians, registered refugees, have fled for their lives, mainly to neighboring coun- 1.2 Refugees by country of tries (see Figure 1.1). Turkey has conflict received half of them. Some of the Syrian refugees have set- The eruption of the Syrian tled in Europe, residing mainly conflict in 2011 has generated in Germany and Sweden. Re- massive forced migration and cent estimates show that Eu- internal displacement. Be- rope accommodates 11.6% of tween July 2017 and June 2018, the total Syrian refugees and over 1.2 million Syrians were asylum-seekers in the world.3 displaced mainly from North-

Figure 1.1: Distribution of Syrian refugees across host countries

%60

%50

%40

%30

%20

%10

%0 Lebanon Jordan Iraq Egypt Turkey Germany Sweden Other

Countries

Source: Syria Regional Refugee Response: Inter-agency Information Sharing Portal: http://data.unhcr.org/syrianrefugees/asylum.php (Erişim: 21.11.2017)/ T.C. İçişleri Bakanlığı Göç İdaresi Genel Müdürlüğü (GİGM): http://www.goc. gov.tr/icerik6/temporary-protection_915_1024_4748_icerik (Erişim: 21.11.2017)

1 UN OCHA. (2018, August). 2018 Humanitarian Response Plan (January -December 2018): Syrian Arab Re- public. https://reliefweb.int/sites/reliefweb.int/files/resources/2018_2018_hrp_syria.pdf. 2 UN OCHA. (2018, August). 2018 Humanitarian Response Plan (January -December 2018): Syrian Arab Re- public. https://reliefweb.int/sites/reliefweb.int/files/resources/2018_2018_hrp_syria.pdf. 3 UNHCR. (2018, April). Comprehensive protection and solutions strategy: A roadmap to advance resettlement and complementary pathways in the Syrian crisis. https://data2.unhcr.org/en/documents/details/63759. Chapter 1 | 7

The conflict in Yemen has also having returned, resettled, or caused large forced migration, locally integrated.5 In terms of although, unlike in Syria, most Yemeni refugees, data from of the refugees are internally dis- the UNHCR show that by May placed. According to the United 2019, 318,000 Yemeni had left Nations High Commissioner on the country, with the largest sec- Refugees (UNHCR),4 in mid- tion having fled to Ethiopia 2018, 2.1 million Yemeni had ex- (see Figure 1.2). The case in perienced internal displacement, Libya is similar: the total num- mostly in Houthi controlled ter- ber of the internally displaced ritories, such as Hodeida and in 2018 amounted to 240,000 out Taizz. Still, displacement was of which about 63,000 managed temporarily or partially resolved to return. The number of Libyan for almost half of them. This co- refugees did not exceed 10,000.6 hort includes those identified as

Figure 1.2: Distribution of Yemeni refugees across main host countries

%50.00 %45.00 %40.00 %35.00 %30.00 %25.00 %20.00 %15.00 %10.00

%5.00 %0.00 Ethiopia Oman Somalia Saudi Arabi Djibouti Sudan

Source of data is UNHCR Operational Portal Refugee Situations. https://data2.unhcr.org/en/situations/yemen

4 See UNHCR Mid-Year Trends 2018. 5 See http://www.internal-displacement. https://www.unhcr.org/statistics/ org/countries/yemen. unhcrstats/5c52ea084/mid-year- 6 See UNHCR Mid-Year Trends 2018. trends-2018.html. https://www.unhcr.org/statistics/ unhcrstats/5c52ea084/mid-year- trends-2018.html. 8 | Who are the Arab refugees?

In Iraq, years of conflict, cul- 1.3 Syrian refugees across minating with ISIS controlling the main host countries Mousol, have induced major displacement across the country, The aim of this section is to mainly in the Kurdistan region. explore the demographic and Between 2014 and 2018, over socio-economic characteristics three million Iraqi have been of the Syrian refugees. Depend- displaced, with 360,000 residing on data availability, we ing in informal settlements and carry out this exercise per host abandoned buildings.7 The con- country, where most Syrian ref- flict has also driven over 280,000 ugees have settled, focusing on refugees into neighboring coun- Turkey, Jordan, Lebanon, and tries, mainly Turkey. A UNHCR Germany. We should stress up- report8 in 2016 shows that close front that the extent of analysis to 127,000 Iraqi refugees had set- that we undertake is largely de- tled in Turkey, out of which 43% termined by available indicators are children, 23% are women, for each host country. In this re- and 34% are men (see Figure 1.3). spect, more discussion will be dedicated to Syrian refugees in Figure 1.3: Age composition of Jordan, benefiting from the rich Iraqi refugees in Turkey, 2016 data sets of the Jordanian La- bor Market Panel Survey-2016 %60 (JLMPS). We also use available data from the Syrian Barometer %50 Survey (SB-2017) to undertake %40 the analysis in Turkey and from Statistisches Bundesamt-2016 %30 Females for Germany. As for Lebanon, Males %20 we also use data, documented in existing reports, mostly pub- %10 lished by international institu- %0 tions, mainly the UNHCR. In 0 to 4 5 to 11 12 to 17 18 to 59 60 and addition, to these data sources, above we resort to an existing litera- Source: UNHCR Turkey Factsheet- Au- ture review on Syrian refugees gust 2018 https://reliefweb.int/report/ to complement our analysis. turkey/unhcr-turkey-factsheet-august-2018. 1.3.1 Syrian refugees in Jordan According to the Jordanian population census of 2015, the number of Syrian refugees in Jordan amounts to 1.3 million. 7 See https://www.unhcr.org/iraq-emergency.html. About 660,000 of them are UN- 8 See https://reliefweb.int/report/turkey/ HCR registered refugees. Of unhcr-turkey-factsheet-august-2018. them, 21% have settled in formal Chapter 1 | 9

camps (mainly in Za’atari, Az- istics of Jordan residents, includ- raq, and the Emirati Jordanian ing natives, migrants, as well as Camp), while only 16,000 Syr- Syrian refugees. Starting with ian refugees settled in informal age composition (see Table 1.1), settlements that lack necessary the data suggest that close to infrastructure and services. The half of Syrian refugees in Jordan rest settled in host communities.9 are children (younger than 15), Most of the Syrian refugees constituting a higher share than in the Za’atari camp (85%) came the corresponding Jordanian from Daraa governorate, close share. The share of male Syrian to the northern border of Jor- refugee children (55%) is slight- dan, while the rest come from ly larger than the female share rural Damascus (10%), urban (49%), although, the data show Damascus (2%), and Homs (2%). that the youth cohort (between Markedly, most of refugees in 15 and 29) is disproportionately this camp (87%) come from ru- female. Many young males have ral areas. Those settled outside remained in Syria, partly to take refugee camps come, mostly, part in the conflict or they have from Daraa (47%), Homs (26%), chosen to settle in other host rural Damascus (7%) and ur- countries. As for older cohorts11, ban Damascus (7%). Of this sec- the age composition for males tion of refugees, 58% come from and females are similar. rural areas: they were farmers and tend to be poorer relative to Table 1.1: Age composition of Syrian refugees in Jordan those coming from urban areas.10 Significantly, socio-economic Age Jorda- Syrian Proportion Syrian conditions of the refugees dictate group nian females males the extent to which refugees cope 05_00 22% 14% 22% 22% with economic consequences associated with their presence in 11_06 20% 14% 16% 24% these host countries (see more 14_12 8% 6% 6% 9% discussion below). 19_15 10% 11% 10% 9%

1.3.1.1 Demographic charac- 29_20 11% 17% 16% 7% teristics 39_30 15% 12% 15% 15% The analysis in this section 49_40 9% 11% 8% 10% heavily relies on the JLMP-2016, a rich data set on the demograph- 59_50 2% 6% 2% 2% ic and socio-economic character- 64_60 2% 3% 2% 1% 9 See UNHCR (2017) Syria Regional 65>= 2% 5% 2% 2% Refugee Response: Inter-Agency In- formation Sharing Portal (Jordan). Total 100% 100% 100% 100% http://data.unhcr.org/syrianrefugees/country.php?id=107c Source of data is JLMP-2016 survey. 10 See Stave, Svein Erik, and Solveig Hillesund. (2015) Impact of Syrian 11 Caroline Krafft, Maia Sieverding, Co- Refugees on the Jordanian Labour lette Salemi, and Caitlyn Keo (2018) Market. Geneva: International Labour ‘Syrian Refugees in Jordan: Demo- Office (ILO), Fafo Institute for Ap- graphics, Livelihoods, Education, and plied International Studies. Health’. ERF Working Paper No 1184. 10 | Who are the Arab refugees?

As for other demographic ment rate of school-age Syrians characteristics, about 19% of Syr- (6-17), it varies considerably by ian household heads are women; place of settlement. Table 1.2 slightly higher than for Jordani- shows that among the formal ans (17%). Of the former, 52% are refugee camps, Za’atari camp married, 16% are divorced, 25% got the lowest enrollment rate are widows, while the remaining of 10%. In host communities, share are single women. Mark- about half of school-age children edly, about 8% of married Syr- enrolled in schools. The lowest ian women who live in camps enrollment rate is reported in are younger than 19 relative to informal settlements in Mafraq, 2.5% for those residing in urban plummeting to about 3.5%16 and or rural areas. Families in refu- close to zero enrollment in infor- gee camps may have resorted mal settlement in the Jordan val- to arranging early marriages as ley.17 a coping mechanism to mitigate While existing infrastructure 12 economic hardship. Early mar- might be insufficient to enroll all riage is linked to school dropout school-age children, poor eco- and associated with a negative nomic conditions are most likely social impact for married girls to play a vital role. Many fami- 13 and their offspring. lies of Syrian refugees may have In 2017, the number of Syr- forced their children to drop out ian children enrolled in formal of school and enter the labor schools was 125,000 (UNICEF market as a coping mechanism and MECI, 2017).14 UNICEF has to maximize earnings. Approxi- played a vital role in providing mately 30,000 children were educational infrastructure in for- reported as child labor, mostly mal camps, while the Jordanian engaged in informal activities, Ministry of Education provides mainly working in agricultural Jordanian teachers. Syrian stu- activities or selling items on the dents, who have been enrolled street.18 in government schools in the host communities, either integrated with their Jordanian peers 16 REACH (2014) ‘Syrian Refugees Stay- or were absorbed in evening ing in Informal Tented Settlements in 15 Jordan: Multi-Sector Assessment Re- shifts. In terms of the enroll- port.’ REACH.

12 See Women’s Refugee Commission. 17 For more discussion on education (2014) ‘Unpacking Gender: The Hu- condition see Salemi, Colette, Jay manitarian Response to the Syrian Bowman, and Jennifer Compton Refugee Crisis in Jordan’. (2018) ‘Services for Syrian Refugee Children and Youth in Jordan: Forced 13 See Lee-Rife, S., Malhotra, A., Warner, A., and Glinski, A. M (2012) ‘What Displacement, Foreign Aid, and Vul- Works to Prevent Child Marriage: A nerability.’ Economic Research Forum Review of the Evidence’. Studies in Working Paper .1188. Family Planning, 43(4), 287–303. 18 See UNHCR, UNICEF, and WFP. 14 UNICEF, and MECI (2017) ‘Education (2014) ‘Joint Assessment Review of Sector Dashboard. Inter-Agency Coor- the Syrian Refugee Response in Jor- dination Home Page’ http://scs.raisun- dan’, 1–34. See also UNICEF (2013) hcr.org/dashboards/Education.aspx ‘Syrian Refugee Children in Jordan: 15 See Human Rights Watch ( 2016) ‘‘We’re Assessments Desk Review on the Afraid for Their Future’ - Barriers to Ed- Situation of Syrian Refugee Children ucation for Syrian Refugee Children in Living in Jordan’. Amman. Jordan.’ Human Rights Watch. Chapter 1 | 11

Table 1.2: School enrollment of Syrian refugees in refugee camps and host communities

Location Number of school-age children Enrollment rate Za’atari camp 211,287 10% Azraq camp 11,782 91% Emirate J camp 2,912 82% Host community 169,861 54% Source of data: the figures in this table, based on estimates from MOE and UNHCR, are reported in Salemi et al (2018).19

Table 1.3 explores another an- gees will compete over jobs with gle of human capital, focusing on natives who share similar labor educational attainment across age characteristics, mainly education. cohorts. This exercise is impor- The data show that the education- tant to understand the linkages al attainment for the Syrian refu- between refugee influx and labor gees is lower across all age cohorts market outcomes for Jordanians. relative to Jordanians. In theory, it is expected that refu-

Table 1.3: Educational attainment across cohorts for Syrian refugees and Jordanians

Syrian Illiterates Read and Basic edu- Secondary Graduate total refugees write cation 19_15 19% 69% 12% 0% 100% 29_20 11% 68% 11% 7% 4% 100% 39_30 21% 66% 3% 3% 8% 100% 49_40 22% 42% 17% 15% 3% 100% 59_50 36% 42% 14% 5% 4% 100% 64_60 39% 11% 12% 19% 18% 100% 65>= 71% 11% 1% 0% 17% 100% Jordani- Illiterates Read and Basic edu- Secondary Graduate Total ans write cation 19_15 3% 21% 64% 12% 0% 100% 29_20 4% 11% 35% 19% 31% 100% 39_30 7% 12% 30% 20% 31% 100% 49_40 7% 16% 30% 22% 25% 100% 59_50 11% 21% 24% 17% 27% 100% 64_60 25% 19% 14% 13% 28% 100% 65>= 48% 19% 13% 5% 15% 100% Source of data is JLMP-2016 survey.

19 See Salemi, Colette, Jay Bowman, and Jennifer Compton (2018) ‘Services for Syrian Refugee Children and Youth in Jordan: Forced Displacement, Foreign Aid, and Vulnerability.’ Economic Research Forum Working Paper .1188. 12 | Who are the Arab refugees?

1.3.1.2 Economic characteristics and females, respectively. Table In this section, we assess sev- 1.4 displays differences in LFPR eral labor market outcomes for across age cohort between the the Syrian refugees and compare Syrian refugees and Jordanians. them with Jordanians. These in- The data, limited to men, show clude labor force participation that LFPR is slightly lower for rate (LFPR), unemployment, Syrians for individuals younger employment characteristics, and than 30. The gap widens for old- wages. This helps to indicate the er cohorts. It is possible that Syr- economic wellbeing of the Syrian ians older than 49 live off aid or refugees, key aspects to under- labor income earned by younger stand the likelihood of their re- family members. patriation. Typically, with poor As for the unemployment economic conditions, many Syr- rate, it amounts to 22% for the ian refugees are expected to re- Syrian refugees relative to 13% turn home as security conditions for Jordanians. Comparing un- in Syria improve. employment rate across cohort The data show that Syrian ref- shows a distinct pattern: it is ugees generally suffer poor labor high for all age cohorts, except market outcomes. LFPR20 for the for the 30-39 age group. In ad- Syrian refugees is 22% compared dition, both Syrian refugees and with 36% for Jordanians. Most of Jordanians share a similar un- Syrian labor force participants employment rate for the 15-19 are male (49%) compared with cohort. The unemployment rate 3% for females. The correspond- gap widens for older cohorts, ing LFPR for Jordanians is much more prominently for those old- higher (59% and 16%) for male er than 39.

Table 1.4: Male labor force participation rate and unemployment rate for Syrian refugees and Jordanians by age

LFPR Unemployment Rate age Syrian Jordanians Syrian Jordanians 19_15 10.14% 13.39% 50.85% 52.51% 29_20 60.57% 64.31% 27.27% 21.89% 39_30 63.89% 85.82% 9.64% 7.80% 49_40 55.11% 69.56% 31.01% 6.95% 59_50 17.80% 50.26% 48.28% 3.40% 64_60 1.92% 33.33% 41.54% 0.00%

Source of data is JLMP-2016 survey.

20 Labor force participants are those who either work are actively seek employment during the reference week. Chapter 1 | 13

The JLMP data also make it sified as employers. In terms of possible to analyze the spatial distribution across occupations, distribution of employment. Fig- the JLMPS data show that the ure 1.4 shows that half of Syrian majority of working Syrian refu- refugee workers settle in Am- gees is employed in craft trade oc- man, followed by Irbid, and Zar- cupations, followed with a wide qa. The overwhelming majority margin by services and sales as of Syrian refugees (86%) are wage well as elementary occupations, workers, while 10% are classified respectively (see Table 1.5). as self-employed and 2% clas-

Figure 1.4: Employment share of Syrian refugees across governorates in Jordan

%60

%50

%40

%30

%20

%10

%0 Amman Balqa Zarqa Madaba Irbid Mafraq Karak

Source of data is JLMP-2016 survey.

Table 1.5: Occupation distribution across Syrian refugees

Occupation Proportion Professionals 5.1% Technicians and associate professionals 0.46% Clerical support workers 1.11% Service and sales workers 19.44% Skilled agricultural, forestry and fish 2.01% Craft and related trades workers 47.79% Plant and machine operators, and assembly workers 7.64% Elementary occupations 16.46%

Source of data is JLMP-2016 survey. 14 | Who are the Arab refugees?

Markedly, most of the Syrian and forced Syrian refugees to workers are employed in the in- reside in poor quality and over- formal sector (82%); substantial- crowded apartments.23 The effect ly greater than Jordanians (34%). of Syrian refugees on the Jorda- While this might indicate that nian housing market has been a Syrian refugees compete with central interest of researchers. informal Jordanian workers, re- Al Hawarin et al (2018)24 show cent research shows otherwise that the Syrian refugees have (Fallah et al, 2019).21 Job compe- negatively affected housing con- tition appears to take place with ditions of the locals such that immigrants. Malaeb and Wah- poorer households are dispro- ba (2018)22 show that Syrians portionately affected. They also refugees mainly compete with show that rent increased sig- economic immigrants, mainly nificantly in regions closer to the Egyptians, in the informal sec- Syrian borders. tor. Evidently, 63% of Egyptian Syrian refugees are highly immigrants are employed in the dependent on direct support, informal sector. Limited chances cash or in kind, for basic need.25 to work in the formal sector seem Means of support vary cover- to have negatively affected wag- ing informal assistance, seasonal es for the Syrian refugees: their emergency cash assistance, and median hourly wage amounted regular cash transfers. The lat- to JD 1.15 relative to JD 2 for Jor- ter benefits are eligible for the danians. With low earning and UNHCR registered refugees. high unemployment rate, pov- Monthly cash transfers in 2016 erty has afflicted Syrian refugees. ranged between $112 and $218 Data from the Vulnerability per household. But without Assessment Framework (VAF) down playing the significance show that 86% of the Syrian refu- of the cash support, it does not gees were living below the Jor- cover all aspects of basic needs.26 danian poverty line (68 JD per 23 Overseas Development Institute, month) (see UNHCR, 2015). In UNICEF, and UNHCR. (2017). A terms of spending, rent, for those promise of tomorrow: Effects of UN- living outside refugee camps, ac- HCR and UNICEF cash assistance on Syrian refugees in Jordan. counts for the largest share (69% http://data.unhcr.org/syrianrefu- of monthly expenditure). Hous- gees/download.php?id=14564. ing supply constraint has driven 24 Alhawarin I,. Assad R,. and ElSayed up rent, mainly in urban areas A., (2018) ‘Migration Shocks and Housing: Evidence from the Syrian 21 Fallah, B., Krafft, C, and Wahba, J., Refugee Crises in Jordan’ Economic (2019) ‘The Impact of Refugees on Research Forum, Working Paper No. Employment and Wages in Jordan.’ 1213. Journal ‘Impact of Refugees on Immi- 25 UNHCR, and PU-AMI. (2017). Basic grants’ Labor Outcomes of Develop- Needs Sector Dashboard. http://scs. ment Economics, 139: 203-2016. raisunhcr.org/dashboards/Basic- 22 See Malaeb, Bilal, and Jackline Wah- Needs.aspx. ba. 2018.’ Economic Research Forum 26 See UNHCR. (2016). UNHCR Cash As- Working Paper Series 1194. Cairo, sistance: Improving Refugee Lives and Egypt. Supporting Local Economies. Amman. Chapter 1 | 15

Still, with financial constraints, 1.3.1.3 Social integration some Syrian refugees remain on As the Syrian conflict broke the waiting list.27 out and Syrians started to flee Recent statistics show that to Jordan, welcoming sentiment 40% of the Syrian refugees in defined the relationship with Jordan indicated that their liv- Jordanians. But tension started ing conditions have deteriorat- to evolve as the conflict persisted ed since their arrival. The high and the flow of Syrian refugees cost of living, the security situ- continued to accumulate. The ation, and lack of assistance are reasons mainly hinge on anxiet- the key determinants. Many of ies around the belief that Syrians those reporting deteriorating are crowding out Jordanians in 30 conditions live in Zarqa. Fac- the labor market as well as per- tors such as lack of assistance, ceived competition over public resources, water, education, and both in terms of value and 31 quantity, and cuts in the UN- health facilities. These have lessened the likelihood of social HCR aid and coupons are the integration. Based on polls, pro- main drivers.28 To cope with duced by CSS and Mercy Corps the deteriorating conditions, and UK FCO Conflict Pool (2015), over half of the adult Syrian two thirds of Jordanians prefer to refugees in Jordan indicated have a Jordanian neighbor, while that they have restricted their the nationality of neighbors did own food consumption for the not make a difference to most benefit of their children. Other Syrian refugees. coping mechanisms include re- sorting to cheaper and lower Using the same poll, Table 1.6 quality food, reducing the num- documents perceptions of Jor- ber of meals per day, borrowing danians and Syrian refugees of money, and selling household the linkage between the increase assets.29 in the number of Syrian refugees and the level of security in Jordan. The results suggest that unlike the Syrian refugees, most 27 See Salemi, Colette, Jay Bowman, and Jennifer Compton (2018) ‘Services for Jordanians believe that the level Syrian Refugee Children and Youth in of security will decrease. Consis- Jordan: Forced Displacement, Foreign tently, Shteiwi et al (2014)32 show Aid, and Vulnerability.’ Economic Re- search Forum Working Paper .1188. 30 See Razzaz, S. (2017) ‘A Challenging 28 CARE International. (2018). 8 years Market Becomes More Challenging: into exile: How urban Syrian refu- Jordanian Workers, Migrant Workers, gees, non-Syrian refugees and vulner- and Refugees in the Jordanian Labour able host communities in Jordan are Market’. Amman. coping and meeting challenges, eight 31 See REACH. (2014) ‘Education and years into the Syria crisis. Tensions in Jordanian Communities https://data2.unhcr.org/en/docu- Hosting Syrian Refugees: Thematic ments/ download/66543 Assessment Report’. Amman. 29 CARE International. (2017). 7 years 32 Shteiwi, M., Walsh, J., Klassen, CH into exile: How Syrian refugees, vul- (2014) ‘Coping with the Crisis. A Re- nerable Jordanians and other refugees view of the Response to Syrian Refugees in Jordan are being impacted by the in Jordan. Center for Strategic Studies, Syria crisis. https://data2.unhcr.org/ University of Jordan. http://www.jcss. es/ documents/download/60392 org/Photos/635520970736179906.pdf. 16 | Who are the Arab refugees?

that negative sentiments towards share of Jordanians who do not receiving more Syrian refugees support receiving more Syrian in Jordan increased between refugees increased from 66% to 2012 and 2014. In particular, the 79%.

Table 1.6: Linkages between influx of Syrian refugees and level of security

Do you think that with the increase in the number of Syrian refugees, the level of security in the region where you live in Jordan will decrease? 2013 2015 Syrian Syrian Jordanian (%) Jordanian (%) (%) (%) Yes, it will decrease 71.2 20.9 57 11.4 No, it will stay the same 27 74.4 36.5 81.5 No, it will increase 1.4 4.1 6.5 7.1 No answer 0.3 0.7 0 0 100% 100% 100% 100%

Source of Data: CSS, Mercy Corps and UK FCO Conflict Pool, 2015

1.3.2 Syrian refugees in Lebanon ProGres database, which allows us to explore a limited number of 1.3.2.1 Demographic and socio- demographic and socio-econom- economic characteristics ic characteristics of Syrian refu- The estimated number of gees in Lebanon. These include registered Syrian refugees in age composition, educational 33 Lebanon is about one million, attainment, marital status, and which makes up about 25% of employment distribution per oc- Lebanon’s population. Geo- cupation.34 Unlike in Jordan, the graphically, most refugees re- data do not allow comparisons side in Bekaa (36%), North Leba- with Lebanese cohorts. non (25%), and Mount Lebanon Figure 1.5 displays the age (25%). The following discussion composition for both males and explores the demographic and females, showing that the larg- socio-economic characteristics of Syrian refugees in Lebanon. 34 The statistics exhibited in the section The documented statistics in this are taken from: Verme, Paolo; Gi- section comes from the UNHCR gliarano, Chiara; Wieser, Christina; Hedlund, Kerren; Petzoldt, Marc; Santacroce, Marco. 2016. The Welfare 33 See UNHCR (2016). Syria Regional of Syrian Refugees: Evidence from Refugee Response. Inter-agency Infor- Jordan and Lebanon. Washington, mation Sharing Portal. http://data. DC: World Bank. © World Bank. https:// unhcr.org/syrianrefugees/country. openknowledge.worldbank.org/han- php?id=122. dle/10986/23228 License: CC BY 3.0 IGO.’ Chapter 1 | 17

est cohort consists of children The average household size (younger than 15 years), fol- of Syrian refugees in 2018 is 4.9, lowed by the 15-34 age cohort. representing a substantial de- In terms of marital status, 60% cline relative to 2015 (5.3). This of the Syrian refugees are single declining trend suggests a ten- compared with 37% who are dency for Syrian households, married. More recent statistics since their arrival in Lebanon, show that of all registered Syrian moving more towards nuclear refugees in Lebanon, (81%) are households.36 women and children.35

Figure 1.5: Age composition of Syrian refugees in Lebanon

%70

%60

%50

%40 Male %

%30 Female %

%20

%10

%0 4-0 14-5 54-45 44-35 34-25 24-15 64-55 +65

Source of data: ProGres data.

Syrian refugees in Lebanon labor market. The data show are generally low educated. that the highest educational at- Figure 1.6 exhibits educational tainment of the vast majority of attainment by age cohort. As Syrian refugees is primary edu- highlighted above, education cation. It is worth noting that composition indicates how the two-thirds of Syrian children re- influx of Syrian refugees is ex- ceived no education and that 44% pected to affect the Lebanese of school-age children have not

35 UNHCR. (2018). Operational portal 36 See UNHCR, UNICEF, and WFP - Refugee situations: Syria Regional (2018) VASyR 2018: Vulnerability As- Refugee Response: Lebanon. https:// sessment of Syrian Refugees in Leba- data2.unhcr.org/en/situations/syr- non. https://data2.unhcr.org/en/ ia/location/71. documents/download/67380. 18 | Who are the Arab refugees?

attended school.37 Many Syrian FPA), the American University children are sent to work as street of Beirut, and Sawa for Develop- peddlers and beggars to mitigate ment and Aid, show that over a the dire economic conditions that third of the Syrian refugee girls face their household (ILO, 2014). between the ages of 20 and 24 Another copying strategy is child years old married before the age marriage. Statistics from A recent of 18.38 The corresponding rate, study, conducted by the United among the Syrian families, pri- Nations Population Fund (UN or to the conflict is one fourth.39

Figure 1.6: Educational attainment of Syrian refugees: across cohort analysis

%100 %90 %80 %70 %60 Primary %50 Secondary %40 Tertiary %30 %20 %10 %0 24-15 64-25 +65

Source of data: ProGres data.

In terms of employment, the ementary occupations (20%). On distribution of Syrian refugees the other hand, close to 90% of across occupation differs by sex. employed females work as ser- Table 1.7 shows that most of the vice and sales workers. Work- male refugees are employed in ing conditions for the Syrian plant/machine operators, and refugees in Lebanon are poor. A assembler activities (50%), fol- lack of job opportunities forces lowed with a wide gap by el- many Syrian refugees to work

37 See World Food Programme (WFP), United Nations Children’s Fund 38 See https://www.unfpa.org/news/ (UNICEF), and United Nations High new-study-finds-child-marriage-ris- Commissioner for Refugees (UN- ing-among-most-vulnerablesyrian- HCR). 2013. ‘Vulnerability Assess- refugees ment of Syrian Refugees in Lebanon.’ 39 See https://www.unfpa.org/news/ WFP, Rome. Available at: https:// new-study-finds-child-marriage-ris- data.unhcr.org/syrianrefugees/down- ing-among-most-vulnerable-syrian- load.php?id=3853. refugees Chapter 1 | 19

in exploitative jobs, mostly infor- which 82% borrowed money in mal, and accept low wages.40 the fourth quarter of 2017.43 No- The labor market characteris- tably, financial challenges have tics of Syrian refugees in Lebanon also forced Syrian refugees to cut seem to shape the economic effect back on their housing expenses of the Syrian influx. David et al via shifting towards non-resi- (2018)41 apply a general equilib- dential structures, such as tented rium model to explore the effect settlements, active construction on the labor market, production sites, and agriculture rooms apparatus, and macroeconomic (engine and pump rooms). The indicators. They find that the share of Syrian households resid- effect is negative and mostly ing in such structures increased 44 limited to the most vulnerable from 26% in 2017 to 34% in 2018. Lebanese workers. Interestingly, they show that providing aid in Table 1.7: Occupation distribution for male and the form of investment enhances female Syrian refugees economic growth and improves labor market prospects. Occupation Women Men In line with low earning re- Managers 0 2% percussions, statistics from the Professionals 2% 2% 2018 Vulnerability Assessment Technicians and associate of Syrian Refugees in Lebanon 2% 5% show that 69% of the Syrian professionals refugee households are below Clerical support workers 0 0.50% 42 the poverty line. Borrowing to Service and sales workers 89% 5% meet basic needs is also common Skilled agricultural forestry among Syrian refugees. About 3% 7% and fishery workers 90% of Syrian refugee house- Craft and related trades holds acquired debts, out of 1% 8% workers Plant and machine opera- 40 See ILO (2014) ‘Assessment of the 0% 50% Impact of Syrian Refugees in Leba- tors, and assemblers non and Their Employment Profile.’ International Labour Organization, Elementary Occupations 2% 20% Regional Office for the Arab States. 41 See Anda David , Mohamed Ali Ma- Source of data: ProGres data. rouani, Charbel Nahas3 and Björn Nilsson (2018) ‘The Economics Of The Syrian Refugee Crises in Neighboring 43 UNHCR, UNICEF, and WFP. (2018, Countries: The Case Of Lebanon’ December). VASyR 2018: Vulnerabil- 42 For more discussion on poverty in- ity Assessment of Syrian Refugees in cidence among Syrian refugees in Lebanon. https://data2.unhcr.org/ Lebanon, see Yassin, N (2018) ‘101 en/documents/download/67380. Facts and Figures on the Syrian Refu- 44 UNHCR, UNICEF, and WFP. (2018, gee Crises’. Issam Fares Institute for December). VASyR 2018: Vulnerabil- Public Policy and International Af- ity Assessment of Syrian Refugees in fairs American University of Beirut. Lebanon. https://data2.unhcr.org/ http://www.aub.edu.lb/ifi. en/documents/download/67380. 20 | Who are the Arab refugees?

1.3.2.2 Social attitudes and per- and legal status. In terms of safe- ceptions of the host and refu- ty, the results show that it varies gee communities by regions: 34% of Syrians in Bei- As indicated above, social rut feel unsafe, compared with cohesion and integration of Syr- 30% in the North, 27% in Bekaa, ian refugees in their host com- 18% in Mount Lebanon, and 13% munities are largely driven by in the South. natives’ perceptions, which are The majority of Syrian refu- often driven by a multitude of gees believe that illegal status factors. These include competi- negatively affects their safety. tion over economic and public Furthermore, 76% of Syrian refu- resources, factors that are linked gees feel unwelcome in Beirut, to the level of development at relative to 60% in Bekaa, and the host community. The level of 50% in Mount Lebanon and the development in Lebanon varies South. Markedly, over half of geographically: North Lebanon the Syrian refugees have indi- and Bekaa are considered the cated that the situation is getting poorest. These localities are the worse. In terms of Lebanese per- main host communities for the ceptions, one investigated aspect Syrian refugees. Therefore, it is is hiring Syrian refugees. The re- expected that competition with sults show that 46% of Lebanese the Syrian refugees over public recruiters are not willing to hire resources and jobs as well as a Syrian refugees, while 28% indi- rise in housing rent are driving cated that they would, 18% indi- factors that would amplify lo- cated that they might hire Syr- cals’ negative perceptions.45 ians, and the rest are willing to hire Syrians if it saves cost. The Alsharabati and Nammour majority of those who reported (2015)46 conducted a survey to that they would not hire Syrians explore perceptions of Syrian (69%) justified their decision as refugees about a number of social Syrian refugees take Lebanese aspects including safety, welcome jobs, and 21% indicated that they perception, among other matters. do not like working with Syrians, They also explored perceptions The rest were concerned with a of Lebanese people in host com- lack of work permits. munities regarding access to services and perceived threats to local communities. They show that 1.3.3 Syrian refugees in Turkey Syrian refugees are concerned the 1.3.3.1 Demographic and socio- most for their safety, followed in economic characteristics order by economic conditions, By end of 2017, the number of Syrian refugees in Turkey was 45 See Ruisi and Shteiwi (2016) ‘Econom- 3.7 million. Most of the refugees ic and Social Integration of Migrants and Refugees in Jordan and Lebanon. are settled in host communities 46 See AlSharabati, C., and Nammour, J of major Turkish cities (Kayaog- 47 (2015) Survey on Perceptions of Syr- lu, 2019). To date, readily avail- ian Refugees in Lebanon. Institut de Sciences Politiques – USJ, Volume 1.0. 47 See Kayaoglu Aysegul (2019) ‘Labor https://data.unhcr.org/syrianrefu- Market Activities of Syrian Refugees gees/download.php?id=9399. in Turkey’. Working Paper No. 1290. Chapter 1 | 21

able official statistics on the de- Syrian refugees settled in refugee mographics and socio-economic camps, while the rest settled in characteristics of Syrian refu- host communities. Table 1.8 ex- gees is lacking. We rely on Syr- hibits the distribution of Syrian ian Barometer 2017 (SB-2017) to refugees, in and out of camps, explore these aspects. SB-2017 is across Turkish areas. The data a data set that covers several de- show that Şanlıurfa accommo- mographic and economic characteristics of Syrian and Turkish dates the highest share of Syrian individuals with a special focus refugees, followed by İstanbul on exploring opinions about so- and Hatay. The data also show cial cohesion and the integration that over 50% of Syrian refu- of Syrian refugees.48 gees who reside outside refugee Descriptive statistics from the camps settled in Istanbul, Hatay, SB-2017 show that 34% of the Şanlıurfa, Gaziantep, and Adana.

Table 1.8: Geographical distribution of Syrian refugees in Turkey

Outside Camps camps Total Sanlıurfa 13% 12% 25% Hatay 4% 15% 19% İstanbul 4% 16% 21% Gaziantep 3% 12% 14% Adana 7% 8% 14% Mardin 3% 3% 7% Kilis 5% 5% Mersin 6% 6% İzmir 5% 5% Bursa 4% 4%

Total No. Syrian refugees 1035

Source of data: SB-2017

In terms of age composition, pared with 37% outside camps. Figure 1.7 shows that about 50% The data also suggest that more of Syrian refugees in camps are young and middle-age Syrians younger than 18 years old com- settled outside camps.

48 SB-2017 collected data on 2089 Turk- ish citizens from 26 provinces and in 1235 Syrian families (7591 individuals) settled in 11 provinces (348 families settled in refugee camps and 887 families in host communities). 22 | Who are the Arab refugees?

Figure 1.7: Age composition of Syrian refugees in Turkey

%50 %45 %40 %35 %30 %25 %20 %15 Inside Camps %10 Outside Camps %5 %0

Student Retired Employed House wife In labor force Unemployed Out of labor force Source of data: SB-2017

Similar to Jordan and Leba- For those having access to educa- non, most Syrian refugees in Tur- tion, many are enrolled in tem- key, above 5 years old, are less porary education centers (TECs). educated; about 10% finished TECs often suffer challenges of tertiary education (see Table 1.9). medium including education, Of a million Syrian children of quality of education, and chil- school age, 40% remained out of dren drop out mainly after the school in the 2017-2018 semesters. first and second grade (see Erdo- gan 2017).49

Table 1.9: Educational attainment of Syrian refugees

Inside Outside Total camps camps Illiterate 14.9% 19.7% 18.5% Literate without diploma 17.2% 10.1% 11.8% Primary school 29.6% 27.5% 28% Secondary school 18.8% 20% 19.7% High school 11% 10.9% 10.9% Vocational college 3.7% 3.2% 3.3% Bachelor degree 4.7% 7.9% 7.1% Master’s/PhD 0.1% 0.7% 0.6% Source of data: SB-2017

49 Erdo an, M. Murat (2017), Syrians Barometer 2017. A Framework for Achieving Social Cohesion with Syrians in Turkey. Executive Summary, Hacettepe University Migration and Politics Research Center. https://mmuraterdogan.files.wordpress. com/2016/06/syrians-barometer- executive-summary.pdf Chapter 1 | 23

In terms of attachment to Table 1.10 also shows that the the Turkish labor market, 58% labor force participation (unem- of Syrian refugees, older than ployment) rate is higher (lower) 11 years, have joined the labor among those residing outside the force. Still, many of them, (34%), refugee camps. are unemployed (see Table 1.10).

Table 1.10: Attachment of Syrian refugees to the Turkish labor market

Inside Outside Total Camps Camps In labor force Employed 24.5% 43% 38.6% Unemployed 28.9% 18.1% 19.7% Out of labor force House wife 23.7% 24.8% 24.5% Student 21.3% 8.8% 11.7% Retired 0.4% 0.9% 0.8%

Source of data: SB-2017

Table 1.11 exhibits the dis- plying for work permits has to tribution of employed refugees, be done by employers and that in and out of camps, accord- application needs to be renewed ing to their employment status. if workers change employers. In The overwhelming majority of addition, employers can apply this cohort is waged workers in for work permits with the condi- which little over half of them tion that Syrian workers do not are irregular workers. Interest- exceed 10% of firm’s total work- ingly, regular workers are more force (see Kayaoglu, 2018). common among those settling So far, few studies have in camps. The data also show looked at the effect of Syrian ref- that most of the employer co- ugees on the Turkish labor mar- hort resides outside the camps. ket. Tumen (2016) provides evi- While Syrians have been eligible dence on differential effects such to obtain a work permit since that areas with a larger share of February 2016, the majority are Syrian refugees experience an employed in the informal sector employment loss (gain) in the in which deficiencies in Turkish informal (formal) sector. Interest- language and barriers to obtain ingly, he shows that the overall work permits are key factors. effect is negative, with no impact As of March 2018, only 40,000 on wages in both sectors (see also workers got work permits. Ap- Del Carpio and Wagner, 2015). 24 | Who are the Arab refugees?

Table 1.11: Distribution of employment status among Syrian refugees

Inside Outside Total camps camps )Daily paid job (irregular 38.2% 43.9% 43.1% Regular laborer 48.1% 39.4% 40.7% Employer 0.3% 9.1% 7.8% Seasonal worker 7.3% 6.1% 6.2% Free family worker 4.8% 0.5% 1.1%

Work in my own job (self- )employed 1% 1% 1%

Source of data: SB-2017

1.3.3.2 Social integration of Syr- them as harmless or abused on ian refugees in Turkey the labor market (see Table 1.12). Competition over economic Consistently, Turkish citizens resources, mainly employment, perceive Syrian refugees as cul- has raised tension levels in com- turally distant. An earlier version munities that host a higher share of Syrian Barometer, published of Syrian refugees. This surely af- in 2014, shows that about 80% of fects the nature and extent of in- Turkish citizens think that Syrian tegration within these communi- refugees do not resemble them. ties both in the medium and long Markedly, it also shows that the run. The SB-2017 report provides corresponding share was high- a multitude of indicators that re- er, relative to average share, in provinces near the Syrian border; flect perceptions of Turkish citi- those that are ethnically closer to zens about Syrian refugees. In to- Syrians. Misperception seems to tal, the report provides evidence contribute to lack of acceptance. of a negative perception, which is The SB-2017 shows that the over- a major barrier to smooth social whelming majority (86.2%) of integration. Turkish citizens think that Syrian While the majority of Turkish refugees ensure their sustenance citizens acknowledge that Syrian via aid from the Turkish state. On refugees are victims of war and the other hand, most Syrians who persecution, many regard them responded to the survey indicate as burdensome, dangerous, and that they earn their living via sal- beggars. On the other hand, low aries from work. shares of respondents regard Chapter 1 | 25

Table 1.12: The most appropriate statements to describe Syrian asylum-seekers (multiple answer)

Victims who escaped from war and persecution 57.80% They are people putting burden on us 43% They are dangerous people who will lead us to 39% troubles in the future Beggars/who live only with help 24.40% Our brothers in religion 20.70% Guests in our country 20.30% They are quite different and foreigners 18% Harmless people who live their own lives 14.60% People who are abused in labor force 14.30% Other 0.70% No idea 1.50% Source of data: SB-2017

In terms of the contribution In line with such attitudes, the of Syrian refugees to their host majority of Turkish citizens think economy, most Turkish citizens that Syrian refugees should not think it is little. The SB-2017 have full access to the Turkish shows that 82% think that the formal labor market. In addition, Syrian refugees do not contribute over half of Turkish citizens think to the Turkish economy. The ma- that Syrian refugees should not jority (85%) also think that Syrian get a work permit and that 23.5% refugees do not enrich society’s think that Syrian refugees can get culture. In total, 75% say that they temporary working permits for cannot live in peace with Syrians. certain jobs (see Table 1.13).

Table 1.13: What sort of arrangements should be made regarding the work of Syrians in Turkey?

Definitely should not have working permit 54,6% Temporary working permits for certain jobs 23,5% Temporary working permits for all jobs 8,1% Permanent working permits for certain jobs 6,3% Permanent working permits for all jobs 5,5% No idea/no answer 2,0% 26 | Who are the Arab refugees?

An important aspect that Another important aspect of in- measures the extent of potential tegration is place of residence. social integration in Turkey is cit- A high share of Turkish citi- izens’ opinion on the legal status zen thinks that Syrian refugees of Syrian refugees. The findings should live in safe zones or in from the SB-2017 show that the camps, although only a minor- majority (76%) are against grant- ity wants them deported. Those ing Syrian refugees full citizen- who believe that Syrian refugees ship. Consistently, the majority should live where they want or (86.5%) believe that no political should be evenly distributed rights should be given to them. across Turkey are also a minority (see Table 1.14).

Table 1.14: Where Syrians should live?

They should live in the safe zones in Syria 37.40% Only in the camps 28.10% They should be all deported 11.50% Wherever they want to live in 7.90% They should be distributed evenly in Turkey 7.70% A Syrians exclusive city should be founded 4.80% No idea/no answer 2.60%

1.3.4 Syrian refugees in Europe In terms of a source of data, the discussion heavily relies on According to the UNHCR, Statistisches Bundesamt (2016).50 over a million Syrian refugees ar- This comes with a caveat: the rived to Europe in 2015. Unlike data do not distinguish Syrians in Jordan, Lebanon, or Turkey, by refugee status. Still, since the most Syrian refugees in Europe data provide statistics of Syrians are well educated and more eco- who sought asylum in 2015, it nomically able. As mentioned largely captures those who fled above, this section of refugees due to the conflict. Evidently, the could afford to travel long dis- number of this cohort increased tances and reside in Europe. Dis- substantially after 2012, rising cussion in this section is limited from 50,000 to over 350,000. to discussing demographic and socio-economic characteristics in The majority of Syrians in Ger- Germany. Representative data many are men (61.6%). In terms of on Syrian refugees in other Eu- age composition, Figure 1.8 shows ropean countries are not read- 50 Statistisches Bundesamt (2016). Aus- ily available. We believe that the ländische Bevölkerung - Ergebnisse findings from this section give a des Ausländerzentralregisters (Bev- good picture of the characteris- ölkerung und Erwerbstätigkeit, Fach- serie 1 Reihe 2). Wiesbaden: Statist- tics of Syrian refugees in Europe. isches Bundesamt. Chapter 1 | 27

that around 60% are younger taken in the final year of second- than 30 years old. The mean age ary school in the German educa- is 26.8 years relative to 44.7 for tional system) is higher than that the German population. Figure of German population. Interest- 1.8 displays the education profile ingly, 36% of the Syrians are ei- of Syrians relative to Germans. ther enrolled in education institu- The share of Syrians with no for- tions or not attending school (not mal education levels at 19.2%, at school age). The corresponding compared with 3% for Germans. share of the Germans is 15%: dif- Still, the share of Syrian migrants ferences in age profile are likely with Abitur (a set of examinations to be the key explaining factor.

Figure 1.8: Age composition of Syrians in Germany, 2015

%35

%30

%25

%20

%15

%10

%0 0-15 15-20 20-25 25-30 30-35 35-40 40-45 45-50 50-55

Source of data: Statistisches Bundesamt, 2016b. Note: data on higher age groups are not representative

Figure 1.9: Educational attainment of Syrians and the German population

%40 %35 %30 %25 %20 %15 %10 %5 %0 still receiving No formal Hauptschule Realschule Abitur education/not education eligible

German population Syrian migrants

Source of data: Statistisches Bundesamt, 2016b. 28 | Who are the Arab refugees?

While Syrians in Germany are lum procedures via amending relatively young and more edu- the country’s Asylum act and cated, most of them are economi- Residence Act. The amendments cally inactive. Only 16.9% are now accelerate the asylum pro- employed, 7% are unemployed, cedures, lower the financial cost while the rest are out of labor for states and municipalities, and force. Half of the latter group improve the safety of refugees.52 are either too young to work While Syrian refugees generally (younger than 15 years old) or acknowledge the effort that the currently enrolled in education. German government is investing In addition, Syrians who are go- to assimilate them into the coun- ing through asylum procedures try, they show resentment over have not acquired the right to the perception that the German work documents. Another pre- still perceive them as totally new cluding factor is language: Syr- individuals without considering ians, like other nationalities, will the pre-existing skills that they have to spend considerable time brought to the society.53 learning the German language to ease their access to labor market. These two factors explain why only a small group has so 1.4 Conclusions far joined the labor market. Of the total Syrians, 55% live off un- In 2011, a series of pro-de- employment benefits or govern- mocracy uprisings erupted in ment support. several Arab countries. The outcome for Libya and Syria To enhance the integration turned deadly. Other countries of Syrian refugees, the German also suffered the consequenc- government has put in place a es of military conflict, mainly number of measures. For exam- Iraq and Yemen. As a result, ple, all Syrian children with asy- millions become internally dis- lum status are eligible for educa- placed and refugees. The scale tion. Also, Syrian refugees will for Syria is incomparably large. be granted a two year resident This chapter has explored the permit if they find a permanent employment position.51 52 See Amaral E,. Woldetsadik M,. and Armenta G. (2018). Georgetown The German government in- Journal of International Affairs, troduced a number of laws that https://www.georgetownjournalo- set the stage for easing the asy- finternationalaffairs.org/online-edi- tion/2018/4/20/europes-great-challenge-integrating-syrian-refugees. 51 See Hindy (2018) ‘Germany’s Syr- 53 See Hindy (2018) ‘Germany’s Syr- ian Refugee Integration Experi- ian Refugee Integration Experi- ment’. The Century Foundation; ment’. The Century Foundation; https://tcf.org/content/report/ger- https://tcf.org/content/report/ger- manys-syrian-refugee-integration- manys-syrian-refugee-integration- experiment/?session=1&session=1 experiment/?session=1&session=1 Chapter 1 | 29

salient demographic and so- mographic and socio-econom- cio-economic characteristics of ic characteristics to Germany; refugees from conflicts in these the largest recipient of Syrian countries, focusing mainly on refugees in Europe. the Syrian refugees and largely The labor market perfor- limiting the analysis to coun- mance of Syrian refugees in tries that disproportionately Germany is not promising with accommodate them, including a low LFPR and a high unem- Turkey, Jordan, Lebanon, and ployment rate. With this, over Germany. half of the Syrian refugees live The analysis shows that the off unemployment benefits or characteristics of the Syrian government support. refugees vary by host coun- To the extent that the avail- tries. Those settled in Jordan, able data allow, this chapter Lebanon, and Turkey have also explored the nature of so- mostly come from rural and cial integration of Syrian refu- poorer areas in Syria. Over half gees in the neighboring coun- of the refugees are children tries. The analysis shows that and most of the adults are less social integration of the Syrian educated. Since the eruption refugees, mainly in Turkey is a of the conflict, the economic farfetched objective. Negative wellbeing of Syrian refugees in sentiments, have accumulated these countries has deteriorat- over the years as the influx of ed. Poverty incidence exceeds Syrian refugees accumulated half of Syrian refugee house- with natives’ belief that refu- holds and the unemployment gees crowd them out in the la- rate has been soaring. Those bor market and compete with who are lucky enough to find them over the limited public jobs mostly work in informal resources and housing. or irregular jobs and are paid The findings and insights substantially lower wages. documented in this chapter With such harsh living condi- help to understand the eco- tions, most are highly depen- nomic and social challenges dent on direct support, cash or faced by the Syrian refugees in in kind, for basic need. the main host countries and as- According to the UNHCR, sociated concerns and fears of over a million Syrian refugees the natives. These challenges arrived in Europe in 2015. Un- are key determinants to unlike in Jordan, Lebanon, or derstand the tendency for the Turkey, most Syrian refugees Syrian refugees to repatriate, in Europe are better educated a matter that is explored in and more economically able. Chapter 2. Due to data availability, we limit our discussion on the de-

CHAPTER 2

Enabling conditions, challenges, and processes

2.1 Introduction scale of the refugee crises and the response of the international Repatriation is one of the du- community, integration and re- rable solutions for refugees en- settlement may not be feasible dorsed by the United Nations options. Therefore, even if talk- High Commissioner on Refugees ing about repatriation may seem (UNHCR) on condition that it is premature, it is important to condone voluntarily and without co- sider it as a reality and to ensure ercion (UNHCR, 1995). In prac- that it takes place sustainably. tice, it is difficult to define what One of the purposes of this constitutes voluntary repatria- chapter is to describe the factors tion. For example, it is illegal for that enter the decision-making host countries to drive refugees processes of refugees when they physically to the borders of their choose whether or not to re- country (the non-refoulement prin- patriate. These contain several ciple) (UNHCR, 1995). But they dimensions including security can still apply a degree of eco- concerns, political conditions, nomic and political pressures that economic drivers, and social con- can mean that repatriation is the siderations. The most important only viable option at a time when issues in each dimension will be conditions at home are still unsafe unpacked, describing whether (Gerver, 2018). This will not only they will enable or inhibit repa- put the refugees at risk of vio- triation. A special focus will be on lence but will also create a cycle of households headed by women. displacement if the conditions of As with any process that in- the home country cannot sustain volves decision-making, infor- return (Harlid et al, 2015) mation-gathering is an important Some of the literature claims dimension to repatriation. Since that, especially in the Global most official sources are deemed South as a result of political con- untrustworthy by refugees, in- text and economic burden, repa- formal networks form within ref- triation is the inevitable end to ugee communities with specific most refugee crises, regardless patterns of information diffu- of the situation back home (Stein sion. The most important litera- and Cuny, 1994). Given the large ture findings about information- 31 32 | Enabling conditions, challenges, and processes

gathering and informal networks 2.2 Early repatriates will also be discussed. Once the process of repatria- This section will summarize tion starts taking place, it is im- the reasons given by Arab refu- portant to discuss what condi- gees for their early spontaneous tions allows it to be physically, return to their unsafe homes. economically, and politically sus- Starting with Yemeni refugees in tainable. The conditions of sus- Djibouti, according to news re- tainability will be studied from ports by (Adow, 2017) and (Sav- two angles; the first is the welfare age and Kalfood, 2016), some of the refugees themselves; and refugees (of which many are the second is the welfare of the originally Somalis who sought home country and the communi- refuge in Yemen) decide to go ties left behind. back to Yemen despite the con- The lessons learned through tinuing conflict and threat of this extensive literature review famine and diseases. The reports will be applied to the case of Arab highlight the bad conditions in refugees. This will take the form the refugee camp, including the of a review of various reports and heat, limited food rations, and papers that study Arab refugees’ lack of basic services as reasons motivations and obstacles to re- for return. In addition, around turn. This section will conclude 19,566 Yemeni refugees returned with implications regarding the from Saudi Arabia in 2019. The most important concerns for ref- reasons for their return, how- ugees and policies that will need ever, were not monitored (Flow to be in place to ensure the best Monitoring Dashboard, 2019). outcomes for refugees. Table 2.1: Number of returnees and reasons for return

Country of Country of Number of returnees Most cited reason for return origin asylum Bad conditions in refugee Somalia - Yemen54 camps Saudi Arabia 19,566 (in 2019) - 45,840 (between 2012 Iraq55 Syria Deteriorating security in Syria and 2013) Turkey 62,439 (as of June 2019) Reunion with family members Lebanon 40,230 (as of June 2019) - Syria56 Iraq 33,703 (as of June 2019) - Jordan 39,082 (as of June 2019) Reunion with family members57 Egypt 975 (as of June 2019) -

54 Adow (2017) and Flow Monitoring Dashboard (2019) 56 UNHCR (2017a) 55 Harlid et al (2015) 57 UNHCR (2018c) Chapter 2 | 33

The situation of Iraqi refugees rating security conditions in the and repatriates is more complex. host country (Harlid et al, 2015; Prior to the Syrian civil war, the Internal Displacement Monitor, return of Iraqi refugees was irreg- 2018). In interviews conducted ular and ill-defined, determined by Internal Displacement Moni- on the one hand by the episodic tor (2018), family reunifications nature of the civil war in Iraq and and deteriorating socio-economic on the other by the individual conditions in exile were also some economic and social situation of of the main reasons that refugees households (Harlid et al, 2015). cited for returning. But many of In general, return was limited the returning Iraqis ended up be- and of cyclical nature, as refugees ing internally displaced because would go back and forth between of the segregation of cities and Iraq and the country of asylum. neighborhood across ethnic lines, But with the onset of the Syrian as well as loss of property and de- civil war in 2011, the largest num- teriorating economic conditions ber of Iraqi refugees returned (Internal Displacement Monitor, from Syria because of the deterio- 2018).

Figure 2.1: Syrian returnees by year

Syrian returnees by year58 60.000

50.000

40.000

30.000

20.000

10.000

0 2016 2017 2018 2019

The most extensively avail- Iraq (19%). A small number also able data on repatriation come returned from Egypt (UNHCR, from the small but increasing 2017a). The profile of the return- number of returning Syrians ees is mostly female from Jordan from neighboring host countries. and Lebanon, and mostly male The returnees mostly hailed from from Turkey, Egypt, and Iraq. Turkey (35%), followed by Leba- The majority of returnees are non (23%), Jordan (22%), and small families and single people. Two surveys conducted by the 58 https://data2.unhcr.org/en/situa- UNHCR (2018b) of returning Syr- tions/syria_durable_solutions 34 | Enabling conditions, challenges, and processes

ians from Turkey and Jordan on orating conditions of the country the reasons for return yield the of asylum, and improved secu- following results; for the vast ma- rity situation back home. Most of jority of the repatriates from Tur- those who returned from Turkey key, the return is perceived to be went back to Aleppo while those permanent59 (98%). The reasons who returned from Jordan went for returning are joining family to Daraa. members, followed by the deteri

Table 2.2: Returnees’ reasons for going back to Syria59

Category Reason for return Jordan Turkey (2016-2018) (2016-2017) Family Reunite with family 70.8% 77.0% Attend family event 1.3% - Get married 1.3% - Conditions in Work opportunities - 3.0% country of origin Improved security 6.0% 10.0% Conditions in Lack of income earning opportunities 4.4% country of asylum Limited access to assistance 4.3% 10.0% Living costs outside camp 3.6% Limited freedom of movement 1.6% Other Medical treatment completed 1.1% - Costs of medical treatment 0.8%

The findings of this UNHCR finds that while the majority of survey is in line with the litera- Syrians want to return someday ture, which will be surveyed in (76%), only 4% plan on doing so the next section, with social con- in the next 12 months. Again, the siderations (family reunions) cit- main reason for wanting to return ed as the most important reason within the next 12 months is re- for return. But that does not mean unification with family members, that the role of other variables is with 90% of them planning on re- irrelevant. This is highlighted by turning to the place of origin. But another UNHCR survey (2018a) those who want to return are also on intentions to return among more likely to have intact prop- Syrian refugees also in neigh- erty in Syria and are more likely boring host countries. The study to be ‘severely or highly vulnerable’ (as per socio-economic indi- 59 Permanent versus temporary return cators) compared with the refu- is not captured in the Jordanian sur- gee population. More than half of vey. them also report an information Chapter 2 | 35

deficit, especially regarding the Table 2.3: Intention to return survey security situation and the availability of shelter, services, and Intention Do you hope Do you plan livelihood opportunities. They Survey to return one to go back in rely primarily on information day? the next 12 from their family members and, months? to a lesser extent, media, commu- Yes 76% 4% nity members, friends and social media. The story of disconnected, No 19% 85% poverty-ridden Syrian refugee households returning early is I don’t know 5% 11% collaborated by Kvittingen et al (2018) in their research in Jordan.

Figure 2.2: Demographics of those intending to return versus respondent refugee population

Demographics of those intending to return versus respondent refugee population60 100% 90% 80% Those intending to 70% return within the 60% next 12 months 50% 40% Respondents 30% population 20% 10% 0% Intact property Highly or severly vulnerable Have family in Syria

Another outlook is provided conflict and more likely to return by a complex World Bank (2019) to areas that have experienced study that uses econometric change in control. The last factor analysis to determine the most has a large magnitude and re- important factors for return; de- flects the importance of political mographic variables that are considerations alongside security positively correlated with return concerns. The security dimension include being single, male, and also has parallels with family less-educated. Analysis of pull ties consideration; for example, factors shows that the most im- returnees to high-intensity con- portant determinant of return is flict areas are more likely to be security: refugees are less likely non-immediate family members to return to areas that are under of those left behind. Services are an important consideration, as 60 UNHCR (2018a) lack of provision of education 36 | Enabling conditions, challenges, and processes

and health facilities in the home well as dwindling aid in Jordan country is an effective deterrent as the reason behind some re- to return. As for push factors, turnees’ choices. the results are more complex; for The UNHCR (2018b, 2018b), example, having food security SAWA (2019), and Yahya et al in the host country can increase (2018) reports highlight how pre- the likelihood of returning but so mature the return at this stage is. does living in a refugee camp. Many of those who have returned Smaller scale surveys con- from Lebanon to Syria became ducted by various researchers displaced again after realizing highlight the role of push factors that staying in Syria is not fea- undertaken by countries of asy- sible. Those who returned from lum in premature return, espe- Jordan to Daraa also became re- cially in Lebanon. These include displaced as the conflict reignited a report by SAWA (2019), which (SAWA, 2019; Yahya et al, 2018). invokes the hostile environment In addition, those who sought to in Lebanon towards refugees and go back within the next 12 months the unsustainable living condi- have had insufficient evidence to tions as one of the main reasons make a well-informed decision for premature return. Another especially regarding the security report by Carnegie Group (Ya- situation back home (UNHCR, hya et al, 2018) mentions ‘humili- 2018a). ation’ of refugees in Lebanon as

Demographic Country of origin Country of asylum

• Single, less-educated • Improved security • Poverty (ambigious) male situation • Hostile envionment • Small families • Presence of family • Highly vulnerable members • Presense of intact property

What do these surveys tell us for pull factors from the home about return? There is consensus country, two stand out. The first on which demographic and pull is the social dimension and the factors contribute to early repa- presence of family, and the sec- triation. Demographic variables ond is the (perceived) improved include being a less-educated security in the area of origin. Eco- adult male, having a small fam- nomic pull variables play a role ily, or being an extended family in return, albeit small, mostly member of those left behind. As taking the form of the presence Chapter 2 | 37

of intact property in Syria. It is opposed to individual decision- also worth noting that those who making. Throughout the rest of want to return report a deficit this chapter, however, the situ- of information on the situation ations of Arab refugees and re- back home, which can imply an patriates will be compared with exclusion from informal social historic refugee crises where ap- networks. The most unclear pat- propriate to highlight similarities tern emerges when studying and differences. push factors from host countries; In general, the findings of Syr- some indications show that early ian repatriates’ surveys are in line returnees are overrepresented with the literature conclusions. among the economically vulner- The demographic features of ear- able while others find mixed rely returnees are similar to those sults or the complete opposite. observed elsewhere. Similarly, In order to understand the the most cited pull factors for re- specificities of the Arab case, it turn are common among any re- is useful to contrast it with other turning refugee population. The historic repatriations. Comparing mixed picture when it comes to the responses of Syrian refugees push factors is also not uncom- on the intentions to return with mon in refugee surveys. This will Afghani refugees in Pakistan in be highlighted in the relationship 2011 shows a similar picture. At between poverty and repatria- that time, 84% of Afghani refu- tion in section 2.3.3. Therefore, it gees expressed no intention of re- could very well be the case that turning home (Harlid et al, 2015), early returnees to Syria include a similar to the 85% of Syrian refu- combination of the economically gees who had no plan to return vulnerable and the economically home within 12 months. The ex- well off. It is also not unreason- perience of Afghani refugees also able to conclude from the sur- highlights that the main driver veyed literature that pull factors of repatriation is pull factors, not play a much more important role push factors; despite the dete- in determining return than push riorating economic conditions in factors. countries of asylum, the number of Afghani repatriates decreased because of the worsening security situation in Afghanistan and 2.3 Conditions for return only picked up when the security situation at home improved The general assumption re- (Harlid et al, 2015). Other exam- garding voluntary repatriation ples in the literature tend to be is that once the security threat at geared towards organized return home is reversed, refugees will as oppose to spontaneous return, go back. The idea is that refugees’ which is more concerned with attachment to their home coun- the macro-security situation as try will inevitably drive them to 38 | Enabling conditions, challenges, and processes

return once the conflict is over, der dimension, whereby men and ending the so-called refugee women have different priorities cycle. This approach, however, and act differently. Therefore, the has proven to be too simplistic to experience of households headed describe reality. While refugees’ by women will be discussed sepa- wellbeing is highly dependent rately at the end of each subsection. on supra-individual factors at home and in exile, they remain 2.3.1 The paradox of security con- independent agents with com- cerns plex decision-making processes that accounts for a myriad of fac- Security is the most cited tors (Harlid et al, 2015). While se- concern for refugees against re- curity is undoubtedly one of the turning to their home country. most important considerations, Despite that, ‘spontaneous’ repa- refugees also take account of the triation does takes place before political, economic, and social the official end of a conflict (Stein factors in the host country and and Cuny, 1994; Zimmermann, home country. 2012). This section will attempt to unpack the relationship between As a result, a new outlook safety and repatriation, and elu- to the repatriation decision has cidate how security is reconciled emerged. Refugees are seen as with return when it happens be- purposive actors with at least fore the end of the conflict. These some degree of independence include return to safe areas with- to make their own choices. They in the home country, deteriorat- compare conditions in exile and ing security situations in the host at home, and make up their country, and the difference be- minds accordingly (Harlid et al, tween personal safety and over- 2015; Stein and Cuny, 1994; Zim- arching security situation. mermann, 2012). This sometimes leads to early return, before the Some refugees return to ‘safe’ conditions at home are declared areas in their home countries be- safe (so-called spontaneous re- fore the official end of a conflict. turn) or, alternatively, it can This could be the place they orig- mean that refugees never return, inally repatriated from or a new even when the situation at home location (Harlid et al, 2015; Zim- is deemed safe (Black and Koser, mermann, 2012). Once an area in 1999; Harlid et al, 2015; Stein and the home country is deemed safer Cuny, 1994). The purpose of this and outright threats of violence section is therefore to describe the are curbed, some refugees opt to security, political, economic, and return there over staying in their social variables that have been host countries. The main issue found to have the most influence with this type of repatriation is on the repatriation decision. As that it may not be sustainable; the with any decision-making pro- conflict may erupt again in safer cess, repatriation also has a gen- areas, creating another episode Chapter 2 | 39

of displacement. In addition, the presence in the host country con- continuing conflict means that stitutes a risk to their safety, they the reconstruction process in might opt to take their chances these ‘safe’ areas is delayed, cre- in their home countries. The creating dismal economic conditions ation of an unsafe environment in for returnees (Harlid et al, 2015). host countries can be ignited by Deteriorating security con- the presence of refugees, or it can ditions in host countries may be a calculated move to encour- prompt refugees to return, pre- age the return of refugees (Crisp, ferring to take their chances back 2003). Alternatively, in volatile home even if the conflict is still regions, worsening security con- continuing. As mentioned, the ditions in the host countries inde- decision-making process of refu- pendent of the presence of refugees involves a comparison be- gees can push those displaced tween the situation in the home to repatriate, even if the conflict country and that in the host coun- at home is not yet over (Lischer, try (Zimmermann, 2012). If their 2017).

Table 2.4: Security threat and its consequences on the refugees

Country of Country of Security threat Year Consequence origin asylum Afghanistan61 Pakistan Threats and police 2016 365,000 Afghan abuse in refugee camps. refugees returned. South Sudan62 Ethiopia Security incidents in 2018 20,000 South the Ethiopian refugee Sudanese refugees camp. returned home. Uganda63 Sudan Sudanese civil war 1983- 2005 Both Ugandan refugees and Sudanese fled to Uganda. Iraq64 Syria Syrian civil war 2011 on- Biggest return of going Iraqi refugees from Syria back to Iraq. Somalia65 Yemen Yemeni civil war 2015 on- 4,300 Somali going refugees returned from Yemen Rwanda66 Zaire (DRC) First Congo War 1996-1997 500,000 Rwandan refugees returned home within a four-day period.

61 Human Rights Watch (2016) 64 Harlid et al (2015) 62 Reliefweb (2019) 65 UNHCR (2019) 63 Harlid et al (2015) 66 Refword (1997) 40 | Enabling conditions, challenges, and processes

There is also a need to distin- 2.3.2 How can politics influence guish between personal safety return? threatened by individual perse- Refugees’ political ties as cution and the overarching secu- well as their view of themselves rity situation. In general, security as agents with political influence and safety are discussed in the can have implications for the de- form of an end to fighting or the cision to return. The most impor- end of violence and movement tant mechanism through which towards post-conflict peace- political factors influence refu- building and reconstruction. But gees’ return is through providing in some cases, security concerns the overarching context in which are more personal, as they per- they must react. It also ties in tain to individual fear of persecu- closely with personal security and tion from one side of the conflict expected future economic con- (Hardgrove, 2009; Omata, 2013). ditions. This section will briefly While those who are concerned unpack some of the implications with the general security envi- of the political dimension, includ- ronment may return before the ing political settlement, political official end of the conflict, those agency, minority return, and the worried about personal safety are overarching political context. more likely to condition return In the case of civil conflicts, on a political resolution rather political settlement is an impor- than just an end to the violence tant consideration for refugees, (Omata, 2013). especially as it ties in with per- As with all factors considered sonal safety. Fear of returning in the repatriation decision, there to a home that is still under the is a gender dimension to the secu- control of a dictatorship is espe- rity concern. A study by Black et cially amplified for political refu- al (2004) finds that men more than gees and for young men who re- women cite overarching security fused to take part in mandatory fears as a deterrent to returning. army conscription (Muasher and As for individual security, the Yahya, 2019; Muggeridge and two genders voice different con- Doná, 2006). In addition, political cerns: for men, army conscription considerations may be masked and political persecution are the by what appear to be economic main source of worry. Women, concerns; many refugees condi- however, tend to worry more tion return on ensuring that they specifically about gender-based retain a way of going back to the violence if they decide to return, host country (through owning and they also tend to perceive the property, retaining a host country journey back home as more un- passport, etc.) in case the political safe than their male counterparts situation deteriorates back home (Harlid et al, 2015). (Muggeridge and Doná, 2006). Chapter 2 | 41

As for political agency, refu- geted during (or even before) the gees may view returning home conflict (Graham and Khosravi, as a political act equivalent to 1997). Without effective domestic recognizing the legitimacy of and international policies, eth- the regime in the home country. nic minority refugees are more This is justifiable as no state can likely never to return for fear of achieve validity with a significant safety and deteriorating ties with portion of its population living the homeland, even in condi- outside its borders (Petrin, 2002). tions of political stability (Harlid When refugees decide to return et al, 2015). If they do return in a home, they are effectively show- situation where ethnic division is ing confidence in the ability of geographically maintained post- the state to protect them. In the repatriation, minority returnees absence of such faith, refugees will become internally displaced may choose to remain in exile to as they may be forced away from deprive the state of political le- their original locations (Petrin, gitimacy, especially in the eyes 2002). of the international community The Table 2.5 below gives (Graham and Khosravi, 1997). examples of minority/majority Repatriates are also seen as an refugees reacting to minority/ important tool for rebuilding the majority governments in their state after the end of the conflict country of origin. While these ex- through various forms of human amples are useful in extrapolat- and physical capital that can be ing realities that can be applied mobilized in the reconstruction to the Arab cases at hand, it is process (Petrin, 2002). But if refu- important to keep in mind that gees do not feel connected to the the refugees’ reactions are highly state or feel threatened by it – as context-specific. Conditions in is likely to be the case if it is the the country of asylum matters, reason they fled in the first place so does the nature of identity – then they will have no interest and the ties to the homeland. For in returning or contributing to example, Shi’a Iraqi refugees in the rebuilding process in any of Jordan (a majority in a minority its forms. government during the Saddam In the case of an ethnically Hussein era) expressed a desire charged conflict, the nature of the to return home because they did population’s relationship with not feel welcome in Jordan (Chat- the state may irreversibly change. elard, 2003). In contrast, Assyrian For minorities specifically, it may Iraqis in Britain felt more wel- compel them to identify with comed than their Arab/Muslim their religious or ethnic group counterparts and, therefore, ex- rather than with their nationality, pressed less desire to go back (Al particularly if they have been tar- Rasheed, 1994). 42 | Enabling conditions, challenges, and processes

Table 2.5: Examples of minority/majority refugees reacting to minority/ majority governments in their country of origin

Government Refugee/ Government Typology Minority Majority

Context-specific: refugees Likely to raise questions of are more likely to express identity and belonging for desire or plans to return minority refugees but also sometimes conditioning it highly dependent on the on the minority staying in relationship to the dominant power. group and to the homeland Armenians Iranian refugees Minority Sunni Iraqi refugees in Sweden expressed doubts during the 1990s in Jordan, about the idea of Iran expected to repatriate remaining their homeland. but also had better This in contrast with the living conditions in exile Baha’is who, despite severe compared with their Shi’a persecution, remained tied counterparts.67 to the idea of Iran as home.68 Refugees Likely to express hesitation Refugees’ considerations and mistrust regarding are less likely to be about repatriation, even in the identity and more likely case of support from the to be about security and international community political agency. and UNHCR. Hutu Rwandan refugees in Majority Zaire expressed suspicion Interviews with Chilean and reluctance to return refugees show they home even after it was conditioned return on declared safe. Less than regime change and end to 10% of the Hutu Rwandan dictatorship because it is the refugees returned home only way they would feel prior to the conflict in safe.70 Zaire.69

It is important to mention community is willing to carry in that the political dimension of terms of resettling refugees and repatriation features much more providing donations and fund- as an overarching context than ing to host countries to integrate a ‘variable’ that affects the deci- them will determine how soon sion to repatriate. Specifically, and how sustainable the refu- the weight that the international gees’ return will be. While the international community may

67 Chaterlard (2003) 69 Refworld (1997) 68 Graham & Khosravi (1997) 70 Muggeridge & Doná (2006) Chapter 2 | 43

not outright condone premature salient economic factors, namely return, they may exercise a level poverty and deprivation; access of apathy that leaves the refugees to services and work opportuni- with no other real option but to ties; as well as housing and prop- risk going back to their home erty rights. country. This is exacerbated by The relationship between the treatment of repatriation as monetary poverty and repatria- an end in itself as opposed to tion is complex and can go both a means of protecting refugees ways. Poorer households both (Petrin, 2002). Finally, while the repatriate the earliest and the lat- rhetoric of home countries’ gov- est (Harlid et al, 2015). Early re- ernments generally favors returnees from poorer households patriation as soon as possible, it may choose to take their chances might be in its interest to delay it back home over impoverishment or stop it altogether. This could in their host country or may find either be due to its economic in- staying in exile unfeasible any- capability of absorbing returnees more. In other cases, however, or a political desire to preserve poverty acts as constraint on re- the status quo. turning since it involves potential exposure to new sets of vulner- 2.3.3 Economic drivers of repa- abilities that the household may triation not be equipped to handle. The perception of what type of live- More than any other factor, lihood access refugees will have the focus on economic drivers on returning is key in defining of repatriation views refugees as the relationship between repa- ‘purposive actors’ with the abil- triation and poverty. Therefore, ity to make decisions about their it is unclear to what extent depri- future independently and ratio- vation does indeed drive return, nally (Harlid et al, 2015). This is despite it being perceived as a in contrast with the more pas- successful tool that host countries sive view that understands repa- use to push for repatriation (Har- triation as the natural ‘next step’ lid et al, 2015). Some of the cop- once the original causes of dis- ing mechanisms that poor refu- placement have been removed. gees and repatriates use include Consideration of economic fac- inhabiting urban areas in their tors usually takes the form of a home country on returning (even comparison between conditions if they mostly hail from rural ar- in the host and home country eas) or undertaking a gradual or on which refugees subsequently partial process of return to mini- base their decision to return. This mize the risk and establish a kin- section will discuss some of those ship network (Harlid et al, 2015). 44 | Enabling conditions, challenges, and processes

Table 2.6: Impact of poverty on repatriation

Poverty pushed for repatriation71 Poverty hindered repatriation72

Country of origin Country of asylum Country of origin Country of asylum

Iraq Syria Liberia Ghana

Beyond poverty, access to Another important aspect services as well as employment considered in the repatriation de- opportunities are two important cision is the ability to access hous- determinants that are considered ing and land on returning. Many in the decision-making process refugees may have lost access to about returning. In many cases, their houses and lands during host countries provide some the conflict either because of de- services to refugees, most nota- struction, looting, or confiscation. bly education, access to health In that case, if there is no other facilities, and food aid (Harlid means of accessing shelter (for ex- et al, 2015). Some training ser- ample, staying with kin) refugees vices may also be provided by have no choice but to stay in host the UNHCR and other INGOs countries. For refugees from rural (although curbing provision areas, access to land encompasses of services is one way through both means for livelihood and for which some host countries and shelter, making return even more the UNHCR ‘encourage’ repa- challenging (Harlid et al, 2015). triation – see Collins (1996). But Having no option but to repatri- host countries tend to restrict ate, rural refugees may choose refugees from getting work per- to go back to urban areas in their mits, making employment op- home country rendering them portunities scarce and access internally displaced. Seizure of to the labor market limited to property in the homeland may the informal sector. This can be also push for premature repa- particularly frustrating for refu- triation for refugees who wish to gees who worked in professional regain possession of their land or fields prior to exile. On the other house. In some cases, this creates hand, home countries that are cyclical return, with refugees go- still undergoing conflict or that ing back to assess if they have ac- have just been declared safe are cess to their property. According unlikely to have the infrastruc- to Harlid et al (2015), attempts by ture necessary for the provision the international community and of services, especially in rural ar- governments to address concerns eas. In addition, employment op- about house and land access by portunities are likely to be scarce repatriates have generally fallen if the post-conflict reconstruction short. phase is yet to be initiated.

71 Harlid et al. (2015) 72 Harlid et al. (2015) Chapter 2 | 45

The economic drivers of repa- tion as it provides a ‘Plan B’ for triation also have a gender dimen- returnees (Black et al, 2004; Da- sion. Poverty among the refugee dush, 2018; Harlid et al, 2015). population tends to be concentrated in households headed by 2.3.4 Social considerations women. This generally translates into premature repatriation in the While security is often cited case of the presence of kin in the as the most important deterrent home country or late repatriation to repatriation, social reasons are if not (Collins, 1996; Taiwa, 2016). considered one of the most im- Households headed by women portant drivers of it. Therefore, also tend to be more concerned this section will focus on the most with the provision of services, commonly cited social reasons especially schools and hospitals that inhibit or encourage return. for their children, while men are Among these are family circum- more concerned with the avail- stances in exile and in the home ability of working opportunities country; social networks and as- and access to livelihood (Collins, set; and changing gender rela- 1996; Martin, 2004; Taiwa, 2016). tionships and roles. Finally, households headed by Family considerations in exile women may find it particularly and changes in family circum- challenging to access house, land stances back home are often- and property since they have less cited reasons for return (Black et influence and power (Collins, al, 2004). The presence of family 1996; Harlid et al, 2015). back home can act as an impor- There are some economic con- tant pull for refugees, especially siderations to repatriation in the in the case of important events literature that are contested or like weddings and funerals. Fam- context-specific. The main one is ily in exile also plays a role in de- that incentives from home coun- termining return. For example, tries mobilize return more effec- refugees who have young chil- tively than disincentives from dren attending school may want host countries. Some case studies to wait until they finish their edu- support this claim, and it links in cation (Graham and Khosravi, with the argument that poverty 1997). Similarly, households with hinders repatriation. Similar to elderly or sick family members this argument is the question of may find it unfeasible to under- whether integration hinders re- take the journey of repatriation. patriation. While intuitively the In some cases, households may answer is yes since it is one of send some family member – usu- the other durable solutions, some ally young adult men - earlier case studies provide evidence than the rest to get an impression that integration may be one way of the conditions back home be- to encourage (partial) repatria- fore the rest of the family follows. 46 | Enabling conditions, challenges, and processes

Households may also choose to repatriate. This is especially the split themselves between the host case in conditions where being a country and the home country woman implies greater risk and as a back-up plan in case politi- more restricted access to liveli- cal conditions start to deteriorate hood, prompting them to seek or in order to diversify sources of the protection and assistance of income between remittances sent the men in their families (Collins, from abroad and money earned 1996; Taiwa, 2016). at home. This is a common oc- An important social consider- currence, termed by the literature ation in the exile and repatriation as ‘cyclical’ or ‘iterative’ return journey is the changing relation- (Harlid et al, 2015; Omata, 2013). ship between the genders and According to Omata (2013), shifts in gender roles. Women take the ‘extent and quality of social up responsibilities during conflict assets’ are especially important and exodus that may have been determinants of the decision to deemed taboo before leaving the repatriate. The presence of con- home country (Taiwa, 2016). But nections in the homeland can harsh exile conditions also play help to buffer some of the shocks a role in exacerbating already that will inevitably hit the return- existing inequalities between ees during the adjustment peri- the genders (Harlid et al, 2015). od. In the absence of these social While some of the responsibili- networks in the home country, ties taken up may be perceived as return will be delayed. As for ‘positive’, with women heading the presence of social networks their households and undertak- within refugee communities or ing paid employment, it can also host countries, these can either lead to dire consequences, such delay or encourage repatriation. as sex work and taking up arms As will be discussed in a later sec- and fighting. This depends on the tion, social networks are the main circumstances of the household method of information transmis- as well as the living conditions in sion between refugees in the ab- the host country. But experience sence of trust in official informa- has shown that these roles tend tion sources. As a result, those to be reversed once the conflict networks within refugee camps is over. As a result, women may and the diaspora can play an im- justifiably assume that returning portant role in informing the de- will mean being relegated back cision to return. Those who have to traditional and inferior social connections in host countries status, which most of them grew may find it easier to integrate and to reject (Graham and Khosravi, have access to jobs, delaying re- 1997). If they have undertaken turn. For women and households sex work or other roles deemed headed by women, the presence unacceptable, return might of male next-of-kin in the home mean social exclusion and even country is an essential reason for violence. This is especially true Chapter 2 | 47

if the home country experienced as they compare their situation social changes during the con- in exile with the situation back flict towards traditional views home across several economic, of gender roles. The other side political, and social factors. Since of changing gender roles is that refugees cannot go home to men, especially older ones, may gauge conditions by themselves, be more open to returning in or- their decision to repatriate is der to regain their authority and heavily dependent on the avail- breadwinner status (Graham and ability and quality of informa- Khosravi, 1997). tion. Refugees tend to be skepti- There is a need to distinguish cal of official information sourc- between women in households es, be it from their home country, headed by men and those headed host country, or even INGOs, so by women. For all the aforemen- they tend to rely on unofficial in- tioned factors – but especially formation sources, including net- the social dimension – men and works within the refugee camps, women have some diverging the home country, and early re- priorities when it comes to the turnees (Collins, 1996). decision to return. In the case of With the emergence and wide- women in male-headed house- spread use of mobile phones and holds, the decision may be in the internet, the scope and reach made in harmony, but they do of information sources, especially not have the final say. On the oth- informal ones, have expanded er hand, households headed by hugely. This means that refugees women – who often have young are exposed to various informa- children – have more autonomy tion sources, some of which are when it comes to the decision- contradictory, that they will have making process and may come to to consider and evaluate (Gil- different conclusions than male- lespie et al, 2016; Wall et al, 2017). headed households given the This section will briefly explain aforementioned diverging pri- the prevalence of informal in- orities and experiences (Collins, formation sources, the patterns 1996; Taiwa, 2016). through which information is spread, and the implication of the increased use of mobile phones and the internet. 2.4 Other considerations According to Black and Koser (1999), the literature used to at- 2.4.1 Networks and information- tribute the diverging outcomes of gathering refugees to the different informa- The previous sections high- tion available to them. But with lighted some of the most im- the emergence of the view of ref- portant factors that determine ugees as agents with independent refugees’ propensity to return, and complex decision-making 48 | Enabling conditions, challenges, and processes

processes, the focus became more formation is coming from infor- on the frequency and accuracy mal sources. In male-dominated of information flow when try- societies, households headed by ing to assess the conditions of a women can be excluded from potential destination as opposed the information loop, leading ei- to only the accessibility of infor- ther to premature repatriation in mation. Therefore, refugees will the case of lack of integration or only take the decision to return delayed return in the case of mis- home once they have evaluated information about the situation the reliability, accuracy, and con- back home. tent of the information gathered One of the most important (Black and Koser, 1999). sources of information for refu- Dissemination or restriction of gees are early returnees who are the flow of certain information in perceived as the most reliable refugee communities can also be evaluators of the situation back used as a tool by the various actors home since they have experi- to push for and encourage return enced being refugees themselves or to discourage immigration and and understand what kind of curb the flow of refugees (Carlson information is valued (Collins, et al, 2018). This creates an issue 1996). As for formal information of mistrust, making the source of sources, the most mistrusted are information a very important con- those that come from govern- sideration in determining its reli- ments (whether in home or host ability among refugee communi- countries) since they tend to be ties. In lieu of official information contradictory, inaccurate, or serv- sources (such as institutions or ing ulterior motives. Information the media), refugees rely on infor- from NGOs or INGOs (such as mal channels, such as social and the UNHCR) are perceived to be kin-based networks. In almost more reliable if too general to be all cases, informal channels are useful in most cases. These or- considered the most trusted, reli- ganizations, however, go out of able, and up-to-date information their way to ensure that house- source on the security or econom- holds headed by women are in- ic conditions in the home country cluded in the information loop (Collins, 1996). by, for example, allowing them There are some commonali- a tour of the home area (Collins, ties in the ways that information 1996). All in all, refugees do not is spread across formal and in- discard any piece of information formal channels, and the meth- even if it is from an untrusted ods through which it is used by source; they evaluate and com- refugees. According to Collins bine the information to get the (1996), female refugees may re- most holistic picture before they ceive less information than male undertake any final decisions or refugees especially when the in- actions (Black and Koser, 1999; Collins, 1996). Chapter 2 | 49

With the emergence and dis- the way that refugees are exposed semination of mobile phones and to and spread information, others the internet, social networks can find that these claims are exag- be extended through social me- gerated since the infrastructure dia to encompass both strong to make use of this technology is and weak ties, and diversify the not always available to refugees, sources of information available and when it is, it can create social to refugees and forced migrants divides and facilitate the spread (Gillespie et al, 2016). In addi- of misinformation (Gillespie et al, tion, it can ensure that the social 2016). ties with those left behind in the home country can be maintained. 2.4.2 Welfare and the sustainabil- The use of mobile phones can re- ity of return sult in both positive and negative outcomes. For example, depend- While in theory repatriation ing on the quality of the informa- is the most desirable outcome for tion disseminated and the ability all stakeholders, this is true for and knowledge to verify it, refu- the home country and the refugee gees can use social media to en- community itself only if certain sure that their decisions are well- conditions are in place. The pro- informed (Gillespie et al, 2016; cess of repatriation not only af- Wall et al, 2017). In addition, be- fects returning refugees but also ing directly and instantaneously the communities that they will be in touch with family and friend returning to. As a result, sustain- back home ensures that the infor- ability takes on two dimensions; mation they have on the situation the first is that of the returnees in the country of origin is reliable, themselves, whether conditions uncorrupted, and relevant. are in place to allow them to stay The negative outcome of mo- long-term in their home country bile phone usage among refugees and reintegrate with the com- comes from two dimensions. munities that they have left. The First, the inability to distinguish second dimension is the welfare between reliable and unreliable of the home country and of the information can have dire conse- communities to which repatriates quences that could put the refu- are returning to, whether they gees’ lives at risk and make them can reabsorb and make use of the vulnerable to exploitation. In ad- different types of capital that re- dition, it makes it easier for state turnees bring with them. and non-state actors to survey As will be discussed in the and monitor refugees’ mobility next section, return to any of the and actions through the digital areas of conflict in the Arab world footprint that they leave behind is not sustainable at this stage. (Gillespie et al, 2016). While some Different reports, especially on literature strands claim that mo- Iraq and Syria, show that even the bile phones have ‘revolutionized’ most basic dimension of repatria- 50 | Enabling conditions, challenges, and processes

tion sustainability – the physical the country of asylum premature- dimension – is not feasible, with ly. Those with more education a proportion of the returnees and with a job are also less likely becoming re-displaced once the to want to re-emigrate and have dormant conflicts reignite. This is higher income levels, providing not surprising given the continu- evidence of socio-economic inte- ing nature of the conflict. Despite gration in the home country. The that, especially in the case of Syr- same is true for those who receive ian refugees, repatriation is hap- assistance and those who have a pening at an accelerating rate. secure status abroad. The welfare of repatriates on When it comes to the condi- returning to their home country tions in the home country, factors is subject to the sustainability of such as access to public and social return. There are several dimen- services, housing, education and sions to this sustainability, the security all contribute to the sus- most basic of which is the physi- tainability of return. In addition, cal dimension; settling at home having links with the host coun- without being re-displaced after tries after returning may help in return. The other dimensions of the socio-economic dimension of return are socio-economic, which sustainability through providing refers to the livelihood and living access to diverse sources of live- conditions of returnees and, fi- lihood, such as remittances. It is nally, political dimensions which worth noting that this study was reflects security and access to ser- conducted in a post-conflict sce- vices (Black et al, 2004; Fransen, nario. If the home country is still 2017). The success of return on all subject to bouts of violence, lack these dimensions are subject to of security will be the determin- the demographic of the returning ing factor for sustainable return refugees, the situation that they or lack thereof. left in the country of asylum, and Repatriation can lead to vary- the situation they are returning ing outcomes for the home coun- to (Black et al, 2004). Measuring try and receiving communities, the sustainability of return en- again, depending on the condi- tails comparing the conditions of tions of the homeland and the re- those who returned with those turnees. In the best-case scenario, who were left behind. returnees can be an important Black and Gent (2004) and asset for post-conflict reconstruc- Black et al (2004) find some de- tion and the re-establishment of mographic characteristics that socio-political structures and in- are linked with sustainability of stitutions. This so-called returnee repatriation. For example, be- capital comes from the skills and ing young and male is correlated expertise of returnees, their eco- with a desire to re-emigrate, as nomic and political power, and does being coerced to return from their unique position in attracting Chapter 2 | 51

international development and marized to see which of the afore- humanitarian assistance (Helling, mentioned considerations are the 2007). most relevant in determining But this best-case scenario is return. While the factors consid- only feasible if some factors are ered in the previous section are present in the home country. the most universally applicable, These include laws and regula- context will inevitably lead to ditions, such as property rights and verging experiences. This section housing, political settlements, will encompass the experience of such as disarmament, but also the Arab refugees from Yemen, Iraq attitudes and actions of other ac- and Syria with a focus on Syrian tors involved, including the state refugees since they constitute the and the receiving community majority of those displaced in the (Black and Gent, 2004; Helling, region. It will start by briefly out- 2007). The last is especially rel- lining the actions of other actors evant when it comes to shifting involved, namely the UNHCR, gender roles in exile and how host countries, the international they are perceived by receiving community, and home countries, communities that are likely to be after which the experience of the more conservative (Jallow, 2004). refugees themselves will be discussed.

2.5 What does this mean 2.5.1 Stakeholders for Arab repatriation? To date, the UNHCR has not declared any of these three Arab The previous sections countries safe for return. For Ye- mapped out the most important men and Syria, it recommends factors that enter the decision- that states suspend forcible re- making processes of refugees on turn to the country (UNHCR, whether or not to return home. 2015, 2017b). As for Iraq, it places In addition, it provided an over- a caveat on suspension of forcible view of how information crucial return with ‘Iraqis who originate to the decision-making process is from areas of Iraq that are af- spread across refugee communi- fected by military action, remain ties with the absence of trust in fragile and insecure after having official information sources as been retaken from ISIS, or remain well as the question of how re- under control of ISIS’(UNHCR, turn can be sustainable for refu- 2016). Despite the straightfor- gees, receiving communities, and ward tone of these declarations, the host country. mounting burdens on neighbor- In this section, a review of ing countries and pressure from studies conducted on Arab refu- donor states have put the UN- gees and repatriates will be sum- HCR in a difficult position: on 52 | Enabling conditions, challenges, and processes

the one hand, there is a need to who will be returning to Iraq and, appease funding states that want as a result, have an unclear status a quick solution for the refu- that makes it difficult for them gee crises; and on the other, its to work (Dockery, 2017). Iraqis mandate to protect the displaced in Syria are not faring better; ac- means it cannot endorse return cording to one report, they’re (Crisp, 2017). The compromise is ‘virtually prisoners in the [refu- to plan subtly for return while at gee] camp’(Internal Displace- the same time upholding that it ment Monitor, 2018). Finally, for is unsafe at this stage. In Iraq, for Yemeni refugees, Djibouti has al- example, the UNHCR provides lowed free movement and work limited repatriation packages permits, but the country is facing for those who express a desire its own problems of poverty and to return (Harlid et al, 2015). As deprivation (Adow, 2017). for Syria, it is taking preparatory The reaction from the Euro- steps and scaling up its opera- pean Union and the United States tions inside the country ‘in antici- to refugees from all three Arab pation’ of voluntary repatriation countries has been to find and (Crisp, 2017; Marks, 2018) push for a quick solution (Crisp, As for host countries, their 2017; Harper, 2008). Funding for reactions to the presence of refu- host countries in the region is gees depend on domestic and expensive despite not meeting international pressures, which, the needs, and it was made clear, more often than not, does not especially by the United States, yield ‘a rationally evolved refu- that resettlement is not an option gee policy’ (Jacobsen, 1996). In (Crisp, 2017). While these actions the case of Syrian refugees, both do not directly endorse repatria- Turkey and Lebanon called for tion, they consciously leave no the return of refugees, with Leba- other solution even as the home non undertaking explicit policies countries remain unsafe. to discourage Syrians from staying (Muasher and Yahya, 2019). Finally, the last stakehold- Jordan, on the other hand, main- ers in the refugee crises are the tained that it would neither en- Arab home countries themselves. courage nor discourage repatria- With the war in Yemen continu- tion for Syrian refugees (Muasher ing, it is difficult to talk about a and Yahya, 2019). But there are unified policy or rhetoric about reports of Jordan forcibly return- Yemeni refugees. As for the Iraqi ing some refugees to Syria with- government, it recently started out a valid reason or a chance undertaking steps to encourage to challenge their removal (Hu- refugees explicitly to repatri- man Rights Watch, 2017). As for ate though financial incentives Iraqis in Jordan, they are treated and compensations (Harlid et al, as ‘guests’; perceived by the gov- 2015). The rhetoric of the Syrian ernment as temporary residents government is to encourage repa- Chapter 2 | 53

triation, but policies put in place mer Baathists who fear persecu- regarding property rights and tion (Internal Displacement Mon- mobility within Syria implicitly itor, 2018; Majeed, 2018). Majeed discourage return (Muasher and (2018), for example, describes the Yahya, 2019). response of an Iraqi ex-officer in the military ‘Thus, when I asked him if he had returned since then, 2.5.2 Refugees or if he wishes to return, he cyni- After discussing the main rea- cally emphasized that there is son for return at the beginning nothing for him to return there of this chapter, this section will anymore; neither a house, nor discuss the main reasons for not anything of the symbolic and ma- returning. It is worth reiterating terial past that once marked his that simply reversing refugees’ history, identity, or status.’ concerns is unlikely to lead to In the case of Syrian refu- mass return as every household/ gees, studies conducted on inten- individual situation is different. tions to return show that ties to Therefore, this literature survey the homeland are still intact. As will help to assess the most im- mentioned, the UNHCR (2018a) portant obstacles to return. finds that 76% of Syrian refugees Starting with Yemeni refugees in neighboring country express a in Djibouti, the biggest concern desire to go back at some point. mentioned by news reports is se- But 85% of them do not think it is curity: beyond the general secu- feasible to do so within the next rity in the country, refugees – es- 12 months. The reasons they give pecially young men – are worried are the lack of safety and security about mandatory conscription in Syria, livelihood opportuni- in one of the fighting factions ties, and housing. The 11% who (Adow, 2017). As for Iraqi refu- are undecided on whether they gees, the main hindrance for their want to return one day mention return is the continuing and epi- security improvements, desire for sodic nature of the conflict, con- political change, as well as access fiscation and loss of property, as to basic services as conditions for well as fear of ethnic or religious return. Finally, the 13% who do persecution (ICMC, 2013; Inter- not wish to return at all say that nal Displacement Monitor, 2018). the main reasons are safety, con- In several studies conducted on cern over property, and scarcity Iraqi refugees in Jordan, the idea of livelihood opportunities. This of returning home is no longer on shows that all three groups have the table even in the face of de- almost identical concerns regard- teriorating economic conditions. ing return. The reason some ex- This is especially the case for press reluctance or refusal to re- minority Christians, who do not turn is therefore likely to be an think it is safe for them to go back issue of lack of optimism regard- home, but it is also true for for- ing concrete change in the future 54 | Enabling conditions, challenges, and processes

rather than a desire to stay in the drive refugees to repatriate. As host countries. mentioned at the beginning of While frustration with the liv- this chapter, poorer Iraqi house- ing situation in host countries is holds had to go back to Iraq at the expressed time and time again by onset of the Syrian civil war. Sim- both Syrian and Iraqi refugees, it ilarly, according to Kvittingen et is usually expressed with a caveat al (2018), poorer Syrian refugees about the impossibility of return. in Jordan without social net- For some refugees, going home is works or assets were more likely synonymous with imprisonment to return home, especially with or death so they prefer taking the recent increase in vulnerabil- their chances in the host coun- ity. But it is clear that at this stage, tries even with the deteriorat- the bad living conditions in host ing economic conditions. This is countries are not enough to push expressed by an Iraqi refugee in for mass return. This is because Jordan who was interviewed by going home for refugees means Majeed (2018): worse living conditions, internal ‘… [She] recounted frequently displacement, persecution, or to me, often noting as justifica- death. tion of why she would never re- This brings us to the exten- turn to Iraq ‘Wallah I would rath- sive report by Yahya et al (2018) er die here and live in the street which is based on focus group begging rather than return after discussions conducted in Leba- everything that has happened to non and Jordan with Syrian refu- us.’’ gees in 2017. On asking refugees This sentiment is echoed by what the main obstacle to return Syrian refugees in Lebanon, and is, interviewed Syrians over- indeed seems to reflect the real- whelmingly responded that their ity of the situation. According to most important concern is secu- Yehya et al (2018): rity and safety. According to one ‘Other refugees, particularly refugee interviewed by Yahya et youths, felt pressure to go home al (2018): because of the systematic hu- ‘We will return to Syria if the miliation they faced in Lebanon. regime is gone and there is secu- But several refugees recounted rity, which means there is no kill- stories of young men who had ing and no bombings.’ returned to Syria out of despair, The interesting aspect of this only to die because of forced con- quote, and indeed many other re- scription or conflict in their areas. sponses highlighted in the report, In reflection, Aisha, from Homs, is that refugees make it clear that asked laconically, ‘Would anyone safety and security to them is in- walk toward death on their feet?’’ terrelated with political factors. That is not to say that poverty Most agree that safety means an and bad living conditions do not end to aerial bombings, sieges, Chapter 2 | 55

and checkpoints and express constitutes the consolidation of their belief that violence should power in the hands of the current only be used by ‘legitimate au- regime, without which they will thority’. Political considerations not return. Therefore, while po- are, therefore, intertwined as- litical settlement is seen as neces- pect of security, with majority sary for mass return, there is no of the refugees saying that they agreement on the nature of that will not go back without a politi- political settlement. cal settlement to end the conflict The gender dimension to even if livelihood opportuni- the conditions of return shows ties are available. On the other that women are less likely to hand, they do indicate that they compromise and express a de- would return in the case of politi- sire for return. Consistent with cal settlement even with minimal the literature, women particu- economic or livelihood opportu- larly express the need for basic nities. According to Yahya et al services, in terms of education, (2018): health, and housing. On the other ‘A large number of refugees hand, young men express fears also indicated that even if their of mandatory conscription in the homes were destroyed, they army on returning, especially would still return to Syria if se- with new legislative decree that curity and political conditions al- makes it harder for them to go lowed it and jobs were available. back. Finally, the report mentions They insisted they would pitch a refugees’ reliance on informal in- tent on the rubble of their homes formation sources to gauge the and rebuild them.’ situation back home. This is es- But there are diverging opin- pecially relevant when it comes ions on the specificities of the to the situation in areas of origin political conditions for return. or whether refugees will be al- Different fractions of the Syrian lowed back to the area they hail refugee population, depending from as well as access to land and on political affiliation, have di- property after leaving and the verging ideas on what constitutes procedures they need to follow in a desirable political settlement. order to get them back (Yahya et For example, some of the refu- al, 2018). Unlike most other con- gees stress the need for a regime flicts with an ethnic dimension, change without specifying the the ethnic/religious majority are alternative, while others, espe- equally wary of returning as the cially those who are educated ethnic minorities with worries of and living in Europe, stress the destruction and lack of optimism need for sustainable democracy about the future of Syria (Yahya (Kaya and Orchard, 2019). On the et al, 2018). other hand, for some pro-regime While the focus of this report refugees, the political settlement is on refugees in a neighboring 56 | Enabling conditions, challenges, and processes

host country, one relevant study violence. The similarity in priori- on the intention of return was ties between those in neighboring conducted in Germany with in- host countries and those in Ger- teresting findings summarized many despite the different liv- in the Table 2.7 (Kaya and Or- ing conditions further supports chard, 2019). In line with Yahya the widespread desire to go back et al (2018) and UNHCR (2018a), home and the importance of pull the analysis of the Syrian refu- factors in making it happen. As gee survey shows that the vast mentioned, however, educated majority of them want to return refugees living in Germany par- at some point (90%) and that ticularly stress the importance of safety and security is the most sustainable democracy as a con- important concern, especially for dition for return. those who have been subject to

Table 2.7: How does Syria have to change in the future so you could have the option to go back from Germany?73

Response Response percentage War has to stop 67.8% Regime change 51.5% ISIS leaves country 43.8% Democracy with free and equal election 41.7% Reconciliation between religions and ethnic groups 28.8% Economic situation gets better 25.9% Autonomy for Kurds 16.1% I don’t want to go back to Syria 8.4%

The aforementioned studies litical implication of repatriation are very useful in highlighting is not lost on refugees and they the main concerns that Syrian ref- see their safety as conditional on ugees have against return. Both political changes. Property rights emphasize how security, in terms and services, two factors that are of personal persecution and con- especially concerning for women, tinuing violence, plays the most are also extensively mentioned important role in deterring re- as necessary conditions for return. But it also highlights how turn. While access to livelihood the security and the political di- is also mentioned several times, it mension are intertwined; the po- does not feature among the most important concerns. It is worth noting that with the increased 73 Kaya and Orchard (2019) Chapter 2 | 57

hopelessness in any real politi- imagined that the crisis would go cal change in Syria, refugees in on this long.’ neighboring countries are becom- The biggest takeaway from ing more open to the idea of reset- this literature survey is that mass tlement and less optimistic about return is conditional on security, repatriation (Yahya et al, 2018). which is derived from political If this persists, the situation may change, followed by economic start to mirror that of Iraqi refu- concerns, most notably property gees, who do not see returning rights. The question worth ask- home as an option anymore. In- ing is to what extent is political deed, some refugees have begun change a necessary condition for expressing lost hope. According return? If personal safety is guar- to Kvittingen et al (2018): anteed and the violence stops, will ‘All waited to return to Syria political settlement still be an es- pending safety but, as they ‘lost sential condition for repatriation? hope’ in a resolution to the con- The interviews seem to imply flict, their plans changed. A mid- that the answer is yes, without a dle-aged Syrian man explained: real political change there will be At the beginning of the crisis, no mass return despite the harsh everyone was saying that there conditions in exile. But if this is no was still hope. That there was a longer on the table, the best bet for solution, that the world wasn’t return would rest on providing going to let it get to this point, guarantee for personal safety of or they were hopeful that things returnees and ensuring that their wouldn’t get this bad. No one property rights are respected.

Table 2.8: The necessary factors and conditions for return

Safety and security Political settlement Economic drivers Priorities74 (primary) (co-primary) (secondary) Refugees in Sustainable General safety Economic stability Germany/Europe75 democracy Young male Fear of mandatory Livelihood refugees conscription opportunities - Need for property Female refugees General safety rights and provision of services Ethnic minorities Minority protection Ambiguous Property rights Pro-Regime General safety Regime remains - Anti-Regime General safety Regime change

74 Kaya and Orchard (2019) 75 Kaya and Orchard (2019) 58 | Enabling conditions, challenges, and processes

2.6 Conclusions On studying the experience of Arab refugees, various conclu- The purpose of this chapter sions can be drawn. The first and has been threefold. The first was most important is that return at to discuss the different aspects this stage is not sustainable for that factor into the decision to Yemenis, Iraqis, or Syrians. For repatriate. The second was to in- the case of Yemen, deteriorating troduce other dimensions of re- economic conditions in Djibouti patriation, namely networks and are nonetheless pushing for early information-gathering and the repatriation. As for Iraqis, family sustainability of return. Finally, reunification, the deteriorating the lessons learned were applied security situation in Syria, and to the case of refugees by provid- the bad socio-economic condi- ing an overview of news reports tions in exile are all drivers for re- and surveys conducted on early patriation. For Syrians, the main repatriates and intentions for re- motivation for return seems to be turn. family reunions, but perceived improvement in the security situ- Four dimensions to the deci- ation also plays a role. sion-making process were discussed. The paradox of security Finally, with the continuing concerns was addressed, as well conflicts in all three countries, it is as the political, economic, and no surprise that refugees mention social drivers for repatriation. A security as the main condition for special focus was paid to house- return. This not only takes the holds headed by women given form of an end to the violence their diverging experience. The but also in sustainable peace and approach was to consider refu- guaranteed individual safety. gees as purposive actors with a Consequently, this means that degree of autonomy that they refugees to varying degrees – but exercise by comparing their situ- from all three countries – are im- ation in exile to the situation back plicitly or explicitly conditioning home: only then will they make their return on a concrete political their decision to repatriate. The settlement. Beyond that, property way information is gathered rights and access to shelter fea- throughout refugee communi- ture as the second-most impor- ties was also touched on, as well tant condition to return. The link as the sustainability of the return between political settlements and encompassing the welfare of repatriation will be discussed in the refugees themselves and the more detail in Chapter 3. home country. Chapter 2 | 59

References Collins, J. (1996). An analysis of voluntariness of refugee repatria- Adow, M. (2017). Yemenis find tion in Africa. refuge in Africa’s Djibouti. Crisp. (2017). It’s Far Too Early to Retrieved April 25, 2019, from Talk of Return for Syrian Ref- https://www.aljazeera.com/ ugees. Retrieved December 3, news/2017/11/yemenis- 2018, from Chatham House find-refuge-africas-djibou- website: https://www.cha- ti-171117110454303.html thamhouse.org//node/31020 Al-Rasheed, M. (1994). The myth Crisp, J. (2003). No Solution in of return: Iraqi Arab and Sight : The Problem of Protracted Assyrian refugees in Lon- Refugee Situations in Africa. Re- don. Journal of Refugee Stud- trieved from https://escholar- ies, 7(2-3), 199-219. ship.org/uc/item/89d8r34q Black, R., and Gent, S. (2004). De- Dadush, U. (2018). The Economic fining, Measuring and Influ- Effects of Refugee Return and encing Sustainable Return: The Policy Implications. https:// Case of the Balkans. 23. doi.org/10.1596/1813-9450- Black, R., and Koser, K. (1999). 8497 The End of the Refugee Cycle?: Dockery, W. (2017, October 5). Refugee Repatriation and Recon- What’s Jordan’s policy to- struction. Berghahn Books. wards Syrian and Iraqi refu- Black, R., Koser, K., Munk, K., gees? Retrieved April 25, 2019, Atfield, G., D’Onofrio, L., and from InfoMigrants website: Tiemoko, R. (2004). Under- https://www.infomigrants. standing voluntary return. 80. net/en/post/5470/what-s- jordan-s-policy-towards-syri- Carlson, M., Jakli, L., and Linos, an-and-iraqi-refugees K. (2018). Rumors and Refu- gees: How Government-Cre- Flow Monitoring Dashboard ated Information Vacuums (2019). Flow Monitoring Undermine Effective Crisis Points | Migrant Arrivals Management. International and Yemeni Returns from Studies Quarterly, 62(3), 671– Saudi Arabia in April 2019. 685. https://doi.org/10.1093/ Retrieved July 26, 2019, from isq/sqy018 migration.iom.int website: https://migration.iom.int/ Chatelard, G. (2003). Iraqi forced reports/yemen-%E2%80%94- migrants in Jordan: Condi- flow-monitoring-dashboard- tions, religious networks and april-2019?close=true the smuggling process (No. 2003/34). WIDER Discus- sion Papers//World Institute for Development Economics (UNU-WIDER). 60 | Enabling conditions, challenges, and processes

Fransen, S. (2017). The Socio- Harlid, N., Christensen, A., and Economic Sustainability of Zetter, R. (2015). SUSTAIN- Refugee Return: Insights from ABLE REFUGEE RETURN: Burundi: The Socio-Economic Triggers, constraints, and les- Sustainability of Refugee Re- sons on Addressing the develop- turn. Population, Space and ment challenges of forced dis- Place, 23(1), e1976. https:// placement [Data set]. https:// doi.org/10.1002/psp.1976 doi.org/10.1163/2210-7975_ Gerver, M. (2018). Refugee Re- HRD-3181-2015006 patriation and the Problem of Harper, A. (2008). Iraq’s refu- Consent. British Journal of Po- gees: Ignored and unwant- litical Science, 48(4), 855–875. ed. International Review https://doi.org/10.1017/ of the Red Cross, 90(869). S0007123416000193 https://doi.org/10.1017/ Gillespie, M., Ampofo, L., Chees- S1816383108000179 man, M., Faith, B., Iliadou, Helling, D. (2007). Working Paper E., Issa, A., … Skleparis, D. Series. 60. (2016). Mapping Refugee Human Rights Watch. (2016). Media Journeys. Retrieved Pakistan Coercion, UN Com- April 14, 2019, from https:// plicity: The Mass Forced Re- www.open.ac.uk/ccig/sites/ turn of Afghan Refugees. Re- www.open.ac.uk.ccig/files/ trieved July, 27, 2019, from Mapping%20Refugee%20 Human Rights Watch web- Media%20Journeys%2016%20 site: https://www.hrw.org/ May%20FIN%20MG_0.pdf report/2017/02/13/paki- Graham, M., and Khosravi, S. stan-coercion-un-complicity/ (1997). Home is Where You mass-forced-return-afghan- Make It: Repatriation and refugees Diaspora Culture among Ira- Human Rights Watch. (2017, nians in Sweden. Journal of October 2). Jordan: Syrian Refugee Studies, 10(2), 115–133. Refugees Being Summarily https://doi.org/10.1093/ Deported. Retrieved April 25, jrs/10.2.115 2019, from Human Rights Hardgrove, A. (2009). Liberian Watch website: https://www. Refugee Families in Ghana: hrw.org/news/2017/10/02/ The Implications of Family jordan-syrian-refugees-being- Demands and Capabilities for summarily-deported Return to Liberia. Journal of ICMC. (2013). Iraqi refugees in Refugee Studies, 22(4), 483–501. Jordan, Lebanon and Syria | https://doi.org/10.1093/jrs/ European Resettlement Net- fep029 work. Retrieved April 25, 2019, from https://www. resettlement.eu/page/iraqi- refugees-jordan-lebanon-syria Chapter 2 | 61

Internal Displacement Monitor. Marks, J. (2018). Pushing Syrian (2018). Nowhere to return to: Refugees to Return. Retrieved Iraqis’ search for durable so- December 3, 2018, from Carn- lutions continues. Retrieved egie Endowment for Interna- April 25, 2019, from http:// tional Peace website: http:// www.internal-displacement. carnegieendowment.org/ org/sites/default/files/pub- sada/75684 lications/documents/201811- Martin, S. F. (2004). Refugee Wom- iraq-case-study-report.pdf en. Lexington Books. Jacobsen, K. (1996). Factors Influ- Muasher, M., and Yahya, M. encing the Policy Responses (2019). Refugee Crises in of Host Governments to Mass the Arab World. Retrieved Refugee Influxes. The Inter- April 25, 2019, from Carn- national Migration Review, egie Endowment for In- 30(3), 655–678. https://doi. ternational Peace website: org/10.2307/2547631 https://carnegieendowment. Jallow, A. T. (2004). Handbook for org/2018/10/18/refugee-cri- Repatriation and Reintegration ses-in-arab-world-pub-77522 Activities. 264. Muggeridge, H., and Doná, G. Kaya, S., and Orchard, P. (2019). (2006). Back Home? Refugees’ Prospects of Return: The Case Experiences of their First Visit of Syrian Refugees in Germa- back to their Country of Ori- ny. Journal of Immigrant and gin. Journal of Refugee Studies, Refugee Studies, 1–18. https:// 19(4), 415–432. https://doi. doi.org/10.1080/15562948.20 org/10.1093/refuge/fel020 19.1570579 Omata, N. (2013). The Complex- Kvittingen, A., Valenta, M., Tab- ity of Refugees’ Return Deci- bara, H., Baslan, D., and Berg, sion-Making in a Protracted B. (2018). The Conditions and Exile: Beyond the Home- Migratory Aspirations of Syr- Coming Model and Durable ian and Iraqi Refugees in Jor- Solutions. Journal of Ethnic and dan. Journal of Refugee Stud- Migration Studies, 39(8), 1281– ies, 32(1), 106-124. 1297. https://doi.org/10.1080 Lischer, S. K. (2017). The Global /1369183X.2013.778149 Refugee Crisis: Regional De- Petrin, S. (2002). Refugee return stabilization and Humani- and state reconstruction: A tarian Protection. Daedalus, comparative analysis, Sarah 146(4), 85–97. https://doi. Petrin. Retrieved April 24, org/10.1162/DAED_a_00461 2019, from UNHCR website: Majeed, A. (2018). Unsettling the https://www.unhcr.org/re- Homeland: Fragments of Home search/working/3d5d0ec94/ and Homeland among Iraqi Ex- refugee-return-state-recon- iles in Amman, Jordan. 130. struction-comparative-analysis-sarah-petrin.html 62 | Enabling conditions, challenges, and processes

Refworld (1997). Committee for Taiwa, K. K. (2016). GENDER Refugees World Refugee Sur- IDENTITIES AND THE DE- vey 1997 - Rwanda. Retrieved CISION TO RETURN: THE July 26, 2019, from refworld CASE OF RWANDAN REFU- website: https://www.ref- GEE MEN AND WOMEN IN world.org/docid/3ae6a8b90. NORTH KIVU, DRC. 93. html UNHCR. (1995). Handbook on ReliefWeb (2019). South Sudan – Voluntary Repatriation: Inter- Spontaneous Returns of Refu- national Protection. Retrieved gees from Ethiopia, Retrieved December 4, 2018, from UN- July 27, 2019, from Reliefweb HCR website: https://www. website: https://reliefweb. unhcr.org/publications/ int/report/south-sudan/ legal/3bfe68d32/handbook- south-sudan-spontaneous- voluntary-repatriation-inter- returns-refugees-ethiopia-dg- national-protection.html echo-ingos-echo-daily-flash UNHCR. (2015). Refworld | UN- Savage, R., and Kalfood, M. A. HCR Position on Returns to (2016). All roads lead to Dji- Yemen. Retrieved April 25, bouti as refugees flee Yemen 2019, from Refworld website: even as migrants head there https://www.refworld.org/ | Global development | The docid/5523fdf84.html Guardian. Retrieved April 25, UNHCR. (2016). UNHCR po- 2019, from https://www.the- sition on Iraq return. Re- guardian.com/global-devel- trieved April 25, 2019, from opment/2016/may/23/all- https://www.refworld.org/ roads-djibouti-refugees-flee- pdfid/58299e694.pdf yemen-civil-war-migrants- UNHCR. (2017a). Durable Solu- head-there tions for Syrian Refugees. Re- SAWA. (2019). Unpacking Re- trieved April 25, 2019, from turn: Syrian Refugees’ Condi- https://data2.unhcr.org/ar/ tions and Concern. Retrieved documents/download/59482 April 25, 2019, from https:// UNHCR. (2017b). Refworld | In- reliefweb.int/sites/reliefweb. ternational Protection Consid- int/files/resources/SAWA_ erations with regard to people Unpacking%20Return%20Re- fleeing the Syrian Arab Re- port.pdf public, Update V. Retrieved Stein, B. N., and Cuny, F. C. April 25, 2019, from Refworld (1994). Refugee repatriation website: https://www.ref- during conflict: Protection and world.org/docid/59f365034. post-return assistance. Devel- html opment in Practice, 4(3), 173– 187. https://doi. org/10.1080 /096145249100077811 Chapter 2 | 63

UNHCR. (2018a). FOURTH RE- Yahya, M., Kassir, J., and El- GIONAL SURVEY ON SYR- Hariri, K. (2018). Refugee IAN REFUGEES’ PERCEP- Attitudes Toward a Return TIONS AND INTENTIONS to Syria - Unheard Voices: ON RETURN TO SYRIA. Re- What Syrian Refugees Need trieved April 25, 2019, from to Return Home. Retrieved https://data2.unhcr.org/en/ December 3, 2018, from Carn- documents/download/66198 egie Middle East Center web- UNHCR. (2018b). Monitoring of site: https://carnegie-mec. Voluntary Repatriation In- org/2018/04/16/refugee- terviews of Syrian refugees. attitudes-toward-return-to- Retrieved December 3, 2018, syria-pub-76061 from https://data2.unhcr. Zimmermann, S. (2012). Un- org/en/documents/down- derstanding repatriation: load/63309 Refugee perspectives on the UNHCR. (2018c). Returns of Syr- importance of safety, reinte- ian Refugees in Jordan - Anal- gration, and hope. Population, ysis Dashboard. Retrieved Space and Place, 18(1), 45–57. July 26, 2019 from https:// https://doi.org/10.1002/ data2.unhcr.org/en/docu- psp.647 ments/download/64996 UNHCR. (2019) Amid Wors- ening conditions in Ye- men, more Somali refugees opt to return. Retrieved July 26, 2019 from https:// www.unhcr.org/uk/news/ briefing/2019/5/5cf0e7eb4/ amid-worsening-conditions- yemen-somali-refugees-opt- return.html Wall, M., Campbell, M. O., and Jan- bek, D. M. (2017). Syrian refugees and information precarity. New Media and Society, 19, 240– 254. https://doi.org/10.1177 /1461444815591967 World Bank. (2019). The Mobility of Displaced Syrians An Eco- nomic and Social Analysis.

CHAPTER 3

Repatriation: scenarios for conflict resolution and reconstruction Ibrahim Elbadawi, Roger Albinyana, Maryse Louis and Jala Youssef

Devastating conflicts and and the world’s response, is a big state failure have followed the part of that. Syria accounts for Arab uprisings of 2011 in some 24% of the world’s 22,5 million MENA countries. The Syrian refugees, according to the United war continues to displace Syr- Nations High Commissioner for ians, and with Western countries Refugees (UNHCR, 2018). What collectively resettling fewer Syr- truly distinguishes the current ian refugees, there is intensifying refugee crisis is that nowadays pressure on the country’s neigh- refugees are likely to remain dis- bors, which are already hosting placed for years or even decades, millions of refugees amid high instead of being able either to re- unemployment and regional in- turn home or to be permanently stability. Yemen has endured resettled in a third country. While the complete destruction of its resettlement has always been economy, pushing its popula- relatively rare (only 1% of global tion to massive internal forced refugees get resettled – UNHCR, displacements and putting them 2010), it is most likely to be used at risk of starvation and disease. in cases where it is evident that In Iraq, the eradication of the Is- the situation in refugees’ home lamic State came at great human countries will never get better, cost, causing massive internal and that they need to stop wait- displacements; while in Libya, ing and make a new life. the successive armed conflicts While none of these conflicts have eroded the remaining state has yet fully ended, attention is structures, which were already increasingly being placed on the characterized at the beginning of emerging needs of repatriation the conflict by weakness or even and reconstruction. For example, complete absence. long-term settlement of Syrian At the same time, the global refugees in Turkey, Lebanon, refugee crisis is arguably worse and Jordan entails considerable than ever. The Syrian civil war, financial commitments from the 65 66 | Repatriation: scenarios for conflict resolution and reconstruction

international donor community Credible political settlement since the burden on Lebanon and scenarios will be compared to Jordan to cater for Syrian refu- the de facto situation, with subse- gees has surpassed their financial quent discussion of different re- capacities. Despite the UN pledg- construction plans. The European ing $12 billion for host coun- Union (EU) and the United States tries in 2016 and an additional are generally leaning towards $6 billion in 2017, the disbursed pushing for repatriation; the US amount has fallen as short as 54% government has made it clear of pledged funding for 2015, 46% that resettlement in the United in 2016 and 43% in 2017 for Leba- States is not an option, and US non, whereas for Jordan, it was and EU humanitarian aid going only 62% in 2016 and 65% in 2017 to host countries is expensive and (UNSCOL, 2018). In these cir- endless (Crisp, 2017). The UN- cumstances, refugees come under HCR takes the view that it can mounting pressure to repatriate neither promote nor facilitate the to Syria. return of MENA refugees, espe- As for reconstruction, the UN cially Syrians, but given its de- Special Envoy has estimated a pendence on Western donors, the cost of $250 billion to reconstruct organization has been engaged in Syria; the World Bank has esti- discreet planning for repatriation mated a cost of $40 billion to re- (UNHCR 2018b). This chapter construct Yemen; while for Libya will also shed light on the linkag- it is not so easy to estimate a cost es between peace agreements (as as the country is enduring a new a proxy for political settlements) outbreak of violence between the and the share of returned refu- forces of Marshal Khalifa Haf- gees on the ground. tar, self-proclaimed leader of the By developing a better under- Libyan National Army, and the standing of the literature on repa- Government of National Accord triation and political settlements, under Prime Minister Fayez Sar- alongside the different post-con- raj, the authority recognized by flict scenarios for Syria, this chap- the international community. ter will elucidate enabling and This chapter discusses the inhibiting factors for repatria- very basic notion and literature tion and sustainability of return. of political settlements and re- Importantly, given the vary- construction plans in the MENA ing degree of cohesion between region, and their implications for refugees and host communities, refugee repatriation and resettle- as well as the role of interna- ment. There is a special focus tional agreements, the outcomes on Syria, which has been one of may be different for refugees in the most devastating conflicts neighboring countries and those in recent history, with consider- in the West. able spillovers onto neighboring countries. Chapter 3 | 67

3.1 Repatriation of refugees to such fear, is unwilling to avail and their rights himself of the protection of that country’ (UN General Assembly, 1951). In this section, we review the literature on refugees and state- The assumption of refugee building through refugee repa- status entails the necessary loss triation. While taking account of of protection for the refugee from the fact that refugee repatriation the country of his/her national- is not only the preferred option ity. Under normal circumstances, by the international community, a social contract would prevail, but also the only solution to mass and this relationship between the exoduses, this section argues that state and the citizen would not refugee repatriation is an oppor- have broken. Therefore, the con- tunity to build more inclusive cept of repatriation implies the states that guarantee their citi- necessary restoration of protec- zens safety over the long term, tion and the rebuilding of a social given that ‘repatriation must be contract between the refugee and understood not in terms of physi- the state (Long, 2012). cal return but as a process of po- Refugee flows are caused in litical rapprochement between many occasions by the loss of citizen, community and state’ state protection on grounds of (Long, 2012). The section also group-based identities, whether provides an overview of the ap- national, political, ethnic, reli- proaches that the international gious, or cultural. Whenever this community, particularly the EU is the case, repatriation can be- and the United States, have taken come instrumental in state-build- to refugees and reconstruction of ing as a holistic process of restor- the four conflict countries. ing the balance between citizen, In order to develop a better nation, and state (Crisp, 2001). understanding of the impact of In recent years, the UNHCR’s repatriation on the rights of refu- repatriation policy has embraced gees, and to recognize that the the importance of post-conflict term repatriation means more state-building. To that end, than simply a return to a refu- greater emphasis has been given gee’s home, one needs to refer to to repatriation programs that ac- the definition embedded in the knowledge the long-term com- 1951 Convention on the Status of plexity of reintegration and the Refugees. This defines a refugee sustainability goal in returnees’ as ‘an individual who owing to reintegration (UN General As- well-founded fear of being per- sembly 1999). Further on, it has to secuted for reasons of race, reli- be noted that the UNHCR has as- gion, nationality, membership of serted that the root causes for ref- a particular social group of politi- ugee fleeing are violence and per- cal opinion, is unable, or owing secution and not poverty (Crisp, 68 | Repatriation: scenarios for conflict resolution and reconstruction

2001). Hence, it is worth consid- Djibouti and the Cambodians ering the political dimension of in Thailand – were all subjected state-building, and how the latter to coerced repatriations, which can be addressed through a pro- were at least partly condoned by cess of refugee repatriation. the UNHCR and the wider inter- From a theoretical point of national community (Crisp and view, the anchor of refugee pro- Long, 2016). These practices were tection is the principle of non-re- neither founded on standards of foulement, as laid down in Article voluntariness nor of safety. In this 33 of the UN Refugee Convention regard, faced with the increasing of 1951, which entails that ‘no number of refugees in the world contracting state shall expel or during that period of time, the return a refugee in any manner Executive Committee Conclusion 29 of 1983, called on states to fa- whatsoever to the frontiers of ter- cilitate the work of the UNHCR ritories where his life or freedom ‘in creating conditions favorable would be threatened.’ to and promoting voluntary re- The contribution that the patriation, which whenever ap- principle of voluntary repatria- propriate and feasible is the most tion makes to refugee protec- desirable solution for refugee tion is obvious. Yet it is equally problems’ (UNHCR 1983), which clear that the limitations such triggered mounting pressure for a principle places on repatria- refugees to repatriate. tion is frequently resented by The UNHCR frequently plays states. Host countries are often an ambiguous role in such nego- impatient to see uninvited refu- tiations, seeking to uphold inter- gees leave. Countries of origin nationally recognized legal and are sometimes impatient to see ethical principles of which it is them return and signal the end the custodian, while simultane- of conflict. Donor states are eager ously meeting the concerns of to bring an end to the long-term these stakeholders. In the worst refugee assistance programs that cases, the UN refugee agency has they fund (Crisp and Long, 2016). capitulated to such pressures and In addition, Chimni (1993) argues actively engaged in repatriation that the contemporary modality operations that are far from safe of repatriation is underpinned by or voluntary (Crisp and Long, the relationship between volun- 2016). Despite the most recent tariness, conceived as an essential evidence of repatriation focusing protection against refoulement, on post-1980 developments, the and safety, which represents the latter has been the most preferred most objective standard for as- solution to tackle the challenges sessing the suitability of return. posed by refugees among the in- In the 1970s and 1980s, a ternational community as stated number of refugee populations by the League of Nations as early – for example, the Ethiopians in as 1922. Chapter 3 | 69

Refugee return is likely to tion must become an opportunity vary from one country to another to build more inclusive societies and from one region to another whereby citizens’ safety is guar- within different countries. It will anteed as a means of refugee re- depend on conditions in the areas patriation. of origin, on the nature of the political bargain that ends the conflict, and the willingness of the ruling elite to allow populations 3.2 Reconstruction: defini- it may consider politically hostile tions, scenarios, and im- at best, to come back to the coun- pact on refugees try or to move back to areas of interest. Return may well mean Discussion of reconstruc- secondary displacement, as refu- tion often begins from economic gees are unable to return to de- needs assessment and templates stroyed homes or to prove their derived from international best ownership of confiscated prop- practices, rather than through erty (Lynch and Yahya, 2018). engagement with the affected If the 1990s was a decade of individuals or with the actual re- repatriation, the last 15 years alities on the ground especially in are best described as ‘a decade the aftermath of civil conflicts. In of protracted emergencies’ (UN- fact, reconstruction will be deeply HCR, 2015). Millions of new refu- intertwined with the reconfigura- gees have been created as a result tion of power relations in societ- of intense violence in Burundi, ies emerging from conflict (Lynch the Central African Republic, and Yahya, 2018). Iraq, Nigeria, South Sudan, Syria, As with political settlements, Ukraine, and Yemen. At the same reconstruction has to start from time, longstanding conflicts in the nationals, or at least they countries such as Afghanistan, should have the most important the Democratic Republic of Con- role in deciding the direction of go, Myanmar, Somalia, and Su- reconstruction. Some might in- dan have gone unresolved (Crisp fer that reconstruction means a and Long, 2016). rebuilding of the status quo ante, Whenever millions of people something that might be neither have been dispossessed from normatively desirable nor politi- their properties, driven into exile cally possible. Others might see at home or abroad due to armed the push to begin thinking about conflicts, which have unfolded reconstruction as a political drive across multiple levels, recon- to force an end to any viable sup- struction can never be separated port for the conflict itself. from politics, and it will never be For example, European and solely triggered by economic or US discussions about how they humanitarian needs. Reconstruc- might ‘win’ the reconstruction of 70 | Repatriation: scenarios for conflict resolution and reconstruction

Syria could be a face-saving way struction funding should not in- to move on from more than a advertently empower the Syrian half decade of attempting to win regime. Starting on a small scale through war. The Assad regime in regions that are not under reviews calls for reconstruction as a gime control could provide a bet- way of signaling the end of con- ter alternative for local rebuilding flict and the beginning of his in- efforts. Any funding should also ternational rehabilitation (Lynch be conditional on the return of and Yahya, 2018). Some might refugees to their homes and ac- also see a focus on reconstruction cess to their property. A vetting as a way to avoid dealing with process should be established to the difficult issues of responsibil- ensure that local entities receiv- ity, especially in contexts such as ing international funding have Syria where the key party to the not been involved in war crimes conflict has been called out for and are not regime fronts. forced population transfers and As we witness considerable for crimes against humanity. A disagreements among interna- focus on physical rebuilding in tional actors on how to end these such a context implies that any armed conflicts, the question justice mechanism, including of where reconstruction fund- transitional justice or the articula- ing will come from is also very tion of shared memories, can and uncertain. It is clear that the im- will take a back seat to the eco- portant role of the international nomic opportunities and political community and international gambits that define post-conflict. donors in the reconstruction can- Critically, such a focus is likely to not be taken for granted. The aim undermine international norms of international intervention is to of accountability and justice, and provide support to establish le- will serve to reward those ac- gitimate political authority that cused of such crimes (Lynch and is ‘democratic, accountable and Yahya, 2018). self-sustained’ (Bojicic-Dzelilovic Reconstruction in places like et al, 2016). The state’s capability Syria is particularly complicated to assume its functions in an equi- by the questions of how assis- table, impartial, and effective way tance can be given to a regime guarantees the exit of external ac- that was in large part responsi- tors. But reconstruction history ble for the country’s devastation points to some contradictory and and has been implicated in war very complex experiences and crimes. International actors today dilemmas. As the international are struggling with whether and community represents different how to support reconstruction stakeholders, a more complex for Syrian communities while approach to their intervention ensuring that this does not end needs to take account of norms up privileging political support- such as sovereignty, accountabil- ers of the regime. Indeed, recon- ity, legitimacy, and ownership Chapter 3 | 71

while providing needed support posed, which defines a bottom- for local authorities. Striking ex- up approach that is more atten- amples are the US interventions tive to local needs and structures, in both Iraq and Afghanistan in and which defines the relation- the past two decades. ship between the local popula- Coming back to the case of tion and the international actors Syria, on the one hand, Russia (Bojicic-Dzelilovic et al, 2014). and its regional ally Iran do not In fact, the very term ‘globaliza- have the financial strength to tion’ makes it almost impossible contribute significantly to the to achieve reconstruction without rebuilding (of what they were the support of the international complicit in destroying). On the community (Luckham, 2003). other hand, European and US as- While a ‘neutral’ role could be sistance, even if plausible, would played by organization such as be subject to a post-conflict po- the UN (not always a clear orien- litical settlement, which in Syria, tation) under their responsibility for example, is far away. While of observing humanitarian law, China’s President Xi Jinping and to prevent human losses (or pledged $20 billion in loans to abuses), it will be unrealistic to revive the Middle East in 2018 ignore the interests of major in- (Reuters 2018), it is very unlikely ternational players in post-con- that this financial package will be flict reconstructions and what it entirely spent in the reconstruc- involves to balance the powers: tion of countries like Libya, Syria, the United States, the EU, Russia, and Yemen. Only the Gulf coun- Iran, China and rich Middle East- tries have the financial strength ern countries. to fund large-scale reconstruction The role of the international interventions, even though they community remains an important are parties in the conflicts of these part of reconstruction, not only in countries, and any interventions terms of infrastructure and other would be eventually subject to physical considerations, but also political compensations. for rebuilding the state and its Hence, smooth transition to a institutions. The interest of these reconstruction phase is rare. One communities in providing sup- can argue that in a post-conflict port to post-conflict resolution is era, norms of sovereignty and ac- rooted in their demand for politi- countability have already been cal stabilization, but also for gain- affected. Moreover, an ex-ante ing power vis-à-vis counterparts. status will lead to failure of the In this regard, and in the con- state again as new political reali- text of a post-conflict political ties that cannot be ignored would transition, a safe and sustainable have emerged. Therefore, a hy- return of refugees often requires brid model of post-liberal peace a framework that acknowledges for intervention could be pro- the political roots of the armed 72 | Repatriation: scenarios for conflict resolution and reconstruction

conflict rather than just its hu- Amid the conflicts examined manitarian dimension. To this in this chapter, it is Syria and end, safety and security can only Iraq, and to a lesser extent Libya, be guaranteed through a politi- that are the ones whose refugees cal process that creates inclusive have spilled over into the EU. governance mechanisms, ends Currently, the EU and its mem- criminal impunity, and facilitates ber states’ external responses reintegration, demilitarization, to the Syrian refugee crisis are and access to justice. to end the violence and support While this process will take efforts to bring about political time given the existing fragmen- change in Syria, while simulta- tation within post-conflict scenar- neously providing immediate ios, efforts to prepare refugees for humanitarian assistance to refu- a return have to begin in advance. gees and the internally displaced. These could include providing le- These two measures, one political gal assistance to inform refugees and the other humanitarian, are of their rights and to help resolve related and both, directly or in- the many anticipated local dis- directly, assist the Syrian people putes. They could also include by concurrently addressing their establishing a network of trusted immediate and future needs, the community mediators. former being life sustaining aid Above all, refugees’ right to a and the latter being a viable na- voluntary return must be respect- tion to return to once the conflict ed. To encourage host countries subsides (Fargues and Fandrich, to adopt policies that secure the 2012). basic needs of refugees, interna- The number of asylum ap- tional support must include both plications from citizens of non- humanitarian aid and economic EU member states has gradually investments geared towards job increased in EU countries since creation for host country nation- 2008, but it reached a peak with als and refugees. 1,392,610 total applications in 2015 and 1,260,900 in 2016, up from 662,165 total applications 3.2.1. The EU and US approaches to refugees, repatriation, and re- in 2014 and 712,200 in 2017 (Eu- construction rostat, 2019). With the number While many refugees of previ- of applications in 2015 and 2016 ous waves (up to the end of the more than doubling the total 1990s) could reach Europe where number recorded in 2013 and they were hosted under a variety 2014, this is an indication of the of statuses (as refugees or mi- magnitude of the refugee wave in grants, temporary or long-term), those two years in the EU. only a tiny proportion of those Germany received most of fleeing conflicts in the 2000s were these applications with 476,459 admitted to Europe. total applications in 2015, despite Chapter 3 | 73

a large influx of about 890,000 ar- ‘Nobody will win neither the rivals that same year. In the year war nor the peace without a ne- after (2016), Germany registered gotiated political solution’, EU as many as 745,545 asylum ap- High Representative Federica plications, despite only about Mogherini said at an EU side 280,000 of those concerned being event on the margins of the UN new arrivals, while the remaining General Assembly (EEAS, 2018). number were formal registrations The Brussels Conference sup- of protection claims expressed in porting the future of Syria and 2015. Italy and France were also the region in September 2018 among the top destinations for highlighted the need to uphold asylum-seekers, receiving 59,165 a political transition necessary to and 50,840 applications respec- allow the safe return of refugees, tively in 2015, and 123,482 and a goal that the EU supports, and 85,244 applications respectively provide a stage for Syrian wom- in 2016. Sweden witnessed a en, youth, and civil society to drastic decrease in the number build the future of Syria. ‘Every of asylum-seekers from over Syrian dreams to go back home. 162,000 in 2015 to 28,939 in 2016 And we can only support this after the introduction of internal dream’, the High Representative border controls from December said, thanking countries hosting 2015 (Asylum Information Data- refugees (EEAS, 2018). base, 2016 and 2017). But, Mogherini added, only Of those seeking asylum a tiny percentage of refugees are within the EU, Syria has been the willing to return to Syria in the main country of citizenship from current circumstances: ‘Syria is 2013 onwards. For example, in not safe for them to return. Still 2016 in Germany, a total num- today, most Syrians simply do ber of 288,866 applications were not have a home to go back to. received from Syrian nationals Their properties are being con- and 97,162 applications from fiscated, and many would risk Iraqi nationals. Syrians were forced conscription or unlawful among the top three nationalities arrest’. Europeans are willing to of asylum-seekers in 2016 in Ger- contribute to the reconstruction, many, Greece, Austria, Hungary, but a political process must be Sweden, Switzerland, Nether- firmly under way, the High Rep- lands, Bulgaria, Belgium, Spain, resentative said. ‘Most Syrians do Finland, Norway, Cyprus, Croa- not want to go back to a country tia, Ireland, Malta, Slovenia and where there is no accountabil- Serbia, while Iraqis were among the top three nationalities in 2016 ity and where political detainees in Germany, Greece, Austria, are still in jail’, she added. ‘We UK, Sweden, Bulgaria, Belgium, discussed ways in which we can Finland, Croatia, Slovenia and increase even more our humani- Serbia (Asylum Information Da- tarian assistance for Syrians, in- tabase, 2017). cluding inside Syria, while still 74 | Repatriation: scenarios for conflict resolution and reconstruction

keeping a very clear position on bers decreased to 282 and only the fact that EU money for recon- kept dropping from there. The struction in Syria will only come 120-day refugee ban officially once the political process will be expired in October 2017, but was firmly underway under UN aus- replaced with a pause on refu- pices’, Mogherini said, speaking gee admissions for 11 countries, to the press in Vienna (EEAS, including Iraq, Libya, Syria, and 2018). Yemen, while the US government In the last years of US Presi- reviewed vetting procedures. Admissions for Syria restarted dent Barack Obama’s administra- at the end of January 2018, and tion, the United States resettled nine refugees have been admit- tens of thousands of Syrian refu- ted since then (Refugee Process- gees, as well as resettling many ing Center, 2019). refugees from elsewhere: about 70,000 a year from 2013 to 2015, These restrictions on refugee and 85,000 a year in 2016, of admissions laid down by the which 1,682 were Syrian refugees Trump administration, which in 2014, 12,587 in 2015 and 6,557 has set the lowest refugee target in 2016 (Refugee Processing Cen- in modern history with no more ter, 2019). When Donald Trump than 45,000, has also been com- plemented with vocal expres- took office in January 2017, that sions of little support for large- number plummeted, partly due scale US economic assistance to to the 120-day ‘refugee ban’ that rebuilding MENA states. prevented nearly any refugees from being brought into the United States over the summer of 2017, and partly because of 3.3. Review of the literature specific scrutiny facing refugees on political settlements from several countries, including Syria, Iraq, Libya and Yemen. In- In this section, we review the deed, there was the original trav- most recent literature on politi- el ban, which banned all refugees cal settlements and elite pacts, and all arrivals from Syria. Even and how it helps to understand after that ban was put on hold in the complex issue of group-based court, refugee admissions were identity, inclusivity in post-con- stop-and-go for months. flict political processes, diversity In December 2016, the last and pluralism within societies in full month of the Obama admin- conflict. In the absence of many istration, 1,318 Syrian refugees examples of political settlements were admitted to the United in the MENA region, this sec- States (Refugee Processing Cen- tion examines the cases of Alge- ter, 2019), whereas in February ria and Lebanon, both of which 2017, the first full month of the underwent processes of national Trump administration, the num- reconciliation after devastating Chapter 3 | 75

civil wars, in an attempt to draw istration, security, and the rule of a typology of political settlements law. According to these authors, applicable to the four case studies. a political settlement is not a sin- There is a large body of re- gle event but rather a continuing search literature that has used political process. Furthermore, the term ‘political settlement’ settlements can adapt to accom- throughout contemporary histo- modate changes in the relation- ry. It first appeared in the works ship between states and societies of the sociologist Joseph Melling that arise due to different needs (1991) in his study of industrial and emerging demands. capitalism and the welfare state The idea of a political settle- in England. In spite of no precise ment as a negotiated agreement definition of the concept being between states and societies is used, Melling refers to the term rather abstract in its construction, to describe the political process- and it is difficult to identify any es that shaped the emergence of solid empirical or historical cases social welfare policies. An early where such an agreement, or ‘so- reference to political settlements cial contract’ between the state within development studies can and society, can be shown to have be found in Khan (1995), where taken place (Laws, 2012). More- the author analyses the perfor- over, there is an implicit assump- mance of institutions by means tion that an agreement between of the ‘inherited balance of power state and society features a notion or political settlement’. The con- of a unified society, when societ- cept was later defined by Khan as ies tend to be diverse and plural the ‘distribution of organization- with different social, economic, al and political power between and political groupings. competing groups and classes’. In fact, political settlements The concept of political settle- differ from peace-building mod- ment did not gain significant at- els, in the way that the latter need tention in development studies to reflect the real balance of pow- until it was later used by Fritz er to avoid falling back into con- and Menocal (2007). These au- flict. In post-conflict situations, thors disagree with the defini- there is often a trade-off between tion used by Khan and offer an stability and elite buy-in on the interpretation of settlements as a one hand and socially inclusive ‘negotiated agreement (at least in settlements on the other (Plonski principle) binding together state and Walton, 2018). and society and providing the The latter also ignores the necessary legitimacy for those reality that not all groups share who govern over those who are a common relationship with or ruled’. The political settlement have comparable access to the establishes the foundation for ba- state, nor do they necessarily sic state functions: public admin- enjoy comparable civil, political 76 | Repatriation: scenarios for conflict resolution and reconstruction

or economic rights (Laws, 2012). of the political settlement from Distinct ethnic, religious, social, which they are excluded, and its and gender groups within a so- distributional implications. For ciety might be excluded to dif- example, Tadros (2012) asserts ferent degrees from participation that politically and economically in formal institutions of the state. excluded Egyptian youth were Hence, these authors believe that instrumental in eroding the po- it is not so obvious that, unlike in litical settlement under President the opinion of Khan, all societies Hosni Mubarak in 2011, in spite can be conceived of as single en- of their absence of elite repre- tities, and therefore, be party to sentation. Arguably, there is a a negotiated agreement with the common understanding that the state. more exclusive and fractured a Considering the fractured na- society might be, the more dif- ture of all societies, particularly ficult it will be to forge inclusive in developing countries, the in- and long-lasting political settle- ternational community, in the ments (Parks and Cole, 2010), process of supporting or broker- prompting an end to political ing political settlements, favors conflicts and civil outbreaks. In inclusive and power-sharing this regard, the region discussed processes, combined with politi- here is commonly characterized cal participation of previously ex- by these features. cluded social groups by engaging To conclude, most of the with pro-development factions above-mentioned authors as- and coalitions (Parks and Cole, sert that political settlements (a) 2010). Therefore, a comprehensive should not be defined solely as a definition of political settlements contract between states and soci- has to take account of the divisive eties; (b) they involve horizontal nature of societies, particularly negotiations between contending those that are developing. elites, and vertical relationships This concept of inclusiveness between elites and their follow- constitutes a significant axiom of ers; (c) they should not necessar- the literature on political settle- ily be defined as a reflection of a ments (Whaites, 2008; Barnes, common understanding between 2009; Parks and Cole, 2010), even elites; (d) they influence the form, though Parks and Cole (2010) nature and performance of insti- suggest that in some cases, the tutions; and (e) they can be more political exclusion of a particular or less inclusive. group in a society can be a key The research literature is full factor in explaining the stability of well documented examples of a particular settlement, at least of political settlements that have for a certain period of time. When been forged throughout recent excluded groups have sufficient history in different regions of the numbers and resources, they world, mainly in sub-Saharan Af- are likely to challenge the terms rica, South East Asia and Latin Chapter 3 | 77

America. But when assessing from 1997 onwards did the peace- the reality in the MENA region, building process take a successful examples are scarce, despite the direction, as the Islamist insur- conspicuous evidence – past and gents moved to declare a unilat- present – of widespread political eral ceasefire (Ghanem-Yazbeck, turmoil and conflict. 2018). Major initiatives such as the Sant Egidio peace process in 3.3.1 Political settlements in the 1995 in Rome failed due to grow- MENA region ing polarization between the This chapter could offer a few two extremes in the dispute: the analyses of how political settle- regime and the Islamists. In any ments have been forged in some case, the peace-building process MENA countries. Two cases, one facilitated the reintegration into from the Maghreb and the other society of around 15,000 fight- from the Levant region, will be ers (Ghanem-Yazbeck, 2018). In featured: Algeria and Lebanon. addition, the reconciliation mea- Particularly noteworthy is the sures eased tensions in society so absence of any mechanisms for that Algerians could return to a transitional justice in most of the relative normalcy. examples in the MENA region. In spite of a political settle- There are no historical precedents ment that contributed to ending in the region that entail meaning- the conflict and returning the ful transitional justice with the country to peace and stability af- exception of Tunisia, where the ter a long decade of armed con- parliament passed a dedicated flict, the regime under the presi- law in December 2013, and to dency of Bouteflika neglected the some degree Morocco. In the ab- war victims, thus preventing a sence of historical reckoning for process of bringing to justice var- the perpetrated actions during an ious parties involved in atrocities armed conflict, only impunity re- and severe human rights viola- mains, with the consequence that tions (Martínez, 2004), beyond lasting grievances and extremist charters and decrees, all of them options might prevail. issued by state institutions. That was the case of the Civil Concord Law, which was adopted in 1999 3.3.1.1 Algeria after a referendum voted favor- Algeria endured a bloody ably by 98,6%, and which granted civil war from 1991 to 2001 with partial amnesty to former jihad- an estimated 150,000 casual- ists under the condition of arms ties, 7,000 disappeared, and $20 dropping and no previous in- billion of damages, according volvement in major massacres, in to the World Bank. From 1994, which case the law was offering there were unsuccessful dialogue reduced jail sentences. Likewise, rounds between the regime and in 2005, a new law, the Charter the Islamist insurgents, but only for Peace and National Reconcili- 78 | Repatriation: scenarios for conflict resolution and reconstruction

ation exempted members of the massacres of the civilian popula- security forces and pro-govern- tion from late 1996 onwards led ment militias from prosecution to the first massive exodus of (Ghanem-Yazbeck, 2018). Simi- people towards the cities, where larly, it gave full immunity to thousands were forced to move members of the security forces, to the outskirts during summer while it also established compen- and autumn 1997 (Norwegian sation for the victims of the war, Refugee Council, 2006). In this including relatives of abducted regard, there is no publicly avail- victims and dead jihadists. able overview of the number The political settlement that of displaced. According to one Algeria brokered during the last newspaper, President Bouteflika years (and in the aftermath) of said on one occasion that 1.5 mil- the armed conflict was credible lion people had been displaced but incomplete. A strong state (Le Figaro, 2004). But this in- grip by a group of elites on eco- formation seems to contradict a nomic resources (mainly oil) and statement given a few days later political power created a clien- by the minister of the interior, telist structure (Werenfels, 2009). stating that only 500,000 people The induced amnesia about the were internally displaced during root causes of the violence exert- the conflict (El Watan, 2004). ed by the different parties in con- A great effort was necessary flict has neutralized any attempt to revitalize the agricultural sec- at transitional justice during the tor in order to secure a livelihood reconciliation process, which has for the returning population. As ended up producing grievances part of the post-conflict resolu- that persist to this day and which tion package, the government have eroded possibilities for fur- launched a rehabilitation pro- ther democratization. In addi- gram in rural areas to encourage tion, from the political perspec- internally displaced people and tive, such an experience, which migrants to return to their vil- was so much distanced from an lages. The authorities tried, for ideal template, has prompted the example, to repopulate villages activity of extremist movements, by promising direct financial as- such as Al Qaeda, in the Islamic sistance to returnees as well as Maghreb within the territory of implementing programs to reha- Algeria (Abadeer et al, 2018), bilitate houses, increase employ- while reducing the effectiveness ment, and revitalize the agricul- of Algeria’s moderate Islamist tural sector, which suffered dur- parties. ing the 1990s. Nevertheless, as the Algeria’s civil war triggered a procedures for the rehabilitation huge flow of internally displaced program were often delayed due population. Indeed, while fami- to administrative mismanage- lies fled insecurity during the ment, the displaced population first years of the Algerian crisis, was discouraged from returning Chapter 3 | 79

to their home villages because ties, and it called for national living conditions in the country- reconciliation and for ‘spreading side were even harder than in the the sovereignty of the State of towns, with no drinkable water, Lebanon over all Lebanese ter- poor general infrastructure, and a ritory’ through a one-year plan lack of health facilities (El Watan, that included the ‘disbanding of 2006). Overall, there is no collated all Lebanese and non-Lebanese information on the number of re- militias’ (Taif Agreement, 1989). turns, apart from a reference in a In this regard, the agreement was speech given by President Boute- decisive in determining the Arab flika in April 2004, where he said identity of Lebanon, emphasizing that 700,000 of 1.5 million inter- that Lebanon was an indepen- nally displaced had returned (Le dent, sovereign, and free coun- Figaro, 2004). try. It also confirmed the territorial integrity of the country and defined its political system as a 3.3.1.2 Lebanon parliamentary democracy, based Lebanon experienced a tragic on the principles of separation, and multifaceted war from 1975 balance, and cooperation among until 1990, which resulted in the the various branches of govern- deaths of an estimated 120,000 ment. It also anchored the coun- people. Prior to the beginning of try’s socio-economic system as a the conflict, various internal ten- free economy favoring individual sions inherent to the Lebanese initiative and the right to private system and multiple regional property. Finally, it advocated developments contributed to the the abolition of political sectari- breakdown of governmental au- anism as a basic national goal to thority and the outbreak of civil be achieved according to a grad- strife in 1975 (Khalidi, 1979). The ual scheme (Krayem, 1997). cause of the Lebanese civil war Related to this last point, it is was multiple, but neither exclu- worth noting that political con- sively for endogenous nor exclu- fessionalism has been considered sively for exogenous factors, but since the independence of Leba- a combination of both. Precisely, non a temporary arrangement the war ended at a specific his- that should be phased out as soon torical moment when movements as possible, but it has prevailed towards internal reconciliation up to the present time. Although coincided with favorable regional the Taif Agreement reaffirms the and international developments. gradual abolition of the confes- A political settlement en- sional system, it still dominates shrined in a Document of Na- institutional and political life in tional Accord (known as the Lebanon (Krayem, 1997). This Taif Agreement) was adopted was translated into a power- in October 1989 by members of sharing arrangement whereby Lebanon’s Chamber of Depu- the president of the country is a 80 | Repatriation: scenarios for conflict resolution and reconstruction

Maronite Christian, the prime gram, which was funded by a minister is a Sunni Muslim, and combination of grants, soft loans, the speaker of the house is a Shi’a and bilateral loans from western Muslim. The parliament is still countries, Gulf states as well as based on confessional quotas international organizations, aid that cannot be revised. In fact, agencies, and commercial banks to avoid going back to conflict (Najem, 1998). The vast recon- of power, a census has not been struction project carried out after held in the country since 1932. the adoption of the Taif Agree- Arguably, one might conclude ment focused on downtown Bei- that in reality the elimination of rut by means of building a new confessionalism has been rel- socio-economic and political or- egated to a rather distant future, der organized around the luxury which leaves the door open to the real estate and service industry renewal of conflict, and increases (Sharp, 2018). the possibilities of its occurrence. In spite of disagreements This has led to a clientelist rela- among many Lebanese confes- tionship between the elite and the sional communities over political people: dignity and protection issues that led to the Taif Agree- in exchange of loyalty (Hermez, ment, the single issue where 2011). agreement existed was that Pal- Such a situation does not al- estinian refugees should not be low for the revival of a stable po- ‘naturalized’, locally referred to litical system, one that is urgently as ‘implanted’ in Lebanon (Azat needed to revitalize the economy and Mullet, 2002). The rejection and develop further public and of permanent resettlement in private institutions. In addition, Lebanon was inscribed into the other issues not properly ad- Agreement, which states that dressed by the political settle- ‘there shall be no fragmentation, ment were the presence of Syrian no partition, and no repatriation forces on Lebanese soil, Hizbol- (of Palestinians in Lebanon)’ (Taif lah’s status as an armed militia Agreement, 1989). Indeed, Leba- outside the state, as well as the non’s Palestinian refugee prob- Palestinian refugee issue, as we lem predates the civil war and will briefly assess in this section. since their forced exile from Pal- The political settlement in estine more than 60 years ago, the Lebanon came along with an majority of the refugees (and their outstanding effort to transform descendents) have been living in a war economy into a peace squalid shelters and cramped ref- economy after almost 15 years ugee camps. The 350,000 Pales- of devastating conflicts. In 1993, tinian refugees in Lebanon form a the first Rafik Hariri’s govern- disenfranchised minority, suffer- ment launched ‘Horizon 2000’, ing from economic marginaliza- a $20 billion reconstruction pro- tion and fearing being excluded from future peace settlements, Chapter 3 | 81

which would leave them exiled 3.4 Typology of political in Lebanon. Unlike other Arab settlement scenarios countries, such as Jordan and Syria, which also serve as hosts In this part of the chapter, we to Palestinian refugees, Lebanon look at the different scenarios of did not grant the Palestinians the political settlements by build- civic rights and only a few were ing a three-dimensional (3-D) granted work permits. Currently, typology addressing the relation- Lebanon has the highest percent- ships between three complex post- age of camp-dwelling refugees of war or conflict resolution issues: all the countries hosting Palestin- ian refugees. 1. The type of political settlement that could be set up. As the Lebanese civil war did not produce any form of tran- 2. How the reconstruction plans sitional justice, in spite of the have responded to this settle- conflict ending 29 years ago, the ment, particularly the inter- question of civil war memory is national community and the still very vivid for many Lebanese donors. who continue debating the root 3. And how this will affect (or causes and consequences of the is affecting) repatriation (or war. The assassination of Prime return) of the refugees or the Minister Rafik Hariri in 2005, the displaced to their lands. following war between Hizbollah Setting up a framework that and Israel in 2006, the spillovers includes a measure of ‘political into Lebanese territory of the Syr- settlement scenarios’ can hold ian war from 2011 onwards, and many dimensions. In this section, the continued political instability we look at the scenario-building that often hits the country have approach by Kelsall (2016), who only added to the notion among has combined three approaches many of its citizens that political in one: violence is chronic and structural. First, the approach of Bruce As time goes by, many citizens Jones et al (2012), in which the de- feel that addressing challenges gree of horizontal elite inclusion such as the situation created by in the political settlement is a ma- the wave of Syrian refugees since jor determinant of state fragility. 2011 is more urgent than dwell- Second, the approach of Da- ing on and dealing with the past vid Booth (2015), who sets the civil conflict. The only exception development pathway depend- to this pattern was the establishing on the ruling elites’ choice to ment of the Special Tribunal for include rival elites into a politi- Lebanon, which was created to cal settlement by offering them judge the criminals that killed 22 spoils distribution of ‘purposive people, including former Prime coordination’ around a common Minister Rafik Hariri in 2005 vision or goal. (Haugbolle, 2018). 82 | Repatriation: scenarios for conflict resolution and reconstruction

Third, the approach of Brian cal settlement. Arguably, the more Levy (2012) and Levy and Walton inclusive the settlement is, the (2013), who differentiate the dif- more stable the country will be. ferent types of settlements based Inclusiveness can be broad- on types of institutions, especial- ened beyond political parties or ly bureaucratic and cultural. powers to embed social inclusion. This 3-D relationship can cre- The powers of excluded social ate a typology of different type groups, such as youth, women, of political settlements depend- sectoral workers (such as farm- ing on the degree of progress in ers, fishermen, etc.), and certain each of these criteria. The way to territories in destabilizing a state assess these three criteria is to an- should not be underestimated. swer three diagnostic questions The Arab uprisings of 2011 were to see where a country is placed initiated by these social groups, within this 3-D cube. rather than by political parties. This means that the less inclusive 3.4.1 How inclusive or exclusive a political settlement is, the more is a political settlement? violence can be generated, which Inclusiveness is a measure of a could also lead to renewed col- country’s stability within a politi- lapse of the state.

Figure 3.1 How inclusive the political settlement is?

Most inclusive • Majority of Inclusive elites ac- Neutral cept the settlement Exclusive

Most exclusive • Minority of elites accept the settlement

(Source: Authors’ analysis, inspired by Kelsall, 2016)

At the negative extreme of At the positive extreme of ‘most ‘most exclusive’, there is a politi- inclusive’, the majority of elites cal settlement that is based on a accept the settlement, which en- minority of elites accepting the sures stability, sustainability, and settlement, leading to instabil- refraining from the use of vio- ity and potential use of violence. lence. Chapter 3 | 83

3.4.2 What motivates elites to ac- and real purposive coordina- cept a political settlement? tion to serve a common purpose The second criteria relate to with a real will to rebuild and the quality of this inclusion/ex- achieve inclusive development, clusion in the political settlement, or whether it was based on a buy- and its ability to supply public in process to achieve a shallow and collective goods and servic- inclusion, which would create es. It looks into how the inclusion economic rents (power, material was carried out: whether it was gains, etc.) or spoils. carried out on an understanding Figure 3.2 What motivates elites to accept political settlement?

Mostly coordination Some coordi- • real purpo- nation Neutral sive coordination Some spoils

Majority on spoils • buy-in-process with shallow inclusion

(Source: Authors’ analysis, inspired by Kelsall, 2016)

At the negative extreme, ‘ma- the political elite provides ser- jority of spoils’ means that a politi- vices and protection, or resourc- cal settlement is based on spoils or es more generally, in exchange gains rather than a common agree- for loyalty of the constituents ment. At the positive extreme, (Eisenstadt and Roniger, 1980), ‘most coordination’ entails an in- a patron-client type of relation- clusive political settlement that is ship. Clearly the less personal the based on a common purpose. governance, the more the ability to supply public goods and ser- 4.4.3 By what norms is bureau- vices. Interestingly, according cracy governed? to some research (Khan, 2010) This question attempts to modern societies co-exist with measure the type of bureaucracy clientelism, which could have and whether it is based on per- a different variation and mixes sonalized links – that is, relying of distribution of power. More- on patronage, nepotism, and over, in some cases and follow- personal relationships – or that it ing some evolutions, rather than involves meritocratic rules. Cli- dying out, this patron-client rela- entelism has traditionally been tionship can evolve into different defined as a system in which varieties of forms. 84 | Repatriation: scenarios for conflict resolution and reconstruction

Figure 3.3 By what norm bureaucracy is governed?

Mostly impersonal Somewhat • governance impersonal Neutral is based on meritocrat- Moderate ic rules clientelism Extreme clientelism, nepotism, and patronage governance is based on personal relations

(Source: Authors’ analysis, inspired by Kelsall, 2016)

At the negative extreme of settlement has a high probability ‘extreme clientelism’, the politi- of surviving and is capable of in- cal settlement is based on client- ducing peace and development. patron relations between the This type is difficult to achieve state and the people. At the posi- in practice: the authors give the tive extreme, the political settle- example of a post-war settlement ment is based on an impersonal in South Korea, and also a close relationship, where all goods and match could be Mauritius. services are delivered to all. Based on these three questions, we can trace a 3-D cube of the type of political settlement, as we draw lines to the answers to these questions. An ideal scenario would be: an inclusive, coordinated, and impersonal settlement. In this type of ideal scenario, the power is balanced and shared between the most important elites. All This type of stable scenario parties included are coordinated will encourage the participation around one common goal to re- of the international community build the state, and the bureau- and donors in reconstruction, be- cracy between them is impersonal ing more congruent with regards and meritocratic (based on mer- to the stability, transparency, and its). Consequently, the political resilience of the settlement. Chapter 3 | 85

The worst-case scenario: ex- The middle scenario is be- clusive, spoils-driven, and per- tween these two previous sce- sonalized settlement narios. This could be the most Clearly this type of settlement common scenario in practice. is the worst, and where the minor- This type of settlement is set up ity of elites are the ones involved by the most powerful, excluding in state-building, probably all other type of elites. Within these linked together with buy-ins and elites, some are motivated by the spoils of materialistic and power gains they will have while others gains. They are governed by a are still coordinated under a com- system that is based on personal mon vision of state-building and relationships and clientelism. seeking development. The bu- This settlement is quite fragile as reaucracy is a mix of patronage it is constantly threatened by con- but some civil servants are actu- flicts within the settlement itself ally impersonal and care for the (as their relationship is based on public. Some examples given of cronyism) and by conflicts from this type of settlement are Ugan- outside the settlement of other da, Kenya, Indonesia, and Cam- elites that have not been includ- bodia. This type of settlement can ed or more broadly the excluded generate some development but layers of societies. This type of it needs to have a mix of portfo- settlement will make it difficult lios to engage the different layers to act for the overall benefit of of society that are excluded. the state. This of course will affect the development of the state and will make support from the international community for reconstruction quite limited due to its fragility. The author gives the example of the Democratic Re- public of Congo as a country with this type of post-conflict political settlement.

Other types of settlements can be drawn in this 3-D relationship with settlements that are relatively inclusive but where the bureaucracy is affected by personal relationships and cronyism (for example, Cote d’Ivoire and Malawi). Another type can be 86 | Repatriation: scenarios for conflict resolution and reconstruction

dominated by the exclusive elites, analysis above, it would be inter- where they are coordinated by a esting to see where they stand in common vision but their bureau- terms of their type of political set- cracy is personalized. tlements and how this is reflected in real terms in their restructur- ing and the repatriation of their 3.5 Applying a political refugees. settlement typology to The table 3.1 answers the historic conflicts in some three questions related to this ap- MENA countries proach.

Looking closely at some MENA countries and given the

Table 3.1. Political settlement, reconstruction and refugees in Algeria and Lebanon

Algeria Lebanon

1. How inclusive/ex- More towards an exclusive Post-war political settlement clusive is the politi- political settlement: strong in 1989 is leaning towards an cal settlement? state grip on economic re- inclusive model: majority of sources and political power elites. The confessionalism is by a group of elites (tracing a temporary settlement that legitimacy to the revolu- seems to be perpetuated to tion) with weak institutions avoid returning to war. An (Werenfels, 2009). This has arrangement that is currently led to tension (political and being contested even by the social) and rise in violence. Lebanese people.

2. What motivates Political settlement is based Inclusion of elite is based on elites to accept the on powers of elites and ‘spoils of truce’. A fragile political settlement powerful lobbies that benefit relationship that often leads from the system. The settle- to tension among elites. ment is not inclusive and hence fragile.

3. By what norms is The political settlement is The political settlement cre- the bureaucracy based on a clientelist struc- ated a state that is based on governed? ture. Caught in relations of clientelism (patron-client ‘reciprocal predation’ be- relationship): loyalty to the tween state and society that elite in exchange to protec- are expressed in the reper- tion and dignity. This re- tory of political clientelism, lationship was mixed with the Algerian deputies thus corruption, leading to a form become agents of clientelistic of patronage to clientelistic mediation. brokerage (Hemez, 2011). Chapter 3 | 87

Type of political settle- The political settlement This fragile political settlement adopted had ended the ment leads to frequent ten- civil war and was considered sion among elite groups and ‘credible’. But the root causes their followers. It comes as of the conflict have not been no surprise that Lebanon addressed, risking a resur- often suffers from political gence in the conflict. deadlocks, growing instability, potentially leading to a state division.

Impact of reconstruc- This fragile political settle- The plan focused on re- tion plans ment led to bribery, cor- building the capital (luxury ruption nepotism and cli- real estate) and focused on entelism. Protestors went emerging services industry, on the streets demanding leaving behind the rest of the change in the regime and its territories. cronies and ending corruption. (Hirak, Feb 2019)

Impact on repatriation Uncounted thousands of Many Lebanese fled the of refugees Algeria fled following the country during the war violence and tension in the (about one million nationals), country (mostly to Europe but did not take the status but many without finding of- of refugees. They have no ficial asylum). Some negotia- intention of returning given tions are happening for the the political instability in the repatriation of these refugees country and its neighbors. (with Germany, for example) but so far no specific plans are in place.

Source: Authors’ analysis

3.5.1 Implications for current settlement scenarios; their conse- post-conflict countries quences for reconstruction plans; It is not easy to describe po- and the implications of these two tential settlement scenarios for for the repatriation of refugees. post-conflict Arab countries As discussed above, while we drawing on previous experiences highlight three different scenari- in the MENA region. os: best, middle, and worst, there The following table 3.2 is an is a wide spectrum of possibili- attempt to put together the three ties that exists in between these approaches and apply them to scenarios. The main driving fac- the case of Syria: its political tor is the type of political settlement, which will determine the 88 | Repatriation: scenarios for conflict resolution and reconstruction

different trajectories of the state’s reconstruction plans and the re- resilience and development, and patriation scenarios (Di John and on which will depend both the Putzel, 2009).

Table 3.2. Scenarios for Syria: Political settlement, reconstruction and refugees

Political settlement Reconstruction plan Implications for refugees: repatriation, resettlement enabling /inhibiting repatriation and sustainability of return Ideal scenario: Involving the Support and encourage • Inclusive and power- international return sharing communities: Yes These include economic • Coordinated - UN estimate reconstruction (rebuilding • Impersonal bureaucratic reconstruction cost of $250 infrastructure, access to governance billion health and education, Result: sustainable political - The EU will fully etc.) and political reform, settlement participate once a such as security, property settlement process (under rights, and access to justice the UN) is in place.76 (Yahya, 2017). Of total refugees, 80% return rate, using UN (2017) population statistics. The UNHCR (2018) survey (mid-2018) also found that 76% of refugees would want to return to Syria one day.

Middle Scenario A substantial Of total refugees, 45% • Partial political settlement, reconstruction program of return rate based on the where majority of elite are $140 billion (average of the ratio of reconstruction included but some are left high and low scenarios), funds in the first 10 years out. spread evenly over a 20- (moderate versus high • Elites included are partially year period. scenario). included on the basis of spoils • Governance and the relationship between the state and the population are partially on some kind of clientelism, although some political freedoms are given.

76 EU High Representative for Foreign Affairs and Security Policy Federica Mogherini said at an EU side event at the UN General Assembly in 2018: ‘ we discussed ways in which we can increase even more our humanitarian assistance for Syrians, including inside Syria, while still keeping a very clear position on the fact that European Union money for reconstruction in Syria will only come once the political process will be firmly underway under UN auspices.’ (ERF, 2019). Chapter 3 | 89

Worst-case scenario Limited reconstruction Of total refugees, 19% • Political settlements, program of $30 billion, return rate based on the excluding the different largely relying on China, ratio of reconstruction groups and/or go back to Iran and Russia, and funds in the first 10 years an ante settlement with poor spread evenly over a five- (low versus high scenario). balance of power. year period. • The elites are connected by their own interests and there is a common national purpose. • Clientelism is the main feature between the state and the people imposing a fragile state and with limited guarantees for micro- security, property rights, and human rights.

In addition to these, while Also, if we are focusing our political settlement is an internal analysis here on the case of Syria, affair, it needs to take account of the political settlement scenarios the borders and not only focus need to involve a peace-making on the state framework (Plonski process. This involves a large et al, 2018). number of issues that are summarized as follows:

Table 3.3. Peace-Making Process requirements

Source: Barnes (2009) 90 | Repatriation: scenarios for conflict resolution and reconstruction

3.6 The nexus between to be the world’s least peace- political settlements and ful region since it is home to 10 returned refugees in South of the least peaceful countries in Med conflict-affected the world. In addition, no country in the region ranked more countries than 40th. Iraq, Syria, and Yemen witnessed some improvements Regarding conflict severity, in their scores, with Iraq the best the Global Peace Index results for improver (Figure 3.4). 2018 consider the MENA region

Figure 3.4: Global Peace Index, 2010-2018

0 Iraq Libya Syria Yemen 2010 2014 2016 2018

Source: Institute for Economics and Peace Note: The GPI measures a country’s level of Negative Peace using three domains of peacefulness: Ongoing Do- mestic and International Conflict, Societal Safety and Security and Militarization. The GPI comprises 23 indicators of the absence of violence or fear of violence. All scores for each indicator are normalized on a scale of 1-5,

Political settlements are some- provides support in mediation. times associated with events such But it does not provide any sort of as the signing of peace agree- assessment of peace agreements. ments and hence, we will take a Second, the Political Settle- closer look at peace agreements, ments Research Program (PRSP), as a proxy for political settle- led by the University of Edin- ments. To that effect, we rely burgh, has developed the Peace on two databases. First, the UN Agreements Database (PA-X), Peacemaker Database takes stock which, in contrast to the UN da- of existing peace agreements by tabase, codes and maps all peace country, conflict type, and date. agreements since 1990 till the Annex 3.1 provides an extract present (around 1800 documents from this database on all existing in 100 countries). We under- peace agreements in this database stand that the implementation of post-2011 for Libya, Syria, and these political agreements mat- Yemen. This database focuses on ters more than the agreements the agreements texts, and also themselves. But given the fact Chapter 3 | 91

that it is complex to measure the ly, this can have implications for outcomes of political events still the growth-stability trade-off. In in progress, the PA-X database case there is incompatibility be- instead evaluates the outcomes tween institutions and interests of peace agreements. The latter of powerful groups, growth-en- usually represent the result of hancing institutions might want political settlement events. It also to consider the transaction cost allows a comparison of all ele- of resistance and hence make a ments of inclusion, development, trade-off between growth and and rights, using quantitatively stability (Khan, 2010). The nature defendable category definitions of political settlement according- for power-sharing, women, vic- ly shapes the pattern of economic tims, socio-economic rights, and growth and transformation (Osei development (Schünemann and et al, 2017). Lucey, 2015). In terms of governance and socio-economic reconstruction, 3.6.1 Governance and political a closer look at the peace agree- settlements ments in our countries of inter- In terms of governance, Khan est reveals that only three agree- (2010) suggests that the linkages ments among the 82 agreements of political settlements with gov- in Syria since the uprisings in- ernance depend on two factors. cluded a mention of mechanisms First, policies and institutions reforming or establishing new that worked in some advanced political institutions, whereas 16 countries do not necessarily agreements included provisions achieve the same successful re- specifically addressing socio-eco- sults in other developing coun- nomic reconstruction or develop- tries. Second, different policies ment in general terms (Annex 3.2, and institutions are effective in Tables 1 and 2). solving different problems in different contexts. This suggests that 3.6.2 Patterns of return of refu- it is crucial to identify the design gees of governance reform since it will Conflicts in South Med coun- accordingly shape the necessary tries have triggered a refugee cri- growth-enhancing institutions sis. A closer look at the UNHCR according to the context. statistics provides us with some Developing countries differ insights into the repatriation pro- from advanced countries in their cess of refugees. For example, political settlements in that they Syrian refugees are concentrated can have informal institutions mainly in Turkey, Lebanon, and playing a significant role. This Jordan where the share of Syrian can make political settlements in refugees in these countries out these countries clientelist, where of total Syrian refugees in 2018 these informal institutions affect reached 54.4%, 14.2%, and 10.2% the power exercise. Consequent- respectively (Figure 3.5). 92 | Repatriation: scenarios for conflict resolution and reconstruction

Figure 3.5: Countries of asylum or residence for Syrian refugees, 2011-2018

100% 7000

6654.3 s 6308.7 6000 80% total 5000

of 5524.4

4873.2 refugee % 4000 by country an n

s 60% 3887.5 3000 thousands) Syri 40% 2000 2468.3 (in 1000 Total Syrian refugees) 20% 729.0

residence (i 0% 19.9 0 of

Syrian refugee 2011 2012 2013 2014 2015 2016 2017 2018

Turkey Lebanon Jordan Germany Iraq Egypt Sweden Austria Netherlands Others Total Syrian Refugees

Source: UNHCR data We are particularly inter- turned refugees in our four coun- ested in this chapter in the re- tries of interest represent a very turned refugees. The UNHCR small portion out of yearly total differentiates between returned refugees over the period 2012 to internally displaced persons and 2017 (Table 3.4). This share also returned refugees.77 In simple seems to be declining for the four terms, returned refugees are countries over the period since refugees who returned to their the Arab uprisings with some re- place of origin and we will even- covery in Syria in 2017. tually focus on them. The re-

Table 3.4: Returned refugees by origin as a percentage of total refugees by country of origin, 2012-2017

(Percentage of total refugees 2012 2013 2014 2015 2016 2017 by country of origin) Iraq 11.0 15.2 2.9 2.2 0.0 0.2 Libya 20.1 0.0 0.0 0.0 0.0 0.0 Syria 9.4 5.7 0.0 0.0 0.0 1.2 Yemen 0.0 0.2 0.0 0.0 0.0 0.0

Source: UNHCR data

77 UNHCR defines returned refugees as ‘Former refugees who have returned to their country of origin, either spontaneously or in an organized fashion, but are yet to be fully integrated. Such returns would normally take place only under conditions of safety and dignity’. As for returned internally displaced persons, UNHCR defines them as ‘Internally displaced persons are persons or groups of persons who have been forced to leave their home or place of habitual residence, in particular as a result of or in order to avoid the effects of armed conflict, situations of generalized violence, violations of human rights, or natural or man-made disasters, and who have not crossed an international border. Returned internally displaced persons refers to those internally displaced persons who were beneficiaries of UNHCR’s protection and assistance activities, and who returned to their area of origin or habitual residence’. Chapter 3 | 93

3.6.3 Implications from history: total number of peace agreements linkages between peace agree- where there is a mention of refu- ments and returned refugees gees (either merely mentioned, The PA-X database evaluates or the agreement contains provi- whether there is a reference to sions, or deals with it in a substan- refugees or displaced persons, in- tive way) and the average share cluding repatriation, that appears of returned refugees out of total in a peace agreement.78 Despite be- refugees over the period 2012 to ing a critical problem for the con- 2017 for all countries included flict in Syria, only 12 among the 84 in the PA-X database. It seems agreements mentioned refugees that the returned refugees share with different degrees of details, is only weakly positively associ- and only one of them deals with ated with the number of peace the refugees’ problem in a sub- agreements with a mention of stantive way (see Annex 3.3, Table refugees (Figure 3.6), suggesting 1). For Libya, the PA-X database that other considerations are crit- identifies 39 peace agreements ical for triggering a meaningful since the uprisings: 14 of them response in terms of repatriation. mention refugees (see Annex 3.3, The figure also shows that two Table 2). The case for Yemen is of our countries of interest, Syria similar: four agreements mention and Yemen, are below the aver- refugees out of a total of 38 agree- age trend line. This indicates that ments (Annex 3.3, Table 3). the mere mention of refugees in It is important to assess the as- peace agreement initiatives were sociation between these clauses not credible enough to trigger on refugees in peace agreements the required response in terms of and the return of refugees on return of refugees in comparison the ground. Figure 3.6 describes with their peer countries. the correlation between the

Figure 3.6: Correlation between the share of returned refugees and the total number of peace agreements with a mention of refugees, 2012-2017

18% R = 0.0151 16% Angola 14% Dem. Rep. of the 12% Afghanistan Congo 10% Rwanda Burundi 8% Kenya 6% Mali Central African Rep. Colombia 4% Russian Federation Sudan Syrian Arab Rep. a mention to refugees Ukraine Uganda Palestinian 2% South Sudan Zimbabwe Myanmar

total number of agreements with 0% Yemen 0 Ethiopia 2 4 6 8 10 12 Average share of returned refugees out of total refugees, 20122017-

78 The provisions specifically addressing Source: UNHCR data and PA-X database refugees’ issues, such as return, and land claims are included in the positive coding of this variable. Further, it includes references to internally displaced persons or displaced persons. 94 | Repatriation: scenarios for conflict resolution and reconstruction

A closer look at Syria reveals ered in 2017 after a few agree- that the returned refugees trend ments were being discussed in is substantially influenced by 2015. But, the number remains other factors, and that even when much lower than in 2013, which there is a response to the number did not see any agreement as the of peace agreements, it tends to war was still very much at an happen with a lag. For example, early stage of the Syrian conflict the number of repatriates recov- (Figure 3.7).

Figure 3.7: Returned refugees and the number of peace agreements with a mention of refugees in Syria, 2012-2017

160000 5 140761 140000 4 120000 4 3 100000 77212 3 80000 68573 2 60000 2 40000 1 20000 0 189 1 1 0 0 2012 2013 2014 2015 2016 Peace 2017

Returned refugees agreements with a mention to refugees

Source: UNHCR data and PA-X database

3.6.4 Correlation analysis shed light on the robustness of In this section, we undertake peace agreements. From a policy a correlation exercise that aims perspective, this will serve as to assess the association between a sort of inference from history the share of returned refugees out so as to predict what is likely to happen in the case of our coun- of total refugees on the one hand, tries of interest. and some aspects of peace agreements on the other, notably men- For the peace agreements dition of refugees, socio-economic mensions mentioned above, we reconstruction, governance as- take into consideration the following variables: pects, the extent of democracy, and human rights. In particular, • Refugees (GRef): this variable we analyze this correlation at two accounts for references to ref- different disaggregation levels: ugees and displaced persons, the individual agreement level; including repatriation, that and the average by country and appear in the peace agreement. year level. This analysis will also Chapter 3 | 95

• Socio-economic development this one-year lag is that the effect and reconstruction (Dev): of a political agreement would this binary variable takes the materialize and get reflected in value of 1 if the peace agree- the process of return of refugees ment includes any provisions in the subsequent year. specifically addressing socio- economic reconstruction or 3.6.5 Results and policy implica- development in general terms. tions • Governance, political institu- Tables 3.5 and 3.6 present tions (Pol): this variable ac- our correlation results and the counts for any instances of following conclusions can be the agreement including any drawn. First, effective repatria- mention of mechanisms re- tion of refugees is a huge dilem- forming or establishing new ma, requiring sustainable politi- political institutions. This cal settlements that take into con- variable also includes provi- sideration several factors. sions for interim administration and/or new democratic For example, having a policy institutions. on refugees embedded in the political settlement is a must and • State definition (Stdef): this is this seems evident from our cor- a binary variable, taking the relation results at both the indi- value of 1 if the peace agree- vidual agreement level as well ment addresses the issue of as at the average country-year state definition. level, where mentions of refugees • Democracy (HrDem): this in peace agreements seems to be variable accounts for any ref- positively associated with the erences to general commit- return of refugees. On the gov- ments to democracy. ernance side, transition mecha- • Human rights (HrGen): this nisms such as the existence of binary variable takes the value political institutions are crucial to of 1 if the peace agreement in- ensure efficient management of cludes any general references post-conflict challenges (includ- and rhetorical commitment to ing acts of violence, massacres, human rights, principles of hu- security challenges, forced dis- manitarianism/law, interna- placements, etc.). This seems to tional law, or the ‘rule of law’. be reflected in our results where We use UNHCR data to cal- correlations between mentions culate the share of returned refu- of political institutions in peace gees in a certain year out of total agreements are positive and sta- refugees in the same year. We ac- tistically significant at both the count for the peace agreements individual agreements level and assessment variables in a certain the average by country and year year (t), whereas we take the level. share of returned refugees out of To support reconstruction and total refugees in the subsequent safe return of refugees, political year (t+1). The rationale behind settlements are supposed to en- 96 | Repatriation: scenarios for conflict resolution and reconstruction

sure that a democratic transition return. This seems evident from is taking place instead of preserv- the positive and statistically cor- ing the status quo. Our positive relations between mention of hu- and statistically significant cor- man rights in peace agreements relations between democracy and share of returned refugees aspects in peace agreements and at both the individual agreement share of return refugees at both level and average country-year the individual agreement level agreements level. and the average country-year lev- We thus conclude that our cor- el seem to confirm this assump- relation results confirm the impor- tion. On a related front, political tance of the socio-economic recon- settlements are supposed to be struction, governance and inclu- inclusive as much as possible in sion aspects of peace agreements order to guarantee that refugees in order to secure refugees’ return.

Table 3.5: Correlation between returned refugees and peace agreements aspects on the agreements level

log (share of returned refugees

in subsequent year) log (Refugees, GRef) 0.0824** log (Socio-economic reconstruction, Dev) 0.0287 log (Political institutions, Pol) 0.0907** log (State definition, Stdef) 0.0613* log (Democracy, HrDem) 0.112*** log (Human rights, HrGen) 0.135*** N 1860

Note: * p<0.05 ** p<0.01 *** p<0.001’ Source: Authors’ calculations based on UNHCR data and PA-X database

Table: 3.6: Correlation between returned refugees and peace agreements aspects on the average country and year level

log (share of returned refugees in subsequent year) log (Refugees, GRef) 0.147** log (Socio-economic reconstruction, Dev) 0.0445 log (Political institutions, Pol) 0.129* log (State definition, Stdef) 0.105 log (Democracy, HrDem) 0.115* log (Human rights, HrGen) 0.150** N 562 Note: * p<0.05 ** p<0.01 *** p<0.001’ Source: Authors’ calculations based on UNHCR data and PA-X database Chapter 3 | 97

3.7 Conclusions that secure the basic needs of refugees, international support This chapter has discussed the must include both humanitarian literature on political settlements aid and economic investments and reconstruction plans in the geared towards job creation for MENA region and their impli- host country nationals and refu- cations for refugee repatriation gees. and resettlement, with a special The chapter then examined focus on Syria. Based on existing the most recent literature on po- literature, the chapter suggests litical settlements and elite pacts, that refugee repatriation is an op- drawing the following insights portunity to build more inclusive on political settlements: a) they states that guarantee their citi- should not be defined solely as a zens safety over the long term. contract between states and soci- In terms of reconstruction, eties; (b) they involve horizontal the chapter argued that it has negotiations between contending to start from the nationals, or at elites, and vertical relationships least that they should have the between elites and their follow- most important role in deciding ers; (c) they should not necessar- the direction of reconstruction. ily be defined as a reflection of a For example, reconstruction in common understanding between places like Syria is particularly elites; (d) they influence the form, complicated by questions of how nature, and performance of insti- assistance can be given to a re- tutions; and (e) they can be more gime that was in large part re- or less inclusive. sponsible for the country’s dev- Regarding the different sce- astation and has been implicated narios of political settlements, the in war crimes. Furthermore, the chapter proposed a 3-D typology role of the international commu- addressing the relationship be- nity remains an important part tween three complex post-war or of the reconstruction, not only in conflict resolution issues: the type terms of infrastructure and other of political settlement; the recon- physical considerations, but also struction plans’ response to these for rebuilding the state and its settlements; and the impact on institutions. In this regard and the repatriation (or return) of the in the context of post-conflict refugees or the displaced to their political transition, a safe and lands. Setting up a framework sustainable return of refugees that includes a measure of ‘po- often requires a framework that litical settlement scenarios’ can acknowledges the political roots hold many dimensions. This 3-D of the armed conflict rather than relationship can create a typol- just its humanitarian dimension. ogy of different type of political This suggests that to encourage settlements depending on the de- host countries to adopt policies gree of progress on each of these 98 | Repatriation: scenarios for conflict resolution and reconstruction

criteria. The way to assess these moderate, and worst-case scenar- three criteria is to answer three ios. Between the two extremes, diagnostic questions: (1) how in- the moderate scenario represents clusive/exclusive is a political a settlement that includes most settlement?; (2) what motivates of the elites, with partial clien- elites to accept a political settle- telism, which will lead to a mod- ment; and (3) by what norms is erate but still strong support for bureaucracy governed? reconstruction plans, and poten- Based on these three ques- tially close to half of the refugees tions, the type of political settle- returning. ment can be classified as follows: Finally, this chapter ana- (1) the ideal scenario: an inclu- lyzed the linkages between peace sive, coordinated, and imperson- agreements (a proxy for political al settlement; (2) the worst-case settlements) and the share of re- scenario: exclusive, spoils-driv- turned refugees on the ground. en, and personalized settlement; To that effect, using the Peace and (3) the middle scenario: an Agreements (PA-X) database and in-between these two previous UNHCR statistics, a correlation scenarios. This typology has exercise was undertaken to as- been applied to two cases from sess the association between the the MENA region: Algeria and share of returned refugees out of Lebanon. It was found that the total refugees on the one hand, political settlement in Algeria and some aspects of peace agree- can be considered as an exclusive ments on the other, notably men- political settlement, whereas the tions of refugees, socio-economic political settlement in Lebanon reconstruction, governance as- post-war in 1989 is more of an pects, the extent of democracy, inclusive model. But both models and human rights. In particular, are fragile and based on powers this correlation was analyzed of elites and clientelism: they do at two different disaggregation not address the root causes of the levels: the individual agreement conflict. level; and the average by coun- The three approaches were try and year level. Our results then applied to the case of Syria: confirm the importance of the its political settlement scenarios, socio-economic reconstruction, their consequences for recon- governance and inclusion aspects struction plans, and the impli- of peace agreements in order to cations of these two for the re- secure refugees’ return. patriation of refugees. This has indicated three scenarios: ideal, Chapter 3 | 99

References Hermez, S. (2011). On Dignity and Clientelism: Lebanon in Allen, T. and Morsink, H (1994) the Context of the 2011 Arab When refugees go home. Lon- Revolution. Studies in Ethnic- don: James Currey. ity and Nationalism, Vol. 11 No. 3. 2011. ASEN Barnes, C. 2009. Renegotiating the political settlement in Kelsall, T. (2016). Thinking and war-to-peace Transitions, Pa- Working with Political Settle- per commissioned by the UK ments. Shaping policy For De- Department for International velopment. Overseas Devel- Development. Reconciliation opment Institute (odi.org) Resources. Berthier, R. (2018) Les réfugiés Kymlicka, W (1995) Multicultural syriens, une richesse ines- citizenship: a liberal theory of pérée pour l’économie liba- minority rights. Oxford: OUP. naise. Orient XXI. Kymlicka, W. (1993) The social Bojicic-Dzelilovic et al, (2014) contract tradition. In: P. Sing- State-building, nation-build- er, ed. A companion to ethics. ing and reconstruction. In: Oxford: OUP. Kaldor, Mary and Rangelov, Laws, E. (2012) Political Settle- Iavor, (eds.) The Handbook of ments, Elite Pacts, and Gov- Global Security Policy. Wiley- ernments of National Unity - a Blackwell, Malden, MA, USA, Conceptual Study. The devel- pp. 265-281. ISBN opment leadership program, University of Birmingham. Chandler, D (2006) Empire in de- Long, K (2008) State, nation, citi- nial: the politics of state-build- zen: rethinking repatriation. ing. London: Pluto. RSC Working Paper No. 48. Chimni, B (1993) The meaning of Oxford: Refugee Studies Cen- words and the role of UNHCR tre. in voluntary repatriation. In- Long, K (2012) Statebuilding ternational journal of refugee Through Refugee Repatria- law. tion, Journal of Intervention Chimni, B (200) From resettle- and Statebuilding ment to involuntary repatria- Luckham, R. 2003. ‘The Interna- tion: towards a critical history tional Community and State of durable solutions to refu- Reconstruction in War-Torn gee problems. Refugee survey Societies’ in ‘Ebnöther, A. quarterly, 23 (3), 55_73. and Fluri, P. (eds) After In- El Dardiry, G (2016). Thinking tervention: Public Security Beyond ‘Crisis’: Population Management in Post-Conflict Displacement and State Build- Societies-From Intervention to ing in the Middle East. The Sustainable Local Ownership Middle East Institute. 100 | Repatriation: scenarios for conflict resolution and reconstruction

McLoughlin, C. (2017) When Osei, RD., C. Ackah, G. Domfe does service delivery under- and M. Danqah (2017) ‘Politi- mine state legitimacy? Why cal Settlements and Structural perceptions of fairness matter. Change – Why Growth Has The development leadership Not Been Transformational in program, University of Bir- Ghana’ in Pritchet, L., K. Sen mingham. and E. Werker (Eds.) ‘Deals Mushtaq, K. (2010) Political Set- and Development: The Po- tlements and the Governance litical Dynamics of Growth of Growth-Enhancing Insti- Episodes’, Oxford Scholarship tutions. SOAS, University of Online. London, UK. Werenfels, I (2009). An Equili- Parks, T. and W. Cole (2010) ‘Po- birum of Instablity: Dynamics litical Settlements: Implica- and Reproduction Mechan- tions for International Devel- sims of Algeria’s Political Sys- opment Policy and Practice’, tem. L’Harmattan ‘Confluenc- The Asia Foundation Occasional es Méditerranée’ 2009/4 No. Paper No. 2, July. 71 p. 179-194. Perrera, S. (2015) Political En- Wolff, S. (2011) Post-Conflict gagement with Non-State Ac- State Building: the debate on tors in Areas of Limited State- institutional choice, Third hood. The development lead- World Quarterly. Hartzell, C ership program, University of and Matthew Hoddie (2003) Birmingham. Institutionalizing Peace: Pow- er Sharing and Post Civil War Petrin, S. (2002) Refugee return ‐ Conflict Management. Ameri- and state reconstruction: a can Journal of Political Sci- comparative analysis. UN- ence. HCR Working paper Yahya, M et al (2018) Unheard Plonski. S and Walton, O (2018). Voices, what refugees need Conflict and Peace in Border- to go home. Carnegie Middle lands: a View from the Mar- East Center gins, Accord Insight 4 by Con- ciliation Ressources Schünemann, J. and A. Lucey (2015) ‘Success and Failure of Political Settlements: Defining and measuring transformation’, Political Settlements Re- search Program Working Paper No. 2. Chapter 3 | 101

Annex 3.1

Table 1: Peace agreements in Libya, Syria, and Yemen post-2011, as per the UN Peace- maker Database

Country Conflict Agreement Title Date Type

Libya Intra-State Libyan Political Agreement 17/12/2015

Syria Intra-State Memorandum on the creation of de-escalation 6/5/2017 areas in the Syrian Arab Republic

Syria Intra-State Letter dated 29 December 2016 from the 29/12/2016 Permanent Representative of the Russian Federation to the United Nations and the Chargé d’affaires a.i. of the Permanent Mission of Turkey to the United Nations addressed to the President of the Security Council

Syria Intra-State Final Communiqué of the Action Group for 30/06/2012 Syria (Geneva Communiqué)

Syria Intra-State Resolution 2043 (2012): The situation in the 21/04/2012 Middle East

Syria Intra-State Preliminary Understanding on the United 19/04/2012 Nations Supervision Mechanism in the Syrian Arab Republic

Syria Intra-State Six Point Proposal of the Joint Special Envoy 14/04/2012 of the United Nations and the League of Arab States

Yemen Intra-State The Peace and National Partnership Agreement 21/09/2014

Yemen Intra-State Agreement on the Implementation Mechanism 5/12/2011 for the Transition Process in Yemen in Accordance with the Initiative of the Gulf Cooperation Council (GCC)

Source: UN Peacemaker Database 102 | Repatriation: scenarios for conflict resolution and reconstruction

Annex 3.2

Table 1: Peace agreements in Syria with a mention of new political institutions

Agreement Date UN Security Council Resolution 2254 2015-12-18 Statement by the Syrian Revolutionary Factions 2015-09-15 Final Communiqué of the Action Group for Syria (Geneva 2012-06-30 Communiqué)

Source: PA-X database

Table 2: Peace agreements in Syria with a mention of socio-economic reconstruction or development

Agreement Date Joint Statement by Iran, Russia and Turkey on the International 2018-05-15 Meeting on Syria in Astana, 14-15 May 2018 Final statement of the Congress of the Syrian national dialogue, Sochi, 2018-01-30 January 30, 2018 Joint Statement by Iran, Russia, and Turkey on the International 2017-10-31 Meeting on Syria in Astana, 30-31 October 2017 Joint Statement by Iran, Russia, and Turkey on the International 2017-09-15 Meeting on Syria in Astana East Qalamoun Ceasefire 2017-09-05 Ceasefire in Towns of Talbisa, Rastan and Al-Hula in Homs 2017-07-31 Countryside between Syrian Government and Free Syrian Army (FSA) Memorandum on the Creation of De-escalation areas in the Syrian 2017-05-04 Arab Republic Joint Statement by Iran, Russia, Turkey on the International Meeting 2017-01-24 on Syria in Astana, January 23-24, 2017 Syrian Ceasefire Agreement: Standard Conditions and Procedures 2016-09-09 Necessary to Ensure the Cessation of Hostilities UN Security Council Resolution 2254 2015-12-18 Homs Ceasefire 2015-12-07 Statement by the Syrian Revolutionary Factions 2015-09-15 Points of the Truce with the People’s Protection Units (YPG) 2014-04-24 Homs Hudna Agreement 2014-02-07 Final Communiqué of the Action Group for Syria (Geneva Communiqué) 2012-06-30 Six Point proposal of the Joint Special Envoy of the UN and the 2012-04-14 League of Arab States accepted by the government

Source: PA-X database Chapter 3 | 103

Annex 3.3

Table 1: Peace agreements in Syria with a mention of refugees, 2012-2018

Refugees/ Agreement Date Displaced persons Joint Statement by Iran, Russia and Turkey on the 5/15/2018 1 International Meeting on Syria in Astana, 14-15 May 2018 Memorandum on the Creation of De-escalation areas in 5/4/2017 1 the Syrian Arab Republic UN Security Council Resolution 2254 12/18/2015 1 Homs Ceasefire 12/7/2015 1 Final declaration on the results of the Syria Talks in 10/30/2015 1 Vienna as agreed by participants UN Security Council Resolution 2043 4/21/2012 1 Final statement of the Congress of the Syrian national 1/30/2018 2 dialogue, Sochi, January 30, 2018 Announcement of Factions on joining the suspension of 10/8/2017 2 Hostilities Follow up on the implementation of the Previous 8/31/2016 2 Agreement in al-Waer, Homs Wadi Barada Agreement 1/6/2016 2 Homs Hudna Agreement 2/7/2014 2 ‘Hudna’ (truce) of al-Zabadani, Kefriyya and al-Fu’aa 9/20/2015 3

Source: PA-X database Note: 1 = refugees/displaced persons are merely mentioned, the provision can be said to be rhetorical, without detail or substance. 2 = the agreement contains provision(s) on refugees/displaced persons; these provisions are detailed or include a clear anti-discrimination clause. 3 = the agreement deals with issues related to refuges/displaced persons in a substantive and substantial way, providing details on implementation modalities; alternatively, the agreement makes both fairly detailed provisions and includes an anti-discrimination clause. 104 | Repatriation: scenarios for conflict resolution and reconstruction

Table 2: Peace agreements in Libya with a mention of refugees, 2012-2018

Agreement Date Refugees/ Displaced persons

Palermo Conference for and with Libya, Conclusions 2018-11-13 1

Pledge of Reconciliation and Cooperation between the 2018-09-07 1 Family and Friends in the Area of Tajoura and Souk Jumaa

Pledge of Reconciliation Between the Cities of Zawiyyah 2018-05-10 1 and al-Zintan

Closing statement from the reconciliation meeting 2018-04-17 1 between Zintan and Zawiyya

Joint Declaration (Paris) 2017-07-15 1

Final Agreement between Zintan and Mashashiyya 2017-05-18 1

Statement of the Fezzan Forum for Libya 2018-02-20 2

Reconciliation Agreement between the Zintan and 2017-05-18 2 Mashashiyya Tribes

Agreement of Social Honour for the Tribes of Tarhūnah, 2017-02-08 2 and the Tribes of Ghriyān, Mashāshiyyah, al-Qal’ah, Yafrin, Jādū, Kābāw, Nālūt and Wāzin

Minutes of Misrata-Tawagha Agreement on the Return of 2016-08-31 2 Displaced Persons and Compensation for those Affected

Humanitarian Appeal for Benghazi 2016-03-16 2

Libyan Political Agreement (Sukhairat Agreement) 2015-12-17 2

Final Communique: Comprehensive Dialogue for Calm 2015-03-19 2 and Peace in the Nafusa Mountains

Table 3: Peace agreements in Yemen with a mention of refugees, 2012-2018

Agreement Date Refugees or displaced persons

Agreement for the exchange prisoners, detainees, missing 2018-12-17 2 persons, arbitrarily detained and forcibly disappeared persons, and those under house arrest

Agreement regarding the Exchange of Prisoners, 2018-01-18 3 Detainees, Remains and Missing Persons

Statement on Security in Governorate of Ibb 2015-06-02 1

National Dialogue Conference Outcomes Document 2014-01-25 3

Agreement between the Bani Maudhah, Bani al-Shurayfi 2014-01-20 1 and Bani Rassam in Utmah, Dhamar

CHAPTER 4

Repatriation: potential contributions to post-conflict economic development Semih Tumen* 4.1 Introduction cal and human capital back to those countries (especially Syria) Following the Arab uprisings, in the short term is still a remote many countries in the MENA possibility. Nevertheless, policy- region – especially Iraq, Libya, makers should start discussing Syria, and Yemen – have experi- alternative scenarios for repatria- enced violent conflict, resulting tion and economic reconstruc- in huge political, institutional, tion in conflict countries with the and economic destruction. Mil- ultimate purpose of streamlining lions of people have been forced transition and normalization. to leave their homes. The physi- The plan of this chapter is as cal, financial, and human capital follows. It starts with a detailed resources of those countries have analysis of the current economic fled to safety. situation in the four countries This chapter discusses the based on satellite-data estimates. economic cost of conflict, post- The purpose of this exercise is conflict growth scenarios, and to provide rough estimates for the potential contribution of the extent of physical destruc- repatriation to political transition, to demonstrate the lack of tion, post-conflict development, economic opportunities in these economic growth, and recon- conflict countries, and to empha- struction in four Arab countries: size the importance of designing Iraq, Libya, Syria, and Yemen. feasible and sustainable econom- The intensity of violence has de- ic and political reconstruction clined significantly, and political scenarios. The methodology is problems have changed nature described in detail, and a com- in those countries recently. But parison of the baseline estimates the conflicts are not yet over, with other estimates in the litera- and a massive return of physi- ture is performed.

* Associate Professor of Economics. TED University, Ziya Gokalp Cad., No.48, 06420 Kolej, Ankara, Turkey. [email protected]. 107 108 | Repatriation: potential contributions to post-conflict economic development

There follows an analytical multi-dimensional conflict, and framework that could enable state failures. Millions of people policy-makers to understand the have been forced to leave their reconstruction and post-conflict homes. Cities, cultural heritage, growth dynamics. The respec- and physical infrastructure have tive roles of public and private been utterly destroyed. investment are discussed. Em- Research on political conflict pirical estimates of post-conflict in the MENA region has mostly growth are reviewed. Results of focused on the refugee crisis and a simulation exercise focusing on the potential impact of refugee various counterfactual scenarios waves on several key outcomes about post-conflict development for natives residing in host coun- paths in Syria are discussed and tries – such as employment, interpreted. The following sec- wages, consumer prices, housing tion reviews the literature on prices/rents, health status, crime the impact of return migration/ rates, schooling, environmental repatriation on economic growth outcomes, and firm openings performance and other outcomes (see, for example, Del Carpio and in the countries of origin. Coun- Wagner, 2015; Balkan and Tu- try experiences are analyzed, men, 2016; Tumen, 2016; Cerito- and the main lessons are listed. glu et al, 2017; Assaad et al, 2018; The final section lists policy Balkan et al, 2018; Akgunduz et recommendations along with a al, 2018; Akgunduz and Torun, brief discussion of concluding 2018; Fallah et al, 2019; Tumen, remarks. 2018; Tumen, 2019; Aksoy and Tumen, 2019; Mamei et al, 2019). Political economy consequences of refugee movements have been 4.2 The economic situation the main concern driving this in the conflict countries agenda (Altindag and Kaushal, 2017). With a few exceptions, The Arab uprisings have economists seem to have most- shaken the political and econom- ly ignored what has happened ic foundations of many countries within the conflict countries. in the MENA region. Some coun- This negligence is partly due to tries have reacted positively by a lack of systematic and healthy quickly restoring stability and data on the state of the econo- implementing reforms to en- mies in the post-conflict era. hance resilience to shocks, while This section develops a sys- others – such as Iraq, Libya, Syr- tematic approach to generating ia, and Yemen, Libya – have ex- credible data using data science perienced devastating wars and techniques with the ultimate terrorism with huge physical purpose of measuring the eco- and human capital costs, large- nomic cost of conflict in the af- scale institutional degradation, flicted countries. The main data Chapter 4 | 109

source is the satellite images pare the night-light intensity in processed by means of the geo- Syria before and after the conflict graphical information systems began. Using the GIS methods (GIS) methods. The focus will be adopted in related research, they on four afflicted countries: Iraq, construct indices combining the Libya, Syria, and Yemen. These contrast and dispersion of the are the countries that have expe- night-lights, and report the time rienced the most severe physical series evolution of those indices and economic destruction in the both for the entire country and at region. A set of preliminary cal- regional level. The policy impli- culations shows that GDP and cations are discussed. physical capital stocks are almost Technically, this approach halved in some of those coun- has two major advantages. First, tries. night-light intensity is a good Specifically, the data provid- alternative measure of income ed by the National Geophysical growth at national level. Any po- Data Center of the United States tential measurement error due to errors in satellite data will most are used to compare the night- likely be uncorrelated with mea- light intensities before and after surement errors associated with the conflict in those four coun- classical GDP calculation meth- tries. Economists have recently ods. Thus, night-light data can started to use night-light data be a complementary source of in- measured by satellites (see, for formation in measuring national example, Henderson et al, 2012).79 income. The night-light data serve as a proxy for local economic activity Second, economic analysis and exhibit a strong correlation of regional growth and of the impacts of policies and events with other major welfare indi- on cities is hindered by the ab- cators. Indices are constructed sence of regular measures of lo- combining the contrast and dis- cal economic activity. The night- persion of night-lights within light data are available at a far fine-grained geographical re- greater degree of geographical gions, and then the time series fineness than is attainable by evolution of those indices are any standard income and prod- reported. uct accounts. GDP numbers are In a recent series of work, not systematically available at a Ceylan and Tumen (2018, 2019) regional level in most countries. use satellite data provided by the Much of the interesting variation National Geophysical Data Cen- in economic growth takes place ter of the United States to com- within countries rather than between countries. Therefore, 79 See also Michalopoulos and Papa- ioannou (2013a, 2013b), Hodler and the GIS methods provide a ro- Raschky (2014), Alesina, Michalopou- bust basis for measurement and los, and Papaioannou (2016), Rohner, Thoenig, and Zilibotti (2013), and Sto- analysis of economic activity at reygard (2016). regional level. 110 | Repatriation: potential contributions to post-conflict economic development

Our experience suggests that Force satellites – the Defense Me- economic reconstruction in con- teorological Satellite Program – flict-affected regions cannot be have been orbiting the earth 14 successfully achieved without times a day. The satellites have a feasible and sustainable road- used their sensors to record the map. Estimating the economic earth-based lights since the 1970s cost of conflict in Iraq, Libya, – there has been a digital archive Syria, and Yemen – the four most open to public use and available severely destroyed countries in for research since 1992. The satel- the MENA region – would serve lites were originally designed to as a robust and technical start- collect meteorological informa- ing point, and a basis for policy tion, that is, cloud images, mostly debate on post-conflict recon- for forecast purposes; the night- struction and economic growth. lights from human settlements Although economic growth indi- are also collected in the course cators exist for some those coun- of orbiting. Images from every tries, they are highly unreliable, location on earth are recorded and therefore can be misleading by each satellite every night be- due to the lack of a stable/trusted tween (local time) 8.30pm-10pm. national statistical infrastructure. The observations are processed To perform this task, satel- by the National Geophysi- lite data have been collected to cal Data Center, where certain construct economic activity in- data cleaning activities are car- dices using the night-light data ried out. First, the observations for Iraq, Libya, Syria, and Ye- from locations experiencing the men. The indices are constructed bright half of the lunar cycle are at the aggregate level for each removed. Second, summer ob- country. The data period is 1994- servations with late sunsets are 2018, with some breaks in 2014 dropped. Third, the polar lights and 2018 for some regions. It (the auroral lights) are removed. should be noted that the timing Fourth, observations obscured and duration of conflict differ by clouds are dropped. And, fi- across those four countries; so, a nally, observations from the loca- detailed examination of the exist- tions where there are forest fires ing data horizon is required to un- are also eliminated. derstand the impact of the conflict. The purpose of this data Technical details on the methodol- cleaning/processing is to end ogy are provided below. up with a data set consisting solely of lights produced by hu- 4.2.1 Methodology man activity. This data set re- In this part, the measure- flects man-made outdoor and ment methodology is described also some indoor activity, which in detail.80 The United States Air proxy aggregate economic activity for each observation. At the 80 See Henderson et al (2012) for more technical details and additional analyses. end, valid observations are av- Chapter 4 | 111

eraged over nights for each sat- are top-coded to take the value ellite in a given year and a large of 63. As Henderson et al (2012) data set with satellite-year varia- clearly indicate, the quality of ob- tion for each location on earth servations decline with the age of is constructed. Those locations the satellite. So, year fixed effects are classified as ‘pixels’ (around are used to control for the satel- 0.86 square kilometers at the lite age effect in statistical analy- equator). The reported pixels are sis. Figure 4.1 provides an exam- placed between 65 degrees south ple of a satellite image for Syria, and 75 degrees north latitude. Lebanon, Iraq, Northern Jordan, The excluded pixels fall into the and Southeastern Turkey. The Arctic regions, where a very tiny pink color represents the loca- and negligible fraction of human tions for which the light intensity activity is recorded. For each in 2012 declined as of 2016, while pixel, observations are averaged the blue color represents the lo- across satellites to obtain pixel- cations for which the light inten- year data, which is also used in sity in 2012 increased as of 2016. the analysis in this section. Clearly, the picture is consistent The pixel-year data is de- with the movement patterns of scribed by an integer number forcibly displaced Syrians. Be- ranging between 1 and 63. A tiny low, the results of the satellite fraction of data from very dense- data analysis are presented for ly populated and rich locations Iraq, Libya, Syria, and Yemen.

Figure 4.1. Satellite image documenting the movement patterns of forcibly displaced Syrians between 2012 and 2016.

Source: Ceylan and Tumen (2019) and the National Geophysical Data Center of the United States. 112 | Repatriation: potential contributions to post-conflict economic development

4.3 Results for Arab con- the main focus is on the most re- flicts cent major conflict in Iraq, which started in 2003 and continued un- 4.3.1 The conflict in Iraq til the withdrawal of US troops in 2011. After the emergence of ter- The conflict in Iraq has been rorism in the northern region of one of the most profound, com- the country at the end of 2014, a plicated, and long-lasting ones in local conflict has re-emerged es- the MENA region. The 1980-1988 pecially in Northern Iraq, which Iran-Iraq War and the invasion still continues locally and inter- of Kuwait by Iraq in 1990 are mittently. not covered in our data set. So,

Figure 4.2: Aggregate economic activity in Iraq between 1994-2018 measured in terms of satellite-based night-light data.

Iraq (Aggregate Index, 2018-1994) 6

0 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018

Figure 4.2 plots the night- growth over 15 years despite light index constructed to char- booming oil prices – which is low acterize the trends in the level in developing country standards. of overall economic activity in The second period is an episode Iraq between 1994 and 2018. Two of gradual resolution of disputes main episodes are observed dur- and selective lifting of economic ing this time period: 1994-2008; sanctions. The annual rate of eco- and 2008-2018. The first period is nomic growth – measured by the characterized by political insta- evolution of satellite-based night- bility, civil conflict, international lights – is above 10%. With the disputes, wars, and heavy eco- increase in the intensity of local nomic sanctions. The economic conflict after 2016, the economic activity is nearly stagnant, aver- activity has again become stag- aging approximately 3% annual nant. The Iraqi case suggests that Chapter 4 | 113

there is a strong positive correla- which resulted in the ouster tion between conflict, economic and death of the former leader sanctions, and economic growth. Muammar Al-Qaddafi in Octo- ber 2011. The post-Qaddafi era 4.3.2 The conflict in Libya did not bring peace either; the civil conflict has continued and Libya has constantly experi- risks of further outbursts of vio- enced international disputes and lence are still alive. conflict in the past three decades,

Figure 4.3: Aggregate economic activity in Libya between 1994-2018 measured in terms of satellite-based night-light data.

Libya (Aggregate Index, 2018-1994) 0.7

0.6

0.5

0.4

0.3

0.2

0.1

0 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018

Figure 4.3 demonstrates the was ousted. The country has trends in satellite-based night- been experiencing a civil conflict light intensity in Libya between since then, and it has struggled to 1994 and 2018. The period be- reconstruct state institutions in tween 2002 and 2008 has ex- the post-Qaddafi era. As a result, hibited the strongest economic economic activity is stagnant, growth in the recent history of and there are substantial political the country. It was the period and economic uncertainties. The in which the Qaddafi regime Libyan experience also shows built good relationships with the that conflict undermines long- Western world—for example, term economic growth. the US Secretary of State made the highest-level official US visit 4.3.3 The conflict in Yemen to Libya in 2008 since 1953; Qad- The civil conflict in Yemen dafi paid the first official visit to started in 2011 and the main in- Italy the year after. As part of the spiration was again the Arab uprisings in the Arab world, an uprisings. The Houthi expan- anti-Qaddafi uprising emerged sion during the 2011-2015 period in 2011 in Libya and the regime 114 | Repatriation: potential contributions to post-conflict economic development

transitioned into violent armed dramatic humanitarian crisis. conflict after 2015. Yemen has Millions of Yemenis have needed been among the least developed humanitarian assistance, and the countries in the region, and the risks of famine and epidemic dis- heavy conflict has generated a ease are still alive.

Figure 4.4: Aggregate economic activity in Yemen between 1994- 2018 measured in terms of satellite-based night-light data.

Yemen (Aggregate Index, 2018-1994) 1.2

0.8

0.6

0.4

0.2

0 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018

Figure 4.4 depicts the trends overcome the sources of internal in the aggregate index construct- and international disputes. There ed using the night-light series. is also a lack of domestic and in- There are two episodes. The first ternational coordination in terms is until the end of 2011; the sec- of improving the general institu- ond one begins in 2012 and still tional structure and the key insti- continues. During the first pe- tutions in the country. riod, the Yemeni economy expanded rapidly and benefited 4.3.4 The conflict in Syria from the favorable economic The Syrian conflict, which conditions in the region due to started to escalate in the spring booming oil prices. In the second of 2011, has displaced millions period, however, the economy of Syrians. Based on the most re- experienced a rapid decline and cent UN figures, around 5.6 mil- the economic environment re- lion Syrian refugees have fled to turned to its conditions in late neighboring countries, which has 1990s, which basically means had important socio-economic that the extent of economic de- impacts on the entire MENA re- struction has been huge. The gion and most of Europe. The country is still struggling to conflict in Syria is different from Chapter 4 | 115

Iraqi, Yemeni, and Libyan con- large human capital destruction, flicts in that it generated one which may have irreversible of the largest refugee waves in consequences for the country’s the past century. This suggests economic growth potential in the that the Syrian conflict entails long term.

Figure 4.5: Aggregate economic activity in Syria between 1994-2018 measured in terms of satellite-based night-light data.

Syria (Aggregate Index, 2018-1994) 5 4.5 4 3.5 3 2.5 2 1.5 1 0.5 0 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018

Figure 4.5 plots the trends in 4.3.5 Comparison with existing economic activity in Syria based estimates on satellite data covering the pe- The estimates presented in riod between 1994 and 2018. The this section suggest that the scale figure suggests that the Syrian and intensity of the violence and conflict has had devastating eco- destruction in the region have nomic consequences. The level of been unprecedented in recent his- night-light intensity is way be- tory and the extent of economic low its level in the mid-1990s. In destruction is the largest in Syria. comparison with Iraq, Libya, and The death toll in Syria, as of the Yemen, the Syrian economy is in first half of 2019, has reached a much worse situation in terms 570,000 according to the calcu- of physical destruction along lations of the Syrian Observa- with vanished human capital tory for Human Rights. The UN and destroyed institutions. Any Economic and Social Commis- reconstruction efforts for the Syr- sion for Western Asia (ESCWA) ian economy should entail seri- performed a very detailed study ous repatriation efforts in addi- to estimate the economic costs of tion to investments in physical the Syrian civil war and conflict. infrastructure and institutions. According to their estimates, the 116 | Repatriation: potential contributions to post-conflict economic development

destruction of the physical capi- lic infrastructure investments, tal stock by the end of 2017 had private capital investments, reached almost $120 billion (ES- human capital, total factor pro- CWA, 2018). Another detailed ductivity, demographics, and study calculating the cost of the labor market outcomes. Those Syrian conflict was conducted by elements are typically embedded the World Bank. Their estimates in large-scale economic growth suggest that the cumulative loss models in a theory-consistent in GDP reached $226 billion – ap- way. One prominent example is proximately four times Syrian the long-term growth model of GDP in 2010 — between 2011 the World Bank (see, for exam- and 2016 (World Bank, 2017). ple, Hevie and Loayza, 2012; and Finally, Devadas et al (2019) use Loayza and Pennings, 2018) and a version of the World Bank’s its extensions embedding pub- long-term growth model (aug- lic capital/infrastructure – see menting public capital and infra- Devadas and Pennings (2018). structure investments) and show These versions of the model are that real GDP in Syria contracted particularly useful in assessing by 12% per year on average over the post-conflict growth poten- the 2011-18 period. These esti- tial of conflict countries because mates are largely in agreement they practically allow for key with the satellite-based estimates quantitative exercises comparing presented in this section. various counterfactual scenarios under certain assumptions. The analysis in Section 2 demonstrates that Iraq, Libya, Syria, 4.4 Post-conflict growth and Yemen suffer from severe potential degradation in the quantity and quality of physical capital and in- 4.4.1 An analytical framework frastructure, which has substan- and some preliminary findings tially reduced the level of GDP To understand the potential and the growth capacity of those contribution of repatriation to countries. Economic reconstruc- post-conflict economic growth tion of conflict countries in the and development in the conflict region substantially relies on re- countries (and also the interac- building the capital stock in those tion between repatriation and countries. It should be empha- economic growth), it is useful to sized that investment in public think in terms of an analytical capital will be key, especially in framework that appropriately the initial stages. The findings re- incorporates the key drivers of ported by Devadas and Pennings growth in low- and middle-in- (2018) largely confirm this con- come countries with a long-term jecture. Using an extended ver- perspective. Those drivers typi- sion of the World Bank’s model, cally include public capital, pub- they discuss the role of public Chapter 4 | 117

capital/infrastructure in increas- gests that Iraq, Libya, Syria, ing economic growth potential in and Yemen will benefit sig- low- and middle-income coun- nificantly from large amounts tries. Public capital and infra- of public capital investment, structure have been among the especially in the form of es- key drivers of economic growth sential infrastructure. and have become a key pillar • Private investment boosts of development strategies in economic growth in the short emerging economies – China’s term: Private capital invest- infrastructure-led growth strat- ment has a large impact on egy is the best-known example. economic growth in the short Although there is a disagree- term in low- and middle- ment in the literature about the income countries, while the magnitude of the effect of pub- size of the effect fades away lic infrastructure on growth in quickly and becomes smaller developing countries (Calderon in the long term. This implies and Serven, 2014), the consensus that attracting large amounts is that public capital/infrastruc- of private capital investment ture investment positively affects would be important to have the growth potential of emerg- a notable short-term boost in ing economies through various economic growth. In other direct and indirect channels/ words, complementing pub- mechanisms. The main findings lic capital investment with of Devadas and Pennings (2018), private investment in the con- which also have important les- flict countries will have a high sons for the conflict countries, return on economic growth in can be summarized as follows: the short term. • Conflict countries need sub- • High-quality public invest- stantial amounts of public ment generates large growth investment: The effect of pub- dividends: Based on a novel lic investment on growth is index measuring the quality quite high when (1) the initial of public investment, Devadas public capital/output ratio is and Pennings (2018) show that low, that is, when public capi- high-quality public investment tal is scarce; and (2) the public – even when the quantity is investment is in the form of kept constant – improves ef- essential public infrastruc- ficiency and increases growth ture; while the contribution of potential substantially in low- public investment to econom- and middle-income countries. ic growth is small in countries The quality of public capital with abundant public capital. and infrastructure is typically Civil conflict and internal war poor in the region and the have largely destroyed public quality dimension even got capital in conflict countries in worse due to civil conflict in the region. This finding sug- afflicted countries. This find- 118 | Repatriation: potential contributions to post-conflict economic development

ing suggests that new public There is another set of papers investments should also aim arguing that countries exposed to improve the quality and to large negative output shocks efficiency of public capital in performed very well in the fol- those countries. lowing decades due to reconstruction dynamics (Milionis and 4.4.2 Existing empirical evidence Vonyo, 2015). Countries like Iraq, on post-conflict recovery Libya, Syria, and Yemen have ex- To assess the potential for perienced massive destruction of post-conflict recovery in the factors of production due to war MENA region, it would be useful and conflict in the aftermath of to look back into the conflict cas- the Arab uprisings. The lessons es in the past and extract some from the reconstruction efforts insights from the growth experi- after the Second World War and ences of other conflict countries insights from other experiences in recent history. The loss in eco- might be useful in evaluating nomic growth potential is large the post-conflict growth poten- especially in Syria, but history tial of the afflicted countries in suggests that a quick recovery the region. The similarities and along with strengthened socio- differences between the conflict economic fundamentals beyond countries and the countries expe- the pre-war situation need not be rienced civil conflict in the past a remote possibility – if perma- should be carefully assessed to nent political stability is achieved extract useful lessons from the in a peaceful and inclusive way. relevant historical events. Findings of some of the rel- For example, as Devadas et evant literature can be sum- al (2019) suggest, ‘One of the marized as follows. Chen et al key lessons from the experience (2008) perform an event study of post-WWII growth in the Eu- design using cross-country data ropean countries and Japan, on 41 countries between 1960- for example, was that the rapid 2013 for the purpose of analyz- growth impact of the massive ing the post-civil-war economic rebuilding of physical capital situation and developments. The was made possible, not only by main finding is that sustaining peace and political stability in the Marshall Plan resources, but the long term increases the pos- also by the relatively limited war sibility of surpassing the pre- time depreciation of the human conflict economic performance capital base and technological in conflict countries. There are potential (Smolny, 2000).’ The other studies focusing on the Sec- conflict in the region following ond World War and the econom- the Arab uprisings has mostly ic performance of the afflicted generated large refugee flows. countries in the post-war period Although the Syrian case is an (see, for example, Organski and extreme example, long-term war Kugler 1977; 1980). and conflict in the afflicted coun- Chapter 4 | 119

tries led to human capital flight Although Section 2 provides to safety, which is a major chal- estimates for the economic cost lenge. Reallocation of produc- of conflict in the afflicted coun- tion factors in the post-conflict tries, a more theory-consistent era is another important issue. approach is needed to incorpo- In an early paper, Janossy (1969) rate some realistic counterfac- highlights this issue by arguing tual exercises and analyze the that although higher returns to post-conflict growth scenarios. physical capital investment is an Devadas et al (2019) offer such an important factor improving the approach. Specifically, they use growth performance in the post- the World Bank public capital conflict period, the roles of eco- model of Devadas and Pennings nomic reorganization and real- (2019) (i) to produce a simulated location of factors of production ‘no conflict scenario’ as a coun- should not be ignored. terfactual exercise; (ii) then, to calculate the cost of conflict; and 4.4.3 Post-conflict growth sce- (iii) to evaluate the post-conflict narios in Syria growth potential in Syria. The after-war simulations are per- The Syrian crisis has had formed based on three counter- devastating humanitarian conse- factual exercises featuring differ- quences due to the huge volume ent settlement scenarios: of forced displacement it generated during the civil war. Almost · First, a moderate scenario, two thirds of the 21 million Syr- which is the baseline scenario, ians have been forced out of their incorporating the ‘Sochi-plan’ homes due to conflict – this fig- carried out by Iran, Russia, ure includes both refugees and Turkey, and international or- internally displaced individuals. ganizations. As of today, Syrian refugees ac- · Second, an optimistic sce- count for approximately one nario cast based on long-term quarter of the total number of political settlement operated refugees in the world. Similarly, and monitored by the UN. internally displaced Syrians are · And, finally, a pessimistic estimated to be around one fifth scenario featuring fragment- of the total number of internally ed political power. displaced persons in the world. These three scenarios also ac- These numbers suggest a huge company three different levels loss of human capital stock in of reconstruction assistance and Syria, which implies that all re- repatriation volumes (therefore, construction plans should in- different levels of public and clude an effective repatriation private investment; different plan to bring the country’s pro- human capital levels): the opti- ductive human capital and labor mistic scenario features a large resources back. amount of reconstruction as- 120 | Repatriation: potential contributions to post-conflict economic development

sistance and repatriation, while has trivial assumptions on recon- both are lower in the pessimistic struction assistance and repatria- scenario. The no conflict scenario tion volumes.

Figure 4.6: Pre-conflict real GDP growth and projection.

(a) September/October 2018 Vintage (b) April 2011 Vintage

% % 8.0 8.0 6.0 4.9 6.0 4.0 4.0 4.5 5.0 2.0 2.0 0.0 0.0 2.0 - - 2.0 -4.0 - 4.0 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 1995 1996 1997 1998 1999 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2000 2001 2002 2003 2004 2005 IMF WEO, Oct 2018 WDI/MFMod 2018 IMF WEO April -6 ,2011year average, 2016-2011 WDI/MFMod -10 ,2018year average, 2010-2001 IMF WEO Oct -10 ,2018year average, 2010-2001 IMF WEO, April 2011 (dashed line indicates estimate/projection)

Source: Devadas et al (2019).

Under the ‘no conflict’ coun- simulations suggest a cumulative terfactual exercise, the estimates loss in GDP potential of about suggest that annual real GDP $300 billion over 2011-2018. The growth averaged around 5.3% depletion of factors of production between 2011 and 2018 (see Fig- alone may account for about 87% ure 4.6), which would increase of the negative GDP growth on real GDP from $60 billion in 2010 average, and further, about 64% to $91 billion in 2018. Similarly, of the average negative growth is real GDP per capita increased due to physical capital destruc- from $2,857 in 2010 to $3,774 by tion. Physical capital destruction 2018. The cost of conflict is es- reflects the compounded effects timated to be around -12% real of large outright damages, low GDP growth per year on aver- new investments, and a falling age from 2011 to 2018 (see Figure output base, that is adversely af- 4.7). This huge economic destruc- fected by all growth drivers. De- tion in GDP implies that the level mographics and labor account of real GDP as of 2018 is around for about 15%, human capital 7%, $23.2 billion. Devadas et al (2019) and total factor productivity 13% further document that: ‘Compar- of negative GDP growth on aver- ing the conflict versus no conflict age over the conflict years.’ Chapter 4 | 121

Figure 4.7: Conflict years calibration for GDP in Syria.

Source: Devadas et al (2019).

Their optimistic scenario, These findings imply that which features high investment high post-conflict economic per- within the first decade, suggests formance is possible in Syria if that Syria will surpass its 2010 long-term political stability is GDP and GDP per capita levels achieved, market-friendly mech- in 10 years. The pessimistic sce- anisms are redesigned to obtain nario, which embeds the assump- efficient allocation of resources tions of limited guarantees for throughout the economy, appro- micro security/property rights, priately designed reconstruction low reconstruction funds, and and repatriation programs are low investment, suggests that it implemented, and those pro- would take two or three decades grams are supported by sustain- for Syria to catch and surpass its able financing facilities. 2010 GDP and GDP per capita levels. In the optimistic and pessimistic scenarios, average real GDP per capita growth would be 8.1% and 3.1% annually. 122 | Repatriation: potential contributions to post-conflict economic development

4.5 Review of the existing individuals vary across countries results and country experi- and regions. For example, the ences emigration rates are lower in the United States, Canada, Australia, and New Zealand, while they are There are several similarities higher in most of Europe. Those and differences between return who leave the host country either migration and repatriation. Both move to another host country or terms refer to going back to the go back home. home country after spending some time in a destination coun- One main implication of try. The main difference is that, temporary (or return) migration in general, return migration re- is that the ones who choose to fers to going back to the home leave the host country are prob- country for ‘economic’ or volun- ably a selected set of immigrants; tary migrants, while repatriation therefore, focusing only on the refers to going back to the home stayers may lead to biased es- country for ‘forced’ migrants or timates from the perspective of refugees. In this section, a brief the host country. Another impli- overview of the literature dis- cation is that ‘return intentions’ cussing the impact of return mi- shape the economic decisions gration and repatriation on both of immigrants. For example, the the home and host countries is expected duration of stay may presented. Although the litera- determine the intensity of hu- ture on return migration is rather man capital investment, amount rich, there is relatively little aca- of savings, and career choice; demic discussion of the impact moreover, these patterns may of repatriation on various eco- differ across countries and time nomic and social outcomes. The periods. main purpose of this chapter is to Perhaps the questions most compile the existing evidence on relevant for our context are both types of returning behav- whether voluntary immigrants ior, and also to extract key les- or refugees have stronger inten- sons and policy implications for tions to leave the host country today’s refugee crises, especially and how the differences in those the one in Syria. intentions affect their behavior. Contrary to widespread be- Cortes (2004) argues that refu- lief, a non-negligible fraction of gees intend to stay longer in the immigrants leaves the host coun- host country than economic im- try within a few years of arrival migrants, and they are also more (Dustmann and Görlach, 2016). willing to invest in their human According to an OECD report, capital after arriving in the host 20-50% of the immigrants leave country than economic immi- the host country within the first grants. A similar result is report- five years (OECD, 2008). The em- ed by Khan (1997). Dustmann igration patterns of foreign-born (2008) argues that the differences Chapter 4 | 123

in the intentions to stay affect Other than the remittances the human capital investment channel, return migration and/ decisions of the second genera- or repatriation also affect the tions; specifically, the sons of the outcomes in the country of ori- parents who intend to stay lon- gin directly through the human ger in the host country are more capital channel. If the returning likely to attain a post-secondary migrants are high-skilled ones, degree. For savings, a number of then return migration has a clear papers – including Merkle and positive impact on the economic Zimmermann (1992), Dustmann outcomes in the home country – and Mestres (2010a), Pinger even when the initial migration (2010), and Bauer and Sinning decision is highly selective (see, (2011) – show that immigrants for example, Domingues Dos who are more willing to return Santos and Postel-Vinay, 2003). tend to save more and have high- This suggests that permanent er remittances than those who high-skilled migration has nega- are less willing to return. tive effects on the long-term hu- Another very important issue man capital capacity of the home is the potential impact of remit- country. tances, return migration, and/ In a more general setting, or repatriation on the economic temporary migration has two outcomes in the home country. opposing effects on the aggre- There is a large body of literature gate human capital level of the trying to address these questions home country. First, temporary (see Rapoport and Docquier, migration may be reducing the 2006, for a survey). Remittances human capital intensity of tem- can be defined as a major channel porary migrants, which imposes through which migration may a negative impact. Second, re- influence the wellbeing of the in- turn migration generates signifi- dividuals living in the countries cant knowledge diffusion and of origin. Remittances are also improves the technology/know- affected by the temporariness how in the home country (Dust- of migration. Temporary immi- mann et al, 2011). As a separate grants are more likely to leave issue, Spilimbergo (2009) argues their parents behind (Funkhous- that return migrants may recon- er, 1995) and they are less likely struct the institutional structure to move their assets (Dustmann in the home country and increase and Mestres, 2010b). If refugees institutional quality. are less likely to return home These findings suggest that – than economic immigrants, then perhaps due to trauma and other they will also be less likely to psychological factors, economic leave their parents behind and destruction in the home country, more likely to move their assets higher expected returns in the into the host country. host country, etc. – the refugees 124 | Repatriation: potential contributions to post-conflict economic development

might have strong incentives to In a different strand of the stay in their host countries, they literature, Lee et al (2017) inves- may have strong intentions to in- tigate the impact of repatriation vest in both their own and their of Mexicans out of the United children’s human capital, and States on the outcomes of US na- they may have weak intentions tives. They report that, contrary to save and generate remittanc- to the common claim, Mexican es. Possible repatriation policies repatriation did not expand em- should take these findings into ployment opportunities for US account. natives. Similarly, Clemens et There are various studies al (2018) show that ‘the bracero’ investigating the impact of re- – for example, the exclusion of patriation on host countries’ Mexican farmers from US agri- economic outcomes. In an early culture – did not improve wages study, Hunt (1992) investigates and employment for native US the impact of the 1962 Algeria workers in the agricultural sec- repatriates on labor market out- tor. The evidence suggests that comes in France. She reports a repatriation has slightly negative slight decline in wages and a effects on the wages and employ- small increase in the unemployment conditions in countries of ment rate in response to the repa- origin; but it has not improved triation movement. Carrington the labor market outcomes of na- and de Lima (1996) ‘examines the labor market effect of the re- tives in host countries. tornados who immigrated to Por- Although these findings are tugal from Angola and Mozam- useful, the papers reviewed bique in the mid-1970s follow- above deal mostly with devel- ing Portugal’s loss of its African oped countries and their migra- colonies.’ They document larger tion counterparts, while the situ- negative effects of repatriates on ation in Syria is different in vari- the Portuguese labor markets ous ways. Most importantly, Syr- than Hunt (1992) documents for ia has lost a substantial fraction France. In a more recent study, of its workforce in the form of Aydemir and Kirdar (2017) aim forced migration, so the context to estimate the impact of Turkish is different. It is most likely the repatriates from Bulgaria on la- case that successful repatriation bor market outcomes in Turkey. of Syrian refugees – once peace They document significant in- is established, institutions are creases in unemployment rates, rebuilt, and political/economic especially for younger workers stability is achieved in the long and for the ones in direct labor term – will substantially improve market competition with the re- the general economic situation in patriates – that is, similar skill Syria. groups. Chapter 4 | 125

It is important to take a closer were originally residing in rural look at countries and context that areas before the refugee wave are more similar to Syria. Harild had started, large-scale projects et al (2015) address exactly this were concurrently implemented issue and present eight examples to facilitate the rehabilitation of of refugee return for countries irrigation systems, seed and fer- that are rather similar to the Syri- tilizer aid packages were distrib- an context: Afghanistan, Angola, uted, support was provided to Bosnia-Herzegovina, Burundi, help returnees build their homes, Cambodia, Iraq, Liberia, and transport infrastructure was im- South Sudan. These eight cases proved, schools and government are reviewed below along the buildings were repaired/rebuilt, lines Harild et al (2015) discuss and access to sanitized water was in their study. provided. Those projects were of benefit not only to the refugees, 4.5.1 Afghanistan but also to the relevant communities as a whole. Yet funding Following the Communist was delayed several times and coup in 1978, a civil conflict sometimes it was inadequate. In emerged in Afghanistan, which addition, the level of coordina- generated a large refugee wave tion between stakeholders was mostly towards Pakistan and also sub-optimal. These factors Iran. Over almost 10 years of reduced the effectiveness of the war and conflict, more than five assistance programs. After the million Afghanis fled to safety – fall of the Taliban in 2001, anoth- more than three million went to er large wave of refugee returns Pakistan and around two million started: another two million Af- to Iran. A gradual repatriation ghani refugees returned within movement started in late 1989 two years. Building on past ex- after Soviet troops left Afghani- periences, the UN and other stan. Refugees started to return stakeholders responded more in huge masses after the collapse promptly and in a much more of the Communist regime in late organized way to the repatria- 1991. Although civil conflict in tion wave. Afghanistan continued in different forms, the repatriation wave Because of unending conflict continued and, by 2000, more and despite intensive humanitar- than four million refugees had ian aid efforts spent over almost returned. 30 years in Afghanistan, the core issues, such as housing, security, The repatriation process was and employment, still remain assisted by the UN and other unresolved to a large extent (UN- international and local NGOs, HCR, 2004). Rapidly increasing as well as the Pakistani and Af- population (despite huge refu- ghan governments. To facili- gee waves) and the existence of tate the return of refugees, who economic/political uncertain- 126 | Repatriation: potential contributions to post-conflict economic development

ties have substantially reduced main problem was that most of the effectiveness of reintegration the refugees did not wait for the programs. Moreover, the compo- formal approval of multilateral sition of returnees has changed treaties, which put them at large over time and more recent repa- risk (HRW, 2005). In addition, triates have headed to urban ar- they could not benefit from any eas: this has limited the impact of monetary and non-monetary as- policy interventions, which have sistance programs and this sud- mostly targeted the rural areas. den movement posed additional The human capital stock and in- challenges for the humanitarian stitutional structure of the coun- aid process. One main problem try have been heavily damaged. was that part of the returnees Relatively less attention has been were ‘forced’ rather than go- devoted by the relevant authori- ing back voluntarily, which also ties to rebuilding the key insti- complicated the repatriation pro- tutions that would supplement cess and reduced the effective- economic development efforts ness of humanitarian aid efforts. in Afghanistan. High-skilled Since the return was not motivat- Afghani refugees seem to have ed by strong pull factors – such permanently left the country and as economic reconstruction and have not yet returned. sustained political stability – the existing conditions in Angola 4.5.2 Angola were not conducive to resettlement and quick reintegration. Angola declared indepen- The international organizations dence from Portugal in 1972 and and humanitarian aid programs the country immediately fell into could not immediately react to a violent civil war with devas- quickly changing conditions in tating socio-economic conse- the region, and they failed to quences. More than four million understand the socio-cultural Angolans fled their home and forces motivating cross-border around 600,000 of them became movements. refugees. The main host countries were Zambia, Namibia, the Democratic Republic of Congo 4.5.3 Bosnia-Herzegovina (DCR), South Africa, the Repub- Following the fall of the So- lic of Congo, and Botswana. The viet Union in late 1991, violent civil war was ended in 2002 and political conflict in Yugoslavia a large repatriation wave had generated ethnic cleansing and started right after the end of the systematic expulsion/persecu- civil war. tion practices against Bosnian Until 2007, around 80% of people (Toal and Dahlman, the Angolan refugees in the 2011). The Srebrenica massa- host countries had returned, ac- cre of more than 8,000 Muslim cording to UNCHR figures. The Bosnian people in 1995 was of- Chapter 4 | 127

ficially accepted as genocide by 4.5.4 Burundi the international community. As Two large refugee waves a consequence of violent conflict moved from Burundi to Tanza- and civil war, around 4.6 million nia: one in 1972 due to internal Bosnians fled their home and ethnic conflict (300,000 refugees); more than 1.2 million of them and the other in 1993, again due became refugees in various host to inter-ethnic violence (350,000 countries across the Europe. refugees). After 1995, the Tanza- The return of refugees and in- nian government closed the door ternally displaced persons were to new refugee inflows and im- coordinated according to the posed several restrictions on the employment and settlement con- principles reached in the Dayton ditions of Burundian refugees in Peace Agreement in late 1995. Tanzania. Internal conflict was Although the agreement had continuing during this period ambitiously targeted the return and the international organiza- of more than two million people tions (mainly the UN High Com- within a rather short period of missioner on Refugees, UNHCR) time, only one million of them re- were unwilling to initiate a returned (fewer than 500,000 refu- patriation program as political gees) 10 years later. The fact that stability and safety had still not the refugees mostly fled to West been reached in Burundi. European countries with better After the end of civil war in economic opportunities for them early 2001, the Tanzanian gov- and their families reduced the ernment and the UNCHR agreed number of actual returns. One on a repatriation plan; the ini- main problem was that the Day- tiative started in late 2002. By ton Peace Agreement focused 2012, approximately 500,000 Bu- too much on the mechanics of rundian refugees had returned the return (which generated ad- home. Building on past lessons, ditional disputes about past the UNHCR carried out a rela- ownership of housing/property tively successful repatriation and the location of returnees) program as the outcomes and and did not consider or address living conditions of the repatri- the socio-political realities of the ates substantially improved af- Bosnian society, which resulted ter returning – although issues in unusual ethnic ‘un-mixing’ about ancestral land and ear- and segregation. Implementa- lier property ownership are still tion of micro-credit programs largely unresolved (Fransen and and other humanitarian aid proj- Kuschminder, 2012). Extensive ects by various international or- micro-credit and employment ganizations did not largely alter programs along with develop- this situation. ment assistance (despite the existence of sustainability concerns) 128 | Repatriation: potential contributions to post-conflict economic development

also facilitated the quick adapta- refugee wave from Syria towards tion of returnees. Iraq. Existing evidence suggests that the reluctance to return back 4.5.5 Cambodia to Iraq was counterbalanced by the emergence of violent conflict More than 500,000 Cambodi- in Syria (UNHCR, 2015). ans fled to Thailand in response to a violent internal conflict in The political situation in Iraq Cambodia during the 1970s and has been quite volatile and un- early 1980s. Around 400,000 of certain over the past 15 years. the Cambodian refugees went Episodes of conflict and violence back home after the 1991 Paris were often followed by episodes Peace Agreement. The UNHCR of relative stability. But the lack of and the governments of the two a permanent resolution of politi- countries agreed on a voluntary cal issues coupled with the lack return plan, which also included of high-quality protection stan- various repatriation assistance dards have made the UNHCR and monitoring programs. Vo- reluctant to offer a structured cational training programs and repatriation plan for the Iraqi access to micro-credit/finance refugees. The fragile situation in have been identified to be the the region is likely to continue to key channels motivating a suc- prevent the design of systematic cessful reintegration of return- repatriation programs, at least in ees in the Cambodian case. The the short term. The existing re- overall Cambodian repatriation turn movements are in the form experience is generally regarded of a rather compulsory or forced among experts and academics to return due to increased violence be a successful example (East- in destination countries. mond and Ojendal, 1999). 4.5.7 Liberia 4.5.6 Iraq The civil war in Liberia in 1989 The US-led invasion of Iraq has a generated around 800,000 generated a huge wave of forced refugees, who mostly moved to displacement – more than four Ghana to seek temporary refuge. million Iraqis, around 500,000 Political stability was re-estab- of whom were refugees and lished in Liberia after the end of asylum-seekers. The main host the civil war in 2003. The UN- countries in the region were HCR designed and led several Syria, Turkey, Iran, and Palestine waves of repatriation from Gha- (plus Lebanon and Jordan to a na to Liberia starting from 2004. smaller extent). The Iraqi case be- The initial waves were rather came more complicated after the unsuccessful, and the number of spark of a violent internal conflict returnees remained very low. in Syria, which forced Iraqi refu- The Ghanaian government gees back to Iraq and generated a implemented some coercive Chapter 4 | 129

measures to reduce the social (UNOCHA 2014). So, the South and economic incentives to stay Sudan experience suggests that in Ghana. First, the government the repatriation efforts should excluded the refugees from some also address conflict prevention employment services. Second, and peace-building practices, humanitarian assistance was re- and security priorities in an ef- duced substantially, which put fective way. Agricultural subsis- the refugee households in a frag- tence plans had been the key ele- ile position. Finally, as the refu- ments of repatriation assistance. gee population declined over time, economic activity within the refugee population shrank and the economic incentives to 4.6 Concluding remarks stay in Ghana diminished sub- and policy perspective stantially (Omata, 2012). To sum up, several policies framed as The purpose of this chapter the ‘push factors’ were the main has been to discuss the economic motivating forces behind repa- cost of conflict in afflicted coun- triation, rather than favorable, tries, the post-conflict growth supportive, and conducive pull scenarios, and the potential factors originating in the home contribution of repatriation to country. the economic and political development of conflict countries 4.5.8 South Sudan – especially Syria. To achieve this goal, the level of economic The Republic of South Su- destruction in the four main dan was established after a six- conflict countries – Iraq, Libya, year interim period following Syria, and Yemen – were quanti- the 2005 ‘Comprehensive Peace fied using satellite data and GIS- Agreement’, which ended nearly based estimates. Post-conflict five decades of war and violence growth scenarios were discussed in the Sudan region. Uganda was based on an extension version the main host country. Among of the World Bank’s Long-Term more than two million refugees, Growth Model. And, finally, a around 550,000 returned within detailed literature review on the the scope of the repatriation pro- economic consequences of return gram of the UNHCR, while the migration/repatriation for both majority of the remaining refu- the home and host countries were gees returned spontaneously – discussed, along with a detailed that is, with no reference to any analysis of the repatriation expe- repatriation program. The emer- riences of selected countries – Af- gence of new civil conflicts in ghanistan, Angola, Bosnia-Herze- early 2010s generated new waves govina, Burundi, Cambodia, Iraq, of displacement, which compli- Liberia, and South Sudan. cated the picture significantly 130 | Repatriation: potential contributions to post-conflict economic development

Clearly, the repatriation 1. The existing consensus in the process is not straightforward. literature suggests that refu- Many parties with conflicting gees are less likely to leave interests and goals are involved. the host country than eco- The home country needs to of- nomic migrants; they have fer peace, political stability, em- stronger incentives to invest ployment opportunities, and in their own and their chil- other socio-economic benefits dren’s human capital in the for a successful voluntary return host country than economic to take place. The UNHCR, the migrants, which suggests main donor countries, and other that they put a strong effort national/international NGOs to align their skill sets with also have to play important roles the skill requirements of the in the whole process to surmount host country; and they have the potential bottlenecks. lower propensity to accumu- This chapter provides several late savings and send their lessons to be learned in terms of money back to home country the link between the economic as remittances, so their poten- consequences of violent conflict, tial contribution in terms of refugee behavior, and repatria- reconstruction after conflict tion. Several countries in the re- may also be limited. These gion have experienced civil con- behavioral patterns should be flict, but the Syrian experience considered in designing re- is unique in the sense that Syria patriation programs, as post- has been the source country for a conflict trauma experienced huge refugee wave. A healthy re- by refugees and superior eco- construction of the Syrian econo- nomic opportunities offered my may not be possible without in host countries may make the successful design and imple- refugees (especially the high- mentation of a feasible repa- skilled ones) less willing to triation plan. Reconstruction of return. The economic oppor- physical and human capital are tunities offered by a potential complementary to each other, return to the home country which suggests that long-term become crucial at this point. political, institutional, and eco- 2. But the economic condi- nomic stability, and the prosper- tions in conflict countries do ity of Syria may not be achieved not look very prospective. without actively involving Syr- Satellite-based estimates sug- ian people in the reconstruction gest that conflict has severely process. The following main destroyed economic activ- policy lessons and recommenda- ity, and the size of economic tions arise from the discussion destruction is very large in provided in this chapter. Syria. This implies that the repatriation programs should focus on systematic and large- Chapter 4 | 131

scale economic reconstruction support, and employment plans – which should natu- subsidy programs (which are rally embed institutional up- shown in the literature to be grading. extremely useful). These les- 3. Investment in the quantity sons should be taken into ac- and quality of public capital/ count in designing repatria- infrastructure would play im- tion programs for the Syrian portant roles in reconstruct- refugees in the future. ing the conflict countries. 5. Finally, the Syrian refugees Moreover, private capital in- are now in the process of vestment could substantially integrating into the socio- boost economic growth rates economic life of their host in the short term. Conditional countries. Developed coun- on achieving long-term po- tries have been implement- litical stability and peace, ing financial and technical surpassing the economic per- support programs to facilitate formance of the pre-conflict the integration process. For example, Turkey has been era and rapid convergence to implementing large-scale EU- the hypothetical ‘no-conflict funded programs to integrate path’ need not be a remote Syrian children into the Turk- possibility, if feasible and ish public education system. appropriately designed re- The overall enrollment rate construction and repatriation is above 60% (including all programs are implemented in levels and grades) and the conflict countries. enrollment rate is expected to 4. Existing experiences suggest increase rapidly further into that most of the repatriation the near future. Host coun- efforts have been less suc- tries are now discussing how cessful than expected due to to provide official work per- various problems, including mits and even to provide citi- but not limited to: (i) lack of zenship selectively to Syrian understanding of the cultural, refugees. As the resolution religious, and historical con- of conflict in Syria is delayed, text; (ii) lack of coordination those integration efforts will between governments and proceed further, and the Syr- NGOs; (iii) lack of anticipa- ian refugees may become less tion of the recurrence of con- likely to participate in volun- flict in the post-repatriation tary repatriation plans. Inter- period; (iv) failing to address national cooperation is need- the issues of past property ed to end the conflict in Syria and land ownership; and (v) (and other conflict countries) lack of supporting the repa- and to implement well-de- triation programs with ap- signed repatriation programs propriately designed micro- rapidly in order to expedite credit assistance, housing economic reconstruction. 132 | Repatriation: potential contributions to post-conflict economic development

References comparison group matching and relative skill composi- Akgunduz, Y. E., van den Berg, tion,’ European Economic Re- M., and Hassink, W. (2018). view, 98, 282-315. ‘The impact of the Syrian Balkan, B., Ozcan-Tok, E., Torun, refugee crisis on firm entry H., and Tumen, S. (2018). ‘Im- and performance in Turkey,’ migration, housing rents, and World Bank Economic Review, residential segregation: Evi- 32, 19-40. dence from Syrian refugees Akgunduz, Y. E., and Torun, H. in Turkey,’ IZA Discussion (2018). ‘Two and a half mil- Paper #9853. lion Syrian refugees, skill mix Balkan, B. and Tumen, S. (2016). and capital intensity,’ GLO ‘Immigration and prices: Discussion Paper #186. Quasi-experimental evidence Aksoy, C. G. and Tumen, S. from Syrian refugees in Tur- (2019): ‘Local governance key,’ Journal of Population Eco- and the environmental cost nomics, 29, 657-686. of forced migration,’ unpub- Bauer, T. K. and Sinning, M. G. lished manuscript, European (2011). ‘The savings behavior Bank for Reconstruction and of temporary and permanent Development (EBRD). migrants in Germany,’ Jour- Alesina, A., Michalopoulos, S., nal of Population Economics, 24, and Papaioannou, E. (2016). 421-449. ‘Ethnic inequality,’ Journal of Calderon, C. and Serven, L. Political Economy, 124, 428-488. (2014). ‘Infrastructure, Altindag, O. and Kaushal, N. growth, and inequality,’ (2017). ‘Do refugees impact World Bank Policy Research voting behavior in the host Working Paper #7034. country? Evidence from Syri- Carrington, W. J. and De Lima, P. an refugee inflows in Turkey,’ J. (1996). ‘The impact of 1970s IZA Discussion Paper #10849. repatriates from Africa on the Assaad, R., Ginn, T., and Saleh, Portuguese labor market,’ In- M. (2018). ‘Impact of Syrian dustrial and Labor Relations Re- refugees in Jordan on educa- view, 49, 330-347. tion outcomes for Jordanian Ceritoglu, E., Gurcihan-Yuncul- youth,’ ERF Working Paper er, H. B., Torun, H., and Tu- #1214. men, S. (2017). ‘The impact of Aydemir, A. B. and Kirdar, M. G. Syrian refugees on natives’ (2017). ‘Quasi-experimental labor market outcomes in impact estimates of immi- Turkey: Evidence from a qua- grant labor supply shocks: si-experimental design,’ IZA The role of treatment and Journal of Labor Policy, 6, 1-28. Chapter 4 | 133

Ceylan, E. S. and Tumen, S. Devadas, S., Elbadawi, I. A., and (2018). ‘Measuring economic Loayza, N. V. (2019). ‘Growth destruction in Syria from after war in Syria,’ World outer space,’ The Forum: ERF Bank Policy Research Work- Policy Portal, Blog article, 19 ing Paper #8967. June 2018, https://theforum. Devadas, S. and Pennings, S. M. erf.org.eg/2018/06/19/mea- (2019). ‘Assessing the effect of suring-economic-destruction- public capital on growth: An syria-outer-space/. extension of the World Bank Ceylan, E. S. and Tumen, S. Long-Term Growth Model,’ (2019). ‘Estimating the cost of Journal of Infrastructure, Policy, conflict in the MENA region and Development, 3, 22-55. with satellite data: Evidence Domingues Dos Santos, M. and from Iraq, Libya, Syria, and Postel-Vinay, F. (2003). ‘Mi- Yemen,’ unpublished manu- gration as a source of growth: script, TED University. The perspective of a develop- Chen, S., Loayza, N. V., and Rey- ing country,’ Journal of Popula- nal-Querol, M. (2008). ‘The tion Economics, 16, 161-175. aftermath of civil war,’ World Dustmann, C. (2008). ‘Return Bank Economic Review, 22, 63- migration, investment in chil- 85. dren, and intergenerational Clemens, M. A., Lewis, E. G., and mobility: Comparing sons Postel, H. M. (2018). ‘Immi- of foreign- and native-born gration restrictions as active fathers,’ Journal of Human Re- labor market policy: Evidence sources, 43, 299-324. from the Mexican bracero ex- Dustmann, C., Fadlon, I., and clusion,’ American Economic Weiss, Y. (2011). ‘Return mi- Review, 108, 1468-87. gration, human capital ac- Cortes, K. E. (2004). ‘Are refu- cumulation, and the brain gees different from economic drain,’ Journal of Development immigrants? Some empirical Economics, 95, 58-67. evidence on the heterogene- Dustmann, C. and Görlach, J.- ity of immigrant groups in S. (2016). ‘The economics of the United States,’ Review of temporary migrations,’ Jour- Economics and Statistics, 86, nal of Economic Literature, 54, 465-480. 98-136. Del Carpio, X. and Wagner, M. Dustmann, C. and Mestres, J. (2015). ‘The impact of Syrian (2010a). ‘Remittances and refugees on the Turkish labor temporary migration,’ Journal market,’ Unpublished manu- of Development Economics, 92, script, World Bank. 62-70. 134 | Repatriation: potential contributions to post-conflict economic development

Dustmann, C. and Mestres, J. Harild, N., Christensen, A., and (2010b). ‘Savings, asset hold- Zetter, R. (2015). ‘Sustain- ings, and temporary migra- able refugee return: Triggers, tion,’ Annals of Economics and constraints, and lessons on Statistics, 97/98, 289-306. addressing the development Eastmond, M. and Ojendal, J. challenges of forced dis- (1999). ‘Revisiting a ‘repa- placement,’ The World Bank triation success’: The case of Group. Washington, DC. Cambodia,’ In: R. Black and Henderson J. V., Storeygard, A., K. Koser (eds): The end of the and Weil, D. N. (2012). ‘Mea- refugee cycle: Refugee repa- suring economic growth from triation and reconstruction. outer space,’ American Eco- Berghan Books. nomic Review, 102, 994-1028. ESCWA (2018). ‘Syria at War Hevia, C. and Loayza, N. V. – Seven Years On (The Eco- (2012). ‘Savings and growth nomic Implications of the Middle East Develop- Conflict in Syria),’ Presenta- in Egypt,’ tion at Consultation Meeting ment Journal, 4, 1-23. on Policy Alternatives for Hodler, R. and Raschky, P. A. Post-Conflict Reconstruction (2014). ‘Regional favoritism,’ and Economic Recovery. Na- Quarterly Journal of Economics, tional Agenda for the Future 129, 995-1033. of Syria. August. United Na- HRW (2005). ‘Coming home: tions Economic and Social Return and reintegration in Commission for Western Angola,’ March 2005, Vol. Asia. 17, No.2 (A), http://www. Fallah, B., Krafft, C., and Wah- hrw.org/reports/2005/ango- ba, J. (2019). ‘The impact of la0305/angola0305.pdf. refugees on employment and Hunt, J. (1992). ‘The impact of the wages in Jordan,’ Journal of 1962 repatriates from Algeria Development Economics, forthcoming. on the French labor market,’ Industrial and Labor Relations Fransen, S. and Kuschminder, K. Review, 45, 556-572. (2012). ‘Back to the land: The long-term challenges of refu- Janossy, F. (1969). ‘The end of the gee return and reintegration economic miracle: Appear- in Burundi,’ UNHCR: New ance and reality in economic Issues in Refugee Research, development,’ ISAP: New Research Paper #242. York, NY. Funkhouser, E. (1995). ‘Remit- Khan, A. H. (1997). ‘Post-migra- tances from international mi- tion investment in education gration: A comparison of El by immigrants in the United Salvador and Nicaragua,’ Re- States,’ Quarterly Review of view of Economics and Statistics, Economics and Finance, 37, 285- 77, 137-146. 313. Chapter 4 | 135

Lee, J., Peri, G., and Yasenov, V. OECD (2008). ‘International Mi- (2017). ‘The employment ef- gration Outlook 2008,’ Wash- fects of Mexican repatriations: ington, DC and Paris: OECD. Evidence from the 1930’s,’ Omata, N. (2012). ‘Struggling to NBER Working Paper #23885. find solutions: Liberian refu- Loayza, N. V. and Pennings, S. gees in Ghana,’ Research Pa- (2018). ‘The long‐term growth per #234, New Issues in Refu- model,’ The World Bank, gee Research, UNHCR. Washington, D.C. http:// Organski, A. F. K. and Kugler, www.worldbank.org/LTGM. J. (1977). ‘The costs of major Mamei, M., Cilasun, S. M., Lippi, wars: The Phoenix Factor,’ M., Pancotto, F., and Tumen, American Political Science - Re S. (2019). ‘Improve education view, 71, 1347-1366. opportunities for better integration of Syrian refugees Organski, A. F. K. and Kugler, J. in Turkey,’ In: Guide to Mo- (1980). ‘The war ledger,’ Chi- bile Data Analytics in Refu- cago, IL: University of Chi- gee Scenarios—The ‘Data for cago Press. Refugees Challenge’ Study. Pinger, P. (2010). ‘Come back or Eds. Salah, A. A., Pentland, stay? Spend here or there? A., Lepri, B., and Letouzé, E., Return and remittances: The Springer. case of Moldova,’ International Merkle, L. and Zimmermann, K. Migration, 48, 142-173. (1992). ‘Savings, remittances, Rapoport, H. and Docquier, F. and return migration,’ Eco- (2006). ‘The economics of nomics Letters, 38, 77-81. migrants’ remittances,’ Hand- Michalopoulos, S. and Papaioan- book of the Economics of Giving, nou, E. (2013a). ‘Pre‐colonial Altruism, and Reciprocity: Appli- ethnic institutions and con- cations, Volume 2. Eds. Kolm, temporary African develop- S.-C. and Ythier, J. M., pp. ment,’ Econometrica, 81, 113- 1135-1198, Elsevier, Amster- 152. dam and Boston. Michalopoulos, S. and Papaio- Rohner, D., Thoenig, M., and annou, E. (2013b). ‘National Zilibotti, F. (2013). ‘Seeds of institutions and subnational distrust: Conflict in Uganda,’ development in Africa,’ Quar- Journal of Economic Growth, 18, terly Journal of Economics, 129, 151-213. 217-252. Milionis P. and Vonyo, T. (2015). Smolny, W. (2000). ‘Post-war ‘Reconstruction dynamics: growth, productivity conver- The impact of World War gence, and reconstruction,’ II on post-war economic Oxford Bulletin of Economics and growth,’ American Economic Statistics, 62, 589-606. Association: 2016 Annual Conference. 136 | Repatriation: potential contributions to post-conflict economic development

Spilimbergo, A. (2009). ‘Democ- Tumen, S. (2019). ‘Refugees and racy and foreign education,’ ‘native flight’ from public to American Economic Review, 99, private schools,’ Economics Let- 528-543. ters, forthcoming. Storeygard, A. (2016). ‘Farther UNHCR (2004). ‘Afghanistan. on down the road: Transport Challenges to Return,’ United costs, trade and urban growth Nations High Commissioner in sub-Saharan Africa,’ Review for Refugees, Afghanistan. of Economic Studies, 83, 1263- UNHCR (2015). ‘Country Op- 1295. erations Profile 2015 Iraq,’ Toal, G., and Dahlman, C. (2011). http://www.unhcr.org/ ‘Bosnia remade: Ethnic pages/49e486426.html cleansing and its reversal,’ UNOCHA (2014). ‘South Sudan Oxford University Press, Ox- crisis situation report as of 13 ford, UK. February 2014,’ Report #19. Tumen, S. (2016). ‘The economic World Bank (2017). ‘The Toll of impact of Syrian refugees on War – The Economic and So- host countries: Quasi-exper- cial Consequences of the Con- imental evidence from Tur- flict in Syria,’ Washington key,’ American Economic Re- DC: World Bank. view, 106, 456-460. Tumen, S. (2018). ‘The impact of low-skill refugees on youth education,’ IZA Discussion Paper #11869. CONCLUSIONS

Repatriation of refugees from the Arab conflicts: prospects and potential political and economic implications

Samir Makdisi

The major problem for repa- in those countries where persis- triation of refugees is the persis- tent civil conflicts have forced tence of the conflicts from which millions to become refugees, they have fled. These continue mostly in neighboring countries in varying forms and at varying but also beyond. intensities in the four countries Given current conditions, vol- discussed in this Report – Iraq, untary repatriation – as defined Libya, Syrian, and Yemen. Final by Article 33 of the 1951 UN settlements of the conflicts that Refugee Convention (voluntary, may lead to national reconcili- safe, and dignified return) – is ation are yet to be reached. The not commonly feasible. The rea- fact that all these conflicts have sons are the lack of robust and involved extensive direct and credible political settlements of indirect foreign interventions – the conflicts that, among other both from other Arab countries things, would guarantee the and elsewhere in the world – micro security of returnee indi- has not only made their resolu- viduals and communities as well tion more difficult but it has also as both their human rights and made the nature of any settle- their property rights. ment uncertain. In addition, there is an ab- When the Arab uprisings sence of a substantial recon- broke out in 2011, hopes were struction agenda, which, in turn, raised that they would lead to hinges on the achievement of the establishment of democratic meaningful political settlements. regimes in the region and open The relatively limited voluntary the door for broad-based devel- repatriation of certain refugees opment. Yet as of mid-2019, with has been made under specific the exception of Tunisia, these conditions that permitted their hopes have not been realized. It return. remains to be seen whether any With these considerations in final settlements could lead to mind, this Report reaches five genuine democracy, particularly main conclusions. 137 138 Conclusions

Refugees from Arab con- very poor living conditions with flicts: scale and character- a very high unemployment rate istics among those able to work. The prospects for potential integration in their host countries are First, the conflicts in the re- very low. gion (in Syria and Yemen prin- cipally, but also in Libya and By contrast, Syrian refugees Iraq) have caused millions to in Europe are better educated flee their homeland or forced and more economically able. them to be internally displaced Should they decide to settle in within their own countries. It is the countries of their refuge, the estimated that as of mid-2019, 6.4 prospects are probably high. million refugees had fled from Syria with the majority settling in neighboring countries. Turkey Incentives for refugees to has received half of them, while return Lebanon and Jordan have a combined share of about 2.4 million. Second, most refugees would About a million refugees have opt to return to their home coun- settled in Europe, mainly in Ger- tries but there are multifaceted many and Sweden. conditions that govern their re- In Yemen, over two million turn, security at home being one have experienced internal dis- of the most important. At the placement, mostly in Houthi- same time, in deciding on return, controlled territories with spill- refugees take account of the po- overs into neighboring African litical, economic, and social fac- countries. Between 2014 and tors in both the host and home 2018, over three million Iraqis countries. Interestingly, there is suffered internal displacement, also a gender dimension to repa- while over 280,000, mostly wom- triation in that men and women en and children, were forced to have different priorities and act seek refuge in neighboring coun- differently. tries, mainly Turkey. Estimates Restoration of security in cer- for forced internal displacements tain parts of a country in conflict, in Libya are not available. as in Syria, has induced a lim- Refugees who have settled ited return of refugees who hail in Jordan, Lebanon, and Turkey from these parts. But a major is- have mostly come from rural and sue here is that to the extent that poorer areas in Syria. Over half a political settlement has not yet of them are children and most of been reached, this type of return the adults are not well educated. may not be sustainable: the con- Poverty is widespread. It is es- flict may erupt again in areas timated that half of the Syrian considered to be safe or secured refugee households suffer from by the state. Chapter 5 | 139

A wider consideration in this of prospects for reintegrating in regard is the nature of a politi- the communities they had left cal settlement and specifically could influence the decisions of whether it is politically inclusive refugees considering repatria- and could lead to democratic tion. forms of governance. Refugees The considerations outlined may view returning home as a above indicate clearly that in political act equivalent to rec- the absence of a genuine peace, ognizing the legitimacy of the repatriation of refugees faces regime in the home country. Un- many uncertainties. While the less they feel they will be protect- host countries are nonetheless ed by the state, they may opt not eager to push in this direction, to return and contribute to the the voluntary return of refu- rebuilding process. gees will continue to be severely Equally, refugees’ perceived constrained as a consequence of economic interests play a role in these considerations. their decision to return or not – Of those who did return vol- that is, their expectations of the untarily, available UNHCR in- conditions of livelihood in their formation (2017) on Syrian early home country and of oppor- repatriates (about 3% of the total) tunities for work and access to indicates that 58% of the return- social services. Here the gender ees are men citing joining family issue enters the picture. House- members as the main reason for holds headed by women tend return. This is followed by, first, to be more concerned with the the deteriorating conditions in provision of services, especially the country of asylum; second, schools and hospitals for their an improved security situation children, while men are more back home; and finally, the desire concerned with the availability to work.1 The returnees mostly of work opportunities and access hailed from Turkey (40%), fol- to livelihood. lowed by Lebanon (36%), and a Other considerations that in- small number from Iraq, Jordan, fluence the decision of refugees and Egypt. to return or not include family According to a 2019 World circumstances in exile and in the Bank study, demographic vari- home country. For example, the ables that are positively corre- presence of family back home lated with return include being can act as an important pull for single, male, and less educated. refugees; but if their children are Of the pull factors, the most im- in school in the host country, this could delay the decision to re- 1 Available data on the repatriation of ref- turn. Furthermore, consideration ugees from the other countries in conflict is fragmentary and unreliable. 140 Conclusions

portant is security and then avail- identifying trusted community able services, such as education mediators. and health facilities. Push fac- In practice, while the emerg- tors include efforts undertaken ing post-conflict regime is nor- by countries of asylum to induce mally the outcome of national premature return, especially in reconciliation of the main parties Lebanon, a hostile environment to the conflict, it may continue to and unsustainable living condi- exhibit strong elements of clien- tions. telism. In other words, post-conflict governance may fall short of the ideal inclusivity that would serve the common interest rather Rebuilding the social con- than catering to a client-patron tract between refugees and relationship. the post-conflict regime Looking at the MENA region, past political settlements, such as Third, whatever motives those following the Algerian and drive refugees to go back home, Lebanese civil wars (in 2002 and their return should be viewed as 1990 respectively), are a good il- a process of political rapproche- lustration. In the former case, ment between citizen, commu- the settlement was credible but nity, and state, which calls for incomplete: a strong state grip the rebuilding of the social con- by a group of elites of economic tract between the refugee and the resources (mainly oil) and po- state. This implies proper recog- litical power created a clientelist nition of the political roots of the structure, and transitional justice armed conflict and not only its was not enforced. In the latter humanitarian dimension, and of case, the settlement known as the the need for safety and security Taif Agreement only managed to be guaranteed through a polit- to realign the political shares of ical process that creates inclusive religious sects in governance (the governance mechanisms, ends so-called ‘consociational democ- criminal impunity, and facilitates racy’) but did not eliminate ex- reintegration, demilitarization, isting inequality in citizens’ civil and access to justice. rights, as had been hoped. The more genuinely inclu- While a large number of sive the post-conflict regime, the Lebanese citizens who left the greater the incentive for refugees country during the civil war did to return. While this process may return once a political settlement take time given the varying post- was reached, a good number opt- conflict scenarios, efforts to pre- ed not to return, especially those pare refugees for a return have who had emigrated to the United to begin in advance: for example, States. providing legal assistance and Chapter 5 | 141

For the countries currently in struction, then Syria would be conflict, effective repatriation of able to surpass its 2010 GDP and refugees presents a huge chal- GDP per capita levels in 10 years’ lenge. It requires having sustain- time. able political settlements leading But of course, this optimistic to the establishment of demo- scenario may not emerge. An cratic governance that in turn alternative pessimistic scenar- would support the safe return io, which embeds the assump- of refugees and open the door to tions of limited guarantees for their participation in planned re- micro-security/property rights, construction projects. low reconstruction funds, and low investment, suggests that it would take two or three decades for Syria to catch and surpass its Resources for reconstruc- 2010 GDP levels. tion These findings imply that high post-conflict economic per- Fourth, in drawing up re- formance is possible in Syria if construction plans, we should long-term political stability is keep in mind that key drivers achieved, market-friendly mech- of growth in low- and middle- anisms are redesigned to obtain income countries include pub- efficient allocation of resources lic capital, public infrastructure throughout the economy, appro- investments, private capital in- priately designed reconstruction vestments, human capital, total and repatriation programs are factor productivity, demograph- implemented, and those pro- ics, and labor market outcomes. grams are supported by sustain- Making available the resources able financing facilities. needed to finance the required Remittances of Syrians (basi- investment in all of these ar- cally of those in the process of in- eas, as well as the workforce to tegrating into the socio-economic implement them, is a huge chal- life of their host countries) could lenge facing post-conflict recon- be a major channel through struction plans. Foreign financ- which migration may influence ing will play a crucial role, as too the wellbeing of individuals at will potentially returning capital home. as well as remittances from those staying abroad. As for repatriation, if returning migrants are highly skilled, Considering specifically the then repatriation has a clear case of Syria, empirical work positive impact on the economic shows that if there were a large outcomes in the home country. amount of reconstruction assis- Further return migrants would tance and repatriation forthcom- contribute to reconstructing ing, and hence high investment national institutions and raise within the first decade of recon- 142 Conclusions

their quality. On the other hand, of refugees. This would include, one should bear in mind that the among others delineated in the repatriation movement could conclusions above, a lack of co- lead to a limited decline in wages ordination between the parties and a small increase in the unem- concerned with repatriation, a ployment rate in the short run. failure of the home country to address issues of property rights, and a lack of sufficient international support for repatriation New models of reconstruc- programs. tion and repatriation Past repatriation and reconstruction efforts were intended Fifth, a cautionary conclusion to address specific regional or is in order. It is not surprising country-specific concerns. But that, so far, repatriation efforts while the effects of existing Arab have been much less success- conflicts and their associated ref- ful than expected. As indicated ugee problems remain basically above, lack of security in the regional concerns, they have home countries and anticipa- now spilled beyond the region tion on the part of the refugees to become a global problem. This of recurring conflicts is a major implies the need for coordinated explanatory factor. Until hostili- policy action between all the gov- ties are permanently ended and ernments and international orga- a genuine national reconciliation nizations concerned to reach an is achieved in the countries suf- all-encompassing solution that fering from conflict, particularly accounts for this new dimension Syria and Yemen, a sustained of the refugee problem. program of repatriation is clearly Past regional models and/or not expected to succeed to any individual country efforts may great extent. no longer be useful or sustain- But even when peace is re- able. Looking forward to the stored, it is essential for the post-conflict phase, new recon- parties concerned (national struction models and institu- governments and international tions would need to be designed organizations), when designing to avoid the inefficiencies of the future repatriation programs of past and lay the foundations for refugees, to address other hin- sustainable programs of repa- drances that have been associat- triation. ed with the so far limited return This report discusses the issue of repatriation of refugees in impacted countries of the South Med region. Through its four chapters, the authors start by looking into the characteristics of these refugees and the conditions affecting their decisions to return. This overview is followed by an analysis of the possible political settlement scenarios and reconstructions’ potentials, with a focus on the possible role of the international community. The authors then analyse the economic costs of conflicts as well as post-conflict growth scenarios. The report concludes by highlighting the main findings and providing policy insights into how to address this issue to ensure a safe, sustainable and dignified return of refugees to their home countries.