THE TOLL OF THE ECONOMIC AND SOCIAL CONSEQUENCES OF THE CONFLICT IN SYRIA THE TOLL O F WA R THE ECONOMIC AND SOCIAL CONSEQUENCES OF THE CONFLICT IN SYRIA THE TOLL O F WA R THE ECONOMIC AND SOCIAL CONSEQUENCES OF THE CONFLICT IN SYRIA © 2017 The World Bank. does not guarantee the accuracy of the data included in this work. The boundaries, colors, on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Rights and Permissions license, you are free to copy, distribute, transmit, and adapt this work, including for commercial Third-party content—The World Bank does not necessarily own each component of the content contained within the work. The World Bank therefore does not warrant that the use of any third-party-owned individual component or part contained in the work will not infringe on the you. If you wish to re-use a component of the work, it is your responsibility to determine whether permission is needed for that re-use and to obtain permission from the copyright owner. Examples of components can include, but are not limited to, tables, figures, or images. With support from: Front cover: A Syrian woman waits with her daughter near the Syria border at the southeastern town of Suruc ACKNOWLEDGMENTS This study was prepared by a multidisciplinary team led by Harun Onder (Senior Economist, TTL) and Faya Hayati (Young Professional, TTL) comprising Haneen Ismael Sayeed (Program Leader, HDN Group Coordinator), Ibrahim Khalil Dajani (Program Leader, SDN Group Coordinator), Peter Mousley (Program Leader, EFI Group Coordinator), Aakanksha Pande (Senior Health Economist), Abdulhakim Al-Aghbari (Senior Transport Specialist), Amal Talbi (Senior Water and Sanitation Specialist), Anders Jagerskog (Senior Water Resources Management Specialist), Andrea Petrelli (Consultant), Andrew James Judaprawira (Consultant), Angela Elzir (Senior Professional Associate), Angela S. Lisulo (Consultant), Anya Vodopyanov (Consultant), Asbjorn Wee (Senior Operations Officer), Ashwaq Maseeh (Consultant), Barbro Hexeberg (Lead Economist), Bartol Letica (Senior Financial Sector Specialist), Bekzod Shamsiev (Senior Agriculture Economist), Burton Bollag (Consultant), Carlo Rossotto (Lead ICT Policy Specialist), David Andrew Stephan (Economist), David Robalino (Manager), Dhiraj Sharma (Young Professional), Dino Merotto (Lead Economist), Dominick Waal (Senior Economist), Dorte Verner (Lead Agriculture Economist), Ece Ozcelik (Consultant), Ejona Fuli (Consultant), Erhan Artuc (Senior Economist), Eric Roland Metreau (Senior Economist), Eyup Ozveren (Consultant), Hiromi Yamaguchi (Analyst), James Fearon (Consultant), Jan Van Der Goltz (FCV and Jobs Economist), Joern Huenteler (Young Professional), Jose Valdes Martinez (Senior Economist), Joy Aoun (Disaster Risk Management Specialist), Israel Osorio-Rodarte (Consultant), Kent Garber (Consultant), Kinley Salmon (Young Professional), Kirk Hamilton (Consultant), Kohei Ueda (Consultant), Lina Fares (Senior Procurement Specialist), Lisa Stewart (Consultant), Manal Quota (Education Specialist), Mohamed Ali Maourani (Consultant), Nabila Assaf (Senior Economist), Natasha Maherali (Consultant), Paul Prettitore (Senior Public Sector Specialist), Peter McConagy (Financial Sector Specialist), Phoram Shah (Urban Specialist), Rabie Nasr (Consultant), Raja Arshad (Lead Disaster Risk Management Specialist), Raffi Garcia (Consultant), Ray Salvatore Jennings (Consultant), Sally Zgheib (Water Supply and Sanitation Specialist), Shahrzad Mobasher Fard (Consultant), Soraya El Khalil (Research Analyst), Syed Mehdi Hassan (Lead Financial Sector Specialist), Tahir Akbar (Senior Disaster Risk Management Specialist), Tobias Schneider (Consultant), Tracy Hart (Senior Environmental Specialist), Trang Thuy Nguyen (Research Analyst), Tulu Balkir (Consultant), Usaid El Hanbali (Consultant), Waad Tammaa (Consultant), Zaki Mehchy (Consultant), and Ziad Nakat (Senior Transport Specialist), and under the guidance of Eric Le Borgne (Acting Practice Manager, GMF05). The team is grateful for guidance and support from Hafez M. H. Ghanem (Vice President, MNAVP), Ferid Belhaj (Chief of Staff, EXC), Saroj Kumar Jha (Senior Director, GCFDR), Shantayanan Devarajan (Chief Economist, MNACE), Kanthan Shankar (Acting Country Director, MNC02), Claire Kfouri (Country Program Coordinator, MNC02), and Janet Minatelli (Senior Country Officer). The team has received valuable comments from our peer reviewers for the final report, Abdallah Al Dardari, Alan Gelb, Tito Cordella, and Xavier Devictor, and for the concept note, Andrew Harper, Joe Saba, Jose Araujo, and Paolo Verme, as well as from the following colleagues: Andrew C. Kircher, Hideki Matsunaga, Lili Mottaghi, Paul Bance, Omer Karasapan, Sajjad Ali Shah, Siobhan McInerney-Lankford, Stefanie Onder, William Stebbins, and Yezid Sayigh. The team also received excellent support from Katia Stefanova and Syvie Creger. Finally, generous financial support from the State and Peacebuilding Fund, technical and financial support from the European Space Agency EO4SD agriculture and rural development activity cluster (provided by eLEAF and GeoVille), and technical support from various UNESCWA, UNHABITAT, UNHCR, IOM, FAO, WFP, SCPR, and REACH Initiative are gratefully acknowledged. TA B L E O F C O N T E N T S : i Key Findings iii Executive Summary 1 CHAPTER 1: Syria Before the Conflict 3 The Long History of Syria 8 Syria at the Dawn of the Conflict 17 CHAPTER 2: The Wrath of Conflict 18 Physical Damage 47 Demographic Dispersion and Loss of Lives and Opportunities 54 Effects of the Conflict on Economic Activity 68 Human Development Outcomes 77 CHAPTER 3: Disentangling the Economic Impact of the Conflict 79 Aggregate Impact 85 Integrative Framework 101 Concluding Remarks Appendices 104 A. Methodology of Physical Damage Assessments 107 B. Sectoral Situation Reports 111 C. Technical Specifications of the Simulation Model 119 References LIST OF FIGURES i Figure ES.1: Organization of the study 10 Figure 1.1: Labor Force Participation Rate, 2010 10 Figure 1.2: Structure of the Working-Age Population in Syria, 2010 15 Figure 1.3: Governance Indicators in Syri 33 Figure 2.1: Sources of Electricity Supply in Three Conflict-Affected Cities 35 Figure 2.2: Damages to Intra-City Roads 37 Figure 2.3: Changes in Connectivity: Pre-Conflict versus De Facto Distances, March 2017 39 Figure 2.4: Damage to Total Health Facilities, by City 40 Figure 2.5: Damage to Health Facilities, by Facility Type, February 2017 51 Figure 2.6: Destination and Origin Governorates of IDPs 51 Figure 2.7: Direction of Flight for the Forcibly Displaced, by Governorate 55 Figure 2.8: Economic Activity during the Years of the Conflict 62 Figure 2.9: Investment and Trade Dynamics 64 Figure 2.10: Fiscal Deficit and Debt 66 Figure 2.11: Exchange Rates and the Consumer Price Index 70 Figure 2.12: Labor Market Indicators: Pre-Conflict versus Conflict 71 Figure 2.13: Structure of the Working-Age Population in Syria, 2015 72 Figure 2.14: Simulated Poverty Rate With and Without the Conflict, 2007–16 (%) 73 Figure 2.15: Food and Non-Food Foreign Aid Flows in Syria 74 Figure 2.16: Trends in Reported Funding, by Source, 20017–16 (US$, millions) 80 Figure 3.1: Conceptualization of Pre-Shock, Post-Shock, and Counterfactual Relationship 81 Figure 3.2: Comparing the Counterfactual Real GDP per Capita by Different Methodologies with Actual Estimates 83 Figure 3.3: Actual and Counterfactual GDPs and Cumulative GDP Losses 84 Figure 3.4: Indifference Map for Income and Life Expectancy in Syria 87 Figure 3.5: Simulation Diagram 93 Figure 3.6: Dynamics of Impact under Various Shock Scenarios 98 Figure 3.7: Dynamics of Impact under Different Migration Restriction Scenarios 100 Figure 3.8: Dynamics of Impact under Various Conflict End-Date Scenarios L I S T O F TA B L E S 11 Table 1.1: Incidence of the Abject, Extreme, and Overall Poverty Rates and the Gini Coefficient, 1997–2007 12 Table 1.2: Incidence of Extreme Poverty, by Region, 1997, 2004, and 2007 21 Table 2.1: Housing Damage in Cities 26 Table 2.2: Housing Damage Estimates across Governorates 30 Table 2.3: Damage in Water and Sanitation Infrastructure (Eight Governorates) 32 Table 2.4: Status of the Electricity Grid and Major Power Infrastructure in Seven Cities in Syria (%) 48 Table 2.5: Key Demographic Indicators 50 Table 2.6: Total Registered Syrian Refugee Population, by Country/Region, 2012–16 52 Table 2.7: Push/Pull Factor Matrix for Aleppo, Homs, and Dar'a (1: Primary Driver, 2: Secondary Driver, 3: Tertiary Driver) 58 Table B2.2.1: Sample of ISIS Crude Oil Production 61 Table 2.8: Damage by industrial area in nine cities 89 Table 3.1: Structure of the Simulation-Based Analysis 92 Table 3.2: Combined and Isolated Impacts of Shocks 96 Table 3.3: Economic Outcomes under Risk Neutrality 99 Table 3.4: Economic Outcomes under Various Conflict End-Date Scenarios 104 Table A.1: Number of Facilities Assessed per Sector, by City 106 Table A.2: Conflict-Induced Damage Inference Matrix 107 Table B.1: Situation Report: Water and Sanitation Infrastructure 108 Table B.2: Situation Report: Power Infrastructure 109 Table B.3: Situation Report: Health Care Facilities 110 Table B.4: Situation Report: Education Facilities LIST OF MAPS 3 Map 1.1: Land Utilization in Syria 5 Map 1.2: The Ottoman Administration of Geographical Syria 6 Map 1.3: Caravan Routes, 16th to 20th Centuries 19 Map 2.1: Incidence of Conflict, 2011–17 22 Map 2.2: Neighborhood-Level Housing Damage 24 Map 2.3: Governorate-Level Estimates of Housing Damage 27 Map 2.4: Reconstruction, by Governorate and District 28 Map 2.5: Calculated Debris Accumulation, as of February 2017 32 Map 2.6: Estimated Damage to Major Power Sector Assets 35 Map 2.7: Major Roads and Railways 42 Map 2.8: Governorate-Level Damage to the Health Sector 45 Map 2.9: Governorate-Level Damage to Syria’s Education Sector Facilities 49 Map 2.10: Primary Reported Routes through Turkey and the "Balkan Corridor" (as of September 2015) 57 Map B2.2.1: Oil Economy during Conflict 60 Map B2.3.1: Study Areas Included in the Demonstration 60 Map B2.3.2: Change in Vegetation Cover for February in March (Al Hassakeh) from Pre-Conflict in 2011 to Conflict in 2016 75 Map 2.11: Number of Beneficiaries, by Sub-District (December 2016) LIST OF BOXES 4 Box 1.1: Key Historical Dates 49 Box 2.1: The Terminology of Forced Displacement 57 Box 2.2: Oil Production by Non-State Groups 59 Box 2.3: Effects of the Conflict on Syrian Agriculture: A Pilot Study 81 Box 3.1: Counterfactual GDP Calculations 88 Box 3.2: Simulation Model KEY FINDINGS • This study assesses the economic and social consequences of the Syrian conflict as of early 2017. The conflict has inflicted significant damage to the Syrian Arab Republic’s physical capital stock (7 percent housing stock destroyed and 20 percent partially damaged), led to large numbers of casualties and forced displacement (between 400,000 and 470,000 estimated deaths and more than half of Syria’s 2010 population forcibly displaced), while depressing and disrupting economic activity. From 2011 until the end of 2016, the cumulative losses in gross domestic product (GDP) have been estimated at $226 billion, about four times the Syrian GDP in 2010. • Disruptions in economic organization have been the most important driver of the conflict’s economic impact, superseding physical damage. The conflict has not only ended lives and destroyed productive factors; it has also severely diminished economic connectivity, reduced incentives to pursue productive activities, and broken economic and social networks and supply chains. Simulations show that cumulative GDP losses due to such disruptions in economic organization exceed that of capital destruction by a factor of 20 in the first six years of the conflict. This is mainly because a “capital destruction only” shock works like some natural disasters (unanticipated and sudden onset): in a well-functioning economy, its effects on investments are limited (only a 22 percent reduction in simulations). Thus, capital is rapidly rebuilt and further economic repercussions are contained. In comparison, disruptions in economic organization decrease investments significantly (a simulated 80 percent reduction) by reducing profitability; therefore, the initial effects are propagated strongly over time. This contrast between different types of shocks helps us to put future reconstruction efforts into perspective. Without rebuilding economic institutions and restoring economic networks, replacing the capital stock by itself will not go far in helping the economy recover. • The longer the conflict continues, the more difficult the post-conflict recovery will be. Although the rate of deterioration moderates over the course of the conflict, the effects become more persistent. Should the conflict end in its sixth year (baseline), GDP recoups about 41 percent of the gap with its pre-conflict level within the next four years. Overall, the cumulative GDP losses will reach 7.6 times the 2010 GDP by the 20th year. In comparison, GDP recoups only 28 percent of the gap in four years if the conflict ends in its 10th year (alternative scenario), and cumulative losses will be at 13.2 times the 2010 GDP. Simulations also show that outmigration could double between the sixth year of the conflict and the 20th year, in the case of a continued conflict. These results do not capture many other complications, like political economy challenges such as conflict-driven grievances. Adding these factors would only reinforce the main findings of the report: the longer the conflict persists, the deeper the grievances and divisions will run in the Syrian society, rendering it very difficult to build efficient institutions and effective economic mechanisms. THE TOLL OF WAR iii EXECUTIVE SUMMARY The Arab Spring protests marked the beginning of a new era in the Syrian Arab Republic in 2011. Minor public protests began almost immediately after the initial protests in Cairo in January 2011. The first large demonstrations began two months later in March, and the following months saw a process of escalation as demonstrations spread and increased in size within the country. By the summer of 2011, the armed conflict was already unfolding. Now in its sixth year, the Syrian conflict remains active and is bringing much pain and tragedy on a daily basis. This study provides an assessment of the conflict’s impact on economic and social outcomes in Syria as of early 2017. Conflicts destroy tangible and intangible assets and leave deep marks on a country’s social fabric, culture, and collective memories. The Syrian conflict has quickly become a particularly harsh example of this. It is, unfortunately, not possible to capture all of these consequences in a comprehensive manner. As of March 2017, the conflict was active, and some longer-term outcomes and political, social, security-related, and institutional effects were not yet observable beyond anecdotal evidence. In addition, because of the ongoing conflict, the country has remained inaccessible to the team who produced this report, and severe data shortages forced the team to leave out much-needed analysis. Given these constraints, the analysis focuses on taking stock of the effects of the conflict in four areas: (i) physical damage, (ii) loss of lives and demographic dispersion, (iii) economic outcomes, and (iv) human development outcomes. The strength of this study lies in its scope and methodology. The novel part of the data used in this study is provided by remote sensing assessments (with media verification), which focused on ten cities1 and six sectors2, and then extrapolated to eight governorates3 by using conflict intensity and asset base comparisons. Information from partner agencies in demography and sector and economic outcomes supplemented these assessments. The novelty of the analysis itself lies in its integrative framework. The results of the physical damage and loss of lives are used to employ an integrative model to match observed economic outcomes, including GDP patterns and demographic mobility (inter-governorate and outmigration from Syria). This approach provided several advantages. First, the distinct roles played by physical destruction, casualties, and economic disorganization in how conflict has manifested its impacts were separated. This separation is important because it could help prioritize actions in a future recovery and reconstruction process, by facilitating a comparison between alternative scenarios of aid composition. 1 These cities are Aleppo, Raqqa, Dar'a, Douma, Dayr az-Zawr, Homs, Hama, Idlib, Kobani, and Tadmur (also called Palmyra) which were chosen on the basis of several criteria, including conflict intensity, geographical representation, and being an important source or destination of internally displaced persons. 2 These sectors are education, energy, health, housing, transportation, and water and sanitation. 3 These governorates are Aleppo, Dar'a, Dayr az-Zawr, Hama, Homs, Idlib, Raqqa, and Rif Dimashq, which were chosen because of conflict intensity. iv Executive Summary Second, it helped compare the results of alternative scenarios that are not observable from actual data: for example, what is the role played by migration in the conflict’s economic impact? How far can the impact of the conflict on economic outcomes prevail after the cessation of the conflict? The report discusses these in detail. (Figure ES.1 shows the organization of the study.) Figure ES.1: Organization of the study Final Report Synthesis Report - Syria ESIA Background Papers Background Papers - Results by sector Integrative Model Integrative Framework Sectoral Macroeconomics Trade Environment Water & Sanitation Agriculture Poverty Education Health Reports Governance Finance Transport & ICT Energy & Extract Urban DRM Social Protection Labor & Jobs Crosscutting Demography National Accounts Conflict Analysis Studies Data Remote Sensing Partner Agency Data Public Data Sources Syria Before the Conflict On the eve of the 2011 unrest, Syria was a fast-growing, lower-middle-income country. In aggregate terms, the Syrian economy was improving, albeit starting from an unfavorable base, during the 2000s. Gross domestic product (GDP) grew at an average of 4.3 percent per year from 2000 to 2010 in real terms, which was almost entirely driven by growth in non-oil sectors, and inflation averaged at a reasonable 4.9 percent. The strong growth performance, however, did not translate into broad-based economic and political inclusion and further transparency and civil liberties. Syria was comparable to other regional economies in many socioeconomic indicators in 2010. The multidimensional poverty rate (5.5 percent) and income inequality (Gini index: 32.7) stood close to regional averages. However, the labor force participation (LFP) rate in Syria (43.5 percent) was one of the lowest in the world, low even by regional standards (the Arab Republic of Egypt and Tunisia had 49 and 47 percent LFP, respectively). This was primarily driven by the extremely low, and decreasing, role of women in the economy (LFP about 5 percentage point lower than the average for the Middle East and North Africa (MENA), which is already low by global standards, and female unemployment, at 25.2 percent, the highest in the region). Similarly, the country consistently performed unfavorably in governance and civil liberties–related indicators during the 2000s. Syria’s THE TOLL OF WAR v ratings on measures of freedom of association and assembly, and freedom of expression and belief, were low even compared with other countries in MENA. Another important feature of Syria’s governance landscape prior to 2011 was the country’s high levels of perceived corruption and low trust in public institutions. According to global indexes, Syria’s rankings on control of corruption and control of economic monopolies declined after 2005. Although the country had already trailed the middle-income MENA peers, like Egypt, Jordan, and Tunisia, on both fronts in 2005, the gap widened substantially by 2010. These trends also undermined citizen trust in public institutions. Gallup surveys from 2009–10 show that the percentage of Syrians expressing trust in key public institutions, such as local police and the judicial system, was lower than comparators. For instance, in 2010, only 48 percent of Syrians reported trust in local police; in comparison, about 87 percent of Jordanians responded favorably. Added to these enabling conditions were external factors that contributed to the onset of the conflict. The Arab Spring and a sudden shift in the regional context, where armed rebellions became easier, as well as the actions chosen by various parties, led to a quick escalation of the initial protests, from civil unrest to a long and intense armed conflict. Fueled by a “loser loses all” logic, which prevented a reasonable and compromising resolution, the conflict has persisted and intensified over time, leading to devastation for all. The Wrath of Conflict The conflict has inflicted extensive damage on Syria’s physical infrastructure. Cities like Homs, Aleppo, and Damascus, and many smaller towns, have served as battlegrounds for government and rebel offensives, with tragic consequences for their inhabitants. Over time, the conflict has caused the partial or full breakdown of urban systems in many cities by destroying houses and public service–related infrastructure like roads, schools, and hospitals, while leading to economic collapse in many areas. As bridges, water resources, grain silos, and other economically significant assets became strategic targets, the physical damage ratios increased. Across the 10 cities on which this study focuses, 27 percent of the housing stock has been impacted, with 7 percent destroyed and 20 percent partially damaged. The percentage varies across cities, with the highest full destruction occurring in Dayr az-Zawr (10 percent) and the highest partial damage in Tadmur (also called Palmyra, 32.8 percent). With 8 percent destroyed housing units and 23 percent partially damaged, Aleppo is also among the worst impacted cities. Across the eight governorates, about 8 percent of the housing stock has been destroyed and 23 percent partially damaged. The damage has been particularly high in the health sector, as medical facilities were specifically targeted. Estimates show that about half of all medical facilities in the eight governorates studied in this report have been partially damaged, and about 16 percent of them were destroyed. The results are similar in education sector facilities (53 percent partially damaged, and 10 percent destroyed). vi Executive Summary Disruptions in economic networks, human capital, and connectivity have greatly magnified the effects of physical damage on public service delivery. Physical damage reflects only a subset of the effects the conflict has imposed on public service delivery in Syria. The available evidence suggests that physical damage to the electricity infrastructure has been severe, but not devastating: all the country’s hydroelectric dams and six of 18 power plants remain operational, while four more power plants are partially damaged, and one has been destroyed. However, fuel shortages and conflict-driven constraints to operation and maintenance have led to a sharp drop in public power supply. Power generation declined to 16,208 gigawatt-hours (GWh) in 2015, compared with 43,164 GWh in 2010, a drop of 62.5 percent. Much of this decline appears to be due to fuel shortages, as available generation capacity declined by about 30 percent in the same period. The decline in electricity supply has caused major disruptions. The majority of cities receive only a few hours of electricity a day. The government has applied a rationing policy via load shedding throughout the country, which affects the delivery of other services, like water, education, and health care services. Among all the consequences of the conflict, the effects on human lives and demographic displacement have been the most dramatic. The pre-conflict population of Syria was estimated at 20.7 million in 2010 (World Development Indicators). Since 2011, the conflict has created a complex set of pressures on the country’s population. The most recent calculations by the International Organization for Migration (IOM) put the population within Syria at 18.8 million as of November 2016. Data limitations render a precise and comprehensive decomposition of the demographic changes impossible: conflict affects fertility rates and life expectancy alike. In addition, an important portion of demographic movements takes place informally: some refugees remain unregistered and, in certain cases, migrants do not factor into in-country population or refugee totals. The casualties that are directly related to conflict are estimated between 400,000 (UN, as of April 2016) and 470,000 (Syrian Center for Policy Research, as of February 2016). Syria has become the largest forced displacement crisis in the world since World War II. Over half of the country’s pre-conflict population has been forcibly displaced. According to the United Nations High Commissioner for Refugees, the total number of Syrians presently registered as refugees outside the country in Lebanon, Turkey, Jordan, Iraq, Egypt, and North Africa is 4.9 million. In addition, more than 800,000 Syrian nationals are estimated to have sought asylum in Europe in 2015 and 2016. Many of these individuals have moved more than once, and have not been removed from registration lists in their first country of refuge. These numbers also do not include an estimated 0.4 million to 1.1 million unregistered Syrian refugees in Lebanon, Jordan, Turkey and Iraq. The number of internally displaced persons was at 5.7 million as of January 2017, with 56 percent of them remaining within their own governorates. Although those who moved within their governorate may be more likely to return to their original communities, the return migration so far has been small (0.56 million) compared with the total numbers of displaced. THE TOLL OF WAR vii The losses in GDP between 2011 and 2016 sum to about four times the size of the Syrian GDP in 2010. The destruction of physical capital, casualties, forced displacement, and breakup of economic networks has had devastating consequences for Syrian economic activity. Syria’s GDP was estimated to have contracted by 61 percent between 2011 and 2015 in real terms, and by an additional 2 percent in 2016—a 63 percent decline compared with its 2010 GDP. Estimates of national account indicators, including counterfactual GDP numbers estimated by using statistical estimation methods, show that the actual GDP fell $51 billion (in 2010 prices) short of the counterfactual GDP in 2016. Aggregating these differences between counterfactual and actual GDP numbers between 2011 and 2016 shows that the cumulative loss in GDP amounts to $226 billion in 2010 prices, about four times the 2010 GDP. Economic disruption has been particularly devastating in the hydrocarbons sector. The oil GDP declined by 93 percent during the same period, while the non-oil economy contracted by 52 percent due to the severe destruction of infrastructure, reduced access to fuel and electricity, low business confidence, and disruption of trade. Hydrocarbon production plunged from 383,000 barrels per day (bpd) in 2010 to 10,000 bpd in 2015 and 2016, due to Islamic State control over most of the oil-producing areas.4 Agricultural production also registered significant losses as a result of damage to irrigation systems and shortages of labor and inputs such as seeds, fertilizers, and fuel. The most severe contraction of the economy took place in 2012 and 2013, when economic activity shrank by 29 and 32 percent, respectively, as fighting intensified and spread across the country. The Syrian economy suffers from severe twin deficits, depleted foreign exchange reserves, and an unsustainably high public debt. Conflict-related disruptions and international sanctions reduced Syrian exports by 92 percent between 2011 and 2015. The current account deficit was estimated to have reached 28 percent of GDP in 2016, up from 0.7 percent of GDP in 2010. The gap has increasingly been financed by withdrawing foreign exchange reserves, which declined severely, from nearly US$21 billion in 2010 to less than US$1 billion in 2015. Fiscal revenues dropped from 23 percent of GDP in 2010 to less than 3 percent of GDP in 2015. This was mainly due to losses in oil and tax revenues, the collapse of international trade due to sanctions, a growing informal economy, and weak administrative collection capacity. In response to this shortfall, government spending was cut back (especially capital expenditures), but these measures were not enough to offset the fall in revenues. As a result, gross public debt rose from 30 percent of GDP in 2010 to a staggering 150 percent of GDP in 2015. Rapidly shrinking job opportunities and scaled down social security programs have further aggravated a mounting humanitarian crisis. Since the 4 This refers to oil in areas under government control, but with the fields under rebel control, oil production is estimated at 40,000 bpd. viii Executive Summary onset of the conflict, jobs were destroyed at an estimated rate of approximately 538,000 per year on average between 2010 and 2015, adding 482,000 people to the unemployment pool every year. More than three in four Syrians of working age (7.7 percent, or nine million individuals) are not involved in any economic value generation: 2.9 million of them are unemployed and 6.1 million are inactive. Unemployment among youth reached 78 percent in 2015. Facing a mounting fiscal problem, the Syrian Government dramatically decreased subsidies. Prices of fuel oil increased 10-fold from 2011 to 2015. As for rice and sugar, prices increased 2.3-fold in the same period. Estimates for this report suggest that approximately six in 10 Syrians live in extreme poverty today. As of December 2016, 5.8 million individuals received in-kind food assistance. The World Food Programme alone distributes more than four million food baskets (supplements for about 1,700 kilocalories per day) each month. Disentangling the Economic Effects of the Conflict Which channel has been the most important in delivering the conflict’s impact on the Syrian economy? The discussions have so far focused on many of the conflict-driven factors that have affected the well-being of Syrian citizens. These include casualties, forced displacement, physical destruction, limitations on the mobility of goods and people, and a vast amount of foregone opportunities. These multidimensional effects, however, comprise overlapping components. Some of the most useful pieces of information, for example, the relative importance of physical destruction, migration, and so forth, in generating the economic impact of the conflict, are not directly observable. To address some of these concerns, this study adopted an integrative framework that helps disentangle the channels of transmission of the effects of the conflict. The analysis also introduced several policy-related scenarios that could inform future recovery and reconstruction efforts. Capital destruction, by itself, accounts only for a small share of the conflict’s economic impact. If the conflict only destroyed capital (at the same rate it has done so far), without other effects (capital-destruction-only scenario), its impacts on income and welfare would be relatively limited. By the sixth year of the conflict, the cumulative losses in GDP would only be 5 percent of the actual loss if the conflict destroyed only capital without leading to casualties and economic disorganization. This is mainly because without further economic disorganization, casualties, and migration, the effects of physical destruction on profitability remain relatively limited. Thus, investments remain relatively resilient, and the destroyed capital is rebuilt relatively quickly. Simulations show that investments decrease by 80 percent in the baseline, where all shocks associated with the conflict are imposed, and they decrease by only 22 percent in the capital-destruction-only scenario. This mechanism highlights a remarkable difference between natural disasters and conflicts. When a natural disaster destroys built capital in a well-functioning market economy with strong institutions, the recovery of capital stock is rapid, and THE TOLL OF WAR ix impacts are not persistent. Civil wars are different: casualties, demographic movements, and lower returns to investment compound the damage to physical capital. As a result, effective losses due to physical capital damage are more pronounced, and they can linger into the future at a much higher rate than natural disasters. Casualties impose insufferable distress and psychological impacts on families. Although their direct economic impact is comparable to that of the capital-destruction-only scenario, this impact is much more persistent. When conflict leads to casualties and no other effects (casualty-only scenario), change in GDP is comparable to that of the capital-destruction-only scenario: in both scenarios, the decrease in GDP during active conflict years remains lower than 5 percent of pre-conflict levels. Behind this similarity, however, lies a major difference between the two. Casualties are the primary drivers of outmigration: conflict-related casualties have led to a massive exodus of Syrians, whereas capital destruction has not. Simulations show that more than two-thirds of all outmigration can be attributed to casualty-related factors alone. Security-related factors dominate over economic reasons in explaining migration patterns: people move to have better life quality, but better life quality is largely about a lower probability of being killed in this case, not higher incomes. Another result from this analysis is that the economic impact of casualties is much more persistent than that of other factors. Even if the conflict ends in its sixth year, GDP continues to be significantly below its pre-conflict level: only a sixth of cumulative GDP losses within the first 20 years after the onset of conflict takes place during the six years of the conflict. The rest happens after the end of the conflict. Disruptions in economic organization have been the most important channel through which the conflict has manifested its economic impact. Where does the bulk of economic impact stem from if not capital destruction and casualties? It stems from disruptions in how economic activity is organized. Conflict not only destroys productive factors, it also prevents the connectivity of people, reduces their incentive to pursue productive activities, and breaks economic networks and supply chains. Our simulations show that, by the end of the sixth year of the conflict, the cumulative GDP losses in the economic-disorganization-only scenario exceed those of the capital-destruction-only and casualties-only scenarios by about 20-fold each. These results suggest that even reinstating lost capital, by itself, would not be sufficient to bring the economy back to its pre-conflict level, if institutional and organizational challenges are not tackled simultaneously. Regardless of the source of financing, boosting public investments without a comprehensive approach would trigger further rent seeking and cronyism, which were important enabling factors in the conflict’s onset, as described in the first section. Thus, recovery and reconstruction are by no means an engineering issue; our results show that the issue is primarily an economic and social one, where the incentives of Syrian citizens are at the core. x Executive Summary As the conflict continues, economic outcomes will further deteriorate and Syrians will continue to migrate. To provide an assessment of the possible recovery paths, the analysis employed different conflict end-date scenarios. These include a baseline, where the conflict ends in the sixth year, and two alternative scenarios, where the conflict ends in its 10th year or does not end in the near future. In all three cases, capital destruction, casualties, and economic disorganization shocks are employed. Accordingly, in the baseline, GDP recovers by about 20 percentage points (as a share of pre-conflict GDP) within four years after the conflict. In comparison, GDP will continue to deteriorate if the conflict does not end in the near future. In both alternative scenarios, the conflict will continue to be devastating for the country. When the conflict ends in its sixth year, the cumulative loss in GDP reaches 7.6 times the annual pre-conflict GDP by the 20th year. With a continued conflict, this loss stands at 13.2. Simulations also show that outmigration doubles between the sixth year of the conflict and the 20th year. Thus, the supply of migrants and forcibly displaced persons will not cease as long as there is conflict. The rate of deterioration slows, because the conflict is geographically limited, migration drains the labor force in conflict-affected regions, and capital stock reaches a stable level with new damage and investment ratios. As a result, the economy converges to a new equilibrium. The longer the conflict continues, the slower the post-conflict recovery will be. Another important result from the conflict end-date scenarios is that the duration of the conflict also affects the pace of recovery. Although the rate of deterioration moderates over the course of the conflict, the effects become more persistent. When conflict ends in its sixth year (baseline), GDP recoups about 41 percent of the remaining gap with its pre-conflict level within the next four years. In comparison, it recoups only 28 percent of the gap in four years if the conflict ends in its 10th year (alternative scenario). These results do not capture much of the complications like political economy challenges, such as conflict-driven grievances. Adding these factors would only reinforce the main finding here: the more the conflict persists, the deeper will run the grievances and divisions in Syrian society, rendering it very difficult to reinstate efficient institutions and economic mechanisms. 1 The ablution fountains of the Great Mosque of Aleppo or the Umayyad Mosque of Aleppo, Ancient City of Aleppo, Syria. Photo Credit: Yann Arthus-Bertrand, Getty Images. CHAPTER SYRIA BEFORE THE CONFLICT When the Arab Spring protests took place in the Syrian Arab Republic in early 2011, few would have anticipated the forthcoming agony on a mass scale. Although it is not difficult to find “evidence” for the approaching conflict in retrospect, standard development indicators, including inequality and poverty measures, failed to predict what has become one of the most brutal conflicts of our time. Syria was following the regional trends in economic progress: extreme poverty was low; hunger and child and maternal mortality were decreasing; and school enrollment was increasing. However, there were factors that distinguished Syria from others. These included an unfavorable profile of employment and labor force participation, especially those of women; low and deteriorating indicators of governance, including lack of political pluralism and civil liberties; and high levels of perceived corruption. These characteristics have emerged historically from a complex interplay between geography and institutions, which was also manifested in a rapid centripetal momentum that accompanied the independence of the Syrian Arab Republic. 2 Syria Before the Conflict C H A P T E R 1 : SYRIA BEFORE THE CONFLICT The Arab Spring protests marked the beginning of a new era in the Arab Republic of Syria. Minor public protests by the Syrian opposition began almost immediately after the initial protests in Cairo in January 2011. The first large demonstrations began two months later in March, and the following months saw a process of escalation as demonstrations spread and increased in size within the country. By the summer of 2011, the armed conflict was already unfolding. Now in its sixth year, the Syrian conflict remains active and is bringing much pain and tragedy on a daily basis. Notwithstanding present perceptions, few would have anticipated the forthcoming cataclysm in Syria back in 2010. Syria is now deeply associated with conflict, refugees, and agony on mass scale. In retrospect, it is not difficult to find “evidence” for the approaching conflict in the previous decade. Yet, a conflict at this scale was not anticipated before its eruption. Ianchovichina, Mottaghi, and Devarajan (2015) suggest that the standard development indicators, including growth, inequality, and poverty measures, failed to predict the Arab Spring, including in Syria. The region had been making progress in eliminating extreme poverty and boosting shared prosperity; the poverty rate was declining in all economies (except the Republic of Yemen); hunger and child and maternal mortality were decreasing and school enrollment was increasing. An eyewitness to the developments in Syria reported: “…2008-2010 were years of uneasy calm, a period of superficial normality on the brink of revolution. I certainly didn’t sense that a civil war would erupt in Syria in March 2011, a few months after my last visit, and I’d venture to say most Syrians did not either” (Sahner, 2014). This chapter describes the pre-conflict conditions in Syria to understand how a seemingly stable, middle-income country like Syria experienced a brutal conflict and endured devastating impact. Although the Arab Spring marked the beginning of the events that evolved into a fully-fledged conflict, it is not enough to conclude without further analysis that “the Arab Spring caused the Syrian conflict.” One reason is that the same external shock hit all Arab countries, while full-on conflicts developed only in Syria, Libya, and the Republic of Yemen.5 The other republican governments in the region almost all faced protests and riots, but did not have conflicts develop. Similarly, the monarchies of the region experienced even less in the way of protests and no conflicts. Therefore, even if the Arab Spring triggered the conflict, there must be domestic factors that made this possible. These factors are also key to understanding the manifestation of the economic impact of the conflict. 5 In the Republic of Yemen, the Arab Spring did not start a civil war, but played into what was already a set of ongoing civil wars. THE TOLL OF WAR 3 The Long History of Syria The geography of Syria has played an important role in the formation of its institutions.6 The Syrian landscape comprises immense deserts with stretches of arable land along the coastline as well as oases of major cities (map 1.1). From deserts to arable strips of land, population densities vary significantly. This landscape and corresponding demographic adjustment have worked against the establishment of a powerful centralized administration through two channels. Map 1.1: Land Utilization in Syria Cotton Fruits Olives Wheat Tobacco Cultivated land with livestock, emphasis on grains, cotton, fruits and olives Forest area Steppe land with nomadic herding (sheep) and scattered cultivation. Desert and steppe lands with some nomadic herding Source: Cole (2010) Administration of this diverse geography required adaptive solutions. First, in pre-modern times, the necessity of major irrigation schemes fostered the development of centralized states. This “Oriental Despotism” or “Asiatic mode of production” was prominent in the Arab Republic of Egypt, and to a lesser extent in Mesopotamia. Syria formed an outlier among these cases. With limited prospects and returns to large irrigation projects, the capacity of centralized powers to provide public goods and services was diminished and, unlike other places in Mesopotamia, private property and use-right-based exploitation of land prevailed longer (Issawi 1995). The exercise of private property rights was further reinforced by the 19th century, as communal 6 The role played by geography and institutions, and their interaction, in shaping political economy outcomes has become an increasingly popular area of discussion in the economic development literature. For an interesting discussion on this, see Acemoglu and Robinson (2013), Gallup, Sachs, and Mellinger (1999), and Nunn and Puga (2012). 4 Syria Before the Conflict tenure (musha) became increasingly marginalized. Second, vast, thinly populated desert territories rendered it difficult to maintain law and order. Central authorities have often relied on a coalition of nomad and semi-nomad populations (for example, Bedouins, but also Druses, Kurds, Circassians, and others) in providing security. Historically, there was always a segment of the population potentially ready to provide such armed “protection services” as an alternative to engaging in strictly economic activities for making a living (Douwes 2000). Alternating centripetal and centrifugal cycles are commonly observed in Syrian history (box 1.1). Occasionally the central authority lost power and the local potentates took up arms overnight and posed themselves as local “mini states,” as has been the case recently with the so-called Islamic State that ruled over a vast yet thinly populated territory comprising much of Syria’s uncultivable land.7 At other times, the central authorities have time and again reasserted their rule, the last instance of which has been witnessed during the conflict. Some of the zones of influence of the various fighting forces during the conflict, especially the areas controlled by the Islamic State (also called ISIS or ISIL) and the Syrian army, are traceable to the ancient geographical and demographic fault lines. Box 1.1: Key Historical Dates 1516–1918 Syrian territories ruled as provinces of the Ottoman Empire 1920–1946 Syria under French Mandate 1946 Syria gains independence from France 1963 Baath Party comes to power, economic development policy characterized by state dominance and inward-looking import substitution model 1970 Hafez Assad becomes president of Syria, launches Corrective Movement that gives more space to the private sector while keeping the state’s dominant role in the economy 2000 Bashar Assad becomes president of Syria 2005 Tenth Five-Year Plan signals break with Syria’s state-led economic development model, move toward “Social Market Economy” 2011 January. Regional “Arab Spring” uprisings 2011 January. First instance of Arab Spring-inspired unrest, in Damascus souq 2011 March Unrest begins in Daraa 2011 Spring Unrest spreads, driven primarily by rural and peri-rural mobilization Mid-2011- … Escalation of conflict 7 Among the most important historical examples along these lines are the following. In 1857 when the Maronite peasants took up arms against their landlords in the Mount Lebanon region, the popular hero, Tanyus Shahin, acted in the name of “republican government” or “peasant commonwealth”; see Khalaf (1987). During the early period of the French Mandate, two regions became hotbeds of resistance to pacification in the countryside. One was the Alawite Mountain and the other, the northwestern districts of the Aleppo state; see Khoury (1987). Alawis revolted during the French Mandate in 1919–21, seeking their own State of the Alawis and demanding they not be placed under the State of Syria. In Jebel Druze, the Druzes revolted in 1925 to preserve their practical autonomy and resisted the French until 1927; see Hourani (1946). THE TOLL OF WAR 5 The peculiar characteristics of the Syrian landscape also augmented the role of external linkages. Before its independence, the geographical Syria historically often resided as a province of large empires. This meant that it received international “protection”—provided at little or no cost—by larger powers, most notably the Ottoman Empire (map 1.2). This special relationship involved “political and economic exchanges” between the empires and geographical Syria. For instance, the Ottoman Empire provided protection at home as much as it could, and protection from foreign interference, from the early 16th century to the early 20th century, except for a brief intervention in 1860 when French forces intervened in response to the civil war. Geographical Syria also benefitted economically from being integrated into imperial-scale economic networks, which paid off handsomely, albeit unevenly, as will be discussed below. Map 1.2: The Ottoman Administration of Geographical Syria Source: Sullivan (2013). Syria’s access to long-distance trade networks under the protection of large empires shaped its economic and political topography. Geographical Syria comprised two major centers of economic activity: Aleppo and Damascus. During Ottoman Rule, the Eyalet (province) of Aleppo and the Eyalet of Damascus played vital roles in connecting trade routes, the organization and provisioning of periodic caravans, and the regular extension of credit, as well as protection to the merchants and their merchandise. Aleppo was historically integrated more into the Anatolian networks of trade, while Damascus looked in the opposite direction, south. Aleppo was the third largest city (after Istanbul and Cairo) of the Ottoman Empire in the 17th and 18th centuries, as well as “the leading commercial and handicraft center of southwest Asia” (Issawi 1988, 4). Thanks to its easy access to the Mediterranean via Alexandretta, it was the major entrepôt for European goods just as it exported 6 Syria Before the Conflict its famous manufactures (map 1.3). In the early 19th century, each year four caravans left Aleppo for Iraq and Anatolia and two for Persia and Iraq. On average, a caravan consisted of about 1,500 camels. It took 25–30 days to cross the desert directly between Aleppo and Bagdad, and 45 days via Diyarbakır (Owen 1981, 53–54). Damascus was also well connected in general and, to its benefit, it served as the historic point of departure for the annual Hajj caravan. Two or three caravans a year traversed the desert between Damascus and Bagdad in the late 18th century. Yet, Damascus lacked easy, quick, and safe access to the coastline, which only came with the construction of the Beirut-Damascus road (1863) and ultimately the railway (1895). Before then, the dangerous trip of only 70–80 miles from Damascus to Saida or Beirut cost more than crossing the entire Syrian Desert as late as 1848 (Owen 1981, 54). This made Aleppo the preferred location for foreign merchants who could maintain relatively easy ties with Europe via Alexandretta. Map 1.3: Caravan Routes, 16th to 20th Centuries Source: Owen 1981 THE TOLL OF WAR 7 The establishment of nation states in the region shook the foundations of the traditional Syrian economy. Broken trade networks and newly introduced trade frictions at the borders posed significant costs on Syria. After the secession of Beirut and Alexandretta from the historical greater Syria, Aleppo and Damascus lost their strategic and commercial access to the Mediterranean, which has only been partly compensated by the costly development of the lesser Port of Latakia. The merchants of Aleppo, who were dependent on long-distance trade, were cut off from their historical connections. So were the merchants of Damascus; however, their losses could be compensated by expanding the service sector in the new nation’s capital. The truncation of Syria from its economic networks boosted the centralization momentum that came with independence. After independence, the Syrian government launched a path of agriculture-based, increasingly import-substituting industrial and economic development strategy. This required strong “vertical” connections between the foundational agriculture and to-be-developed industrial and service sectors. However, Syria’s land was much exhausted by hundreds of years of hard use, climatic change, and scarce water resources. Hence developing agriculture was a challenge in which Syria had scant comparative advantage. Even so, the Baath government pushed agriculture to the limits of the possible. By doing so, the central government instituted itself in the allocation of resources, in particular via land redistribution and access to irrigation. The reinforcement of a centralized state came with complications. The new government in Damascus faced complex challenges in promoting a cohesive nation state. The policies of empires tolerated, if not emphasized, sectarian and ethnic differences. For instance, under the French administration, minorities were afforded autonomy and rights vis-a-vis neighboring ethnicities, and regional and communal representation was established in the Parliament. Centralization meant the complete absorption of certain groups, like the Alawites, into the young state apparatus and the exclusion of others, like the Kurds. Alawites became reconciled to common Syrian citizenship, and gave up the dream of a separate Alawite state. Kurds faced a harsh transition. With the 1962 census, nearly 120,000 Kurds, 20 percent of the resident Kurdish population, found themselves “stateless” overnight in the Jazira province (HRW 1996). Other harsh measures included the implementation of an “Arab Cordon” on the lands that bordered Turkey. Kurds were expelled, and Arab Bedouins were brought in to resettle the area. As the cordon ended, some 140,000 Kurds had been relocated to the southern desert of Al-Raad (Vanly 1992). By the new millennium, Syria was ruled by a well-established, authoritarian, and seemingly stable central administration. Since 1970, the last successful coup, which brought Hafez al-Assad to power, presidential elections have never been contested and the single candidate (Hafez al-Assad, then his son Bashar) has never won less than 99 percent of the vote. The executive and 8 Syria Before the Conflict Baath Party have dominated all three branches of government (parliaments have been handpicked by the Baath Party and the president). In the economic and social spheres, the government was considered as inclusive with a broad popular base (Batatu 1999). The government distributed relatively generous rewards, especially to those who took the extra step to join the ranks of the Baath Party. Syria at the Dawn of the Conflict The decade before the conflict witnessed a transition from four decades of socialism toward a “social market-economy.” In 2001, Syria requested to become a member of the World Trade Organization and, in 2007, a free trade agreement with Turkey came into effect. In 2004, the government permitted private banks to operate in Syria for the first time after more than 40 years, and in 2009, it opened the stock market. This trend accelerated from 2005, when Syria embraced “Social Market Development” in its Tenth Five-Year Plan. In the period that led up to 2011, Syria undertook several market-friendly reforms. The Syrian government focused in particular on improvements to the business environment and took measures to reduce tariffs, lower the top income tax rate, and unify the exchange rate. The decline in oil production in many ways served as the motivation for these reforms. On the eve of the 2011 unrest, Syria was a fast-growing, lower-middle-income country. In aggregate terms, the Syrian economy was improving, albeit starting from an unfavorable base, during the 2000s. Gross domestic product (GDP) grew at an average of 4.3 percent per year from 2000 to 2010 in real terms, which was almost entirely driven by growth in the non-oil sectors, and inflation averaged a reasonable 4.9 percent. The current account was largely in balance and, by the end of 2010, Syria was estimated to have international reserves to cover more than nine months of imports. Despite the improvements, however, Syria’s GDP per capita in 2010 (at current prices, US$2,949) was significantly lower than those of its neighbors (Iraq: US$4,521, Lebanon: US$8,846, and Israel US$31,665). The opening of Syria was also reflected in its booming international trade. The value of Syrian trade with the rest of the world reached 76.5 percent of GDP just before the global financial crisis, which was relatively high compared with the average for the Middle East and North Africa (MENA) (excluding oil-exporting high-income countries) of 70.3 percent of GDP. Tariffs in Syria were slightly high by regional standards, with a weighted mean rate of 7.2 percent in 2010, compared with 6.1 percent for MENA (excluding high-income countries). However, there was relatively little foreign direct investment through this period. Foreign investment represented 1.3 percent of GDP on average during this period and focused mainly on pharmaceuticals, food processing, and textiles. Therefore, the impact of the global financial crisis on Syrian GDP was relatively moderate, and it was primarily transmitted by changes in the major export markets of Syrian producers, mainly the European Union and Gulf economies. THE TOLL OF WAR 9 Despite the increasing momentum, however, the Syrian economy exhibited several risk factors that may have contributed to the onset and escalation of the conflict. The conventional risk factors that are associated with armed conflict include absolute (objective) and relative (subjective) levels of socioeconomic deprivation, ethnic/sectarian fragmentation, and unfavorable governance indicators, including corruption and low institutional capacity.8 In what follows, the analysis will describe a several areas where improvements in aggregate economic conditions failed to mitigate these potential risk factors in Syria. Employment Growth and market-oriented reforms created more jobs; however, those fell short of keeping up with the demographic dynamics. The high economic growth rate and increasing liberalization efforts led to an expansion of employment at a rate of 1.3 percent annually, on average, in the decade prior to the conflict. The growth of the population (2.9 percent annually) and labor force (1.7 percent annually) dwarfed the growth in employment. Agriculture experienced a contraction in employment in the same period. The sector’s share of employment fell from 32.9 percent in 2000 to just 13.2 percent by 2011, in part because of the droughts that struck the country from 2006 to 2011, which contributed to a major demographic shift from rural areas to urban centers, increasing the shares of employment in services and industry to 55.3 and 31.4 percent, respectively. It has been argued that the 2006–09 drought and concomitant migration to cities was a major risk factor that contributed to the events in 2011; however, it is difficult to verify this empirically. The excess labor, especially women, were pushed out of the labor force. With labor markets failing to absorb the growing size of the labor force, from 2001 to 2010, the aggregate labor force participation rate decreased by about 7 percentage points (from 50.2 to 43.5 percent). The adjustment was more severe for women: their labor force participation (LFP) rate decreased by more than 10 percentage points (from 20.4 to 13.2 percent). By 2010, Syria had one of the lowest LFPs in the world, low even by regional standards (Egypt and Tunisia had 49 and 47 percent LPF rates, respectively). The gap was primarily driven by the extremely low, and decreasing, participation of women in the economy. Female LFP in Syria was about 5 percentage points lower than the MENA average (excluding high-income countries) and 10 percentage points lower than Egypt and Tunisia. Similarly, the unemployment rate among women (25.2 percent), more than four times higher than that of men (6.1 percent), was the highest in MENA (figures 1.1 and 1.2). 8 See Fearon and Laitin (2003) for a detailed discussion on this. 10 Syria Before the Conflict Figure 1.1: Labor Force Participation Rate, 2010 TUN 24% EGY 76% EGY 23% TUN 70% JOR 15% SYR 69% SYR 14% JOR 68% MNA 19% MNA 72% LMC 38% LMC 79% 0% 10% 20% 30% 40% 50% 60% 65% 70% 75% 80% 85% Source: World Development Indicators; Syria Central Bureau of Statistics. Note: The MNA average excludes high-income countries. EGY = Egypt; JOR = Jordan; LMC = lower-middle-income countries; MNA = Middle East and North Africa; SYR = Syrian Arab Republic; TUN = Tunisia. Figure 1.2: Structure of the Working-Age Population in Syria, 2010 Working age population 13million 63% Inactive Active 7.2million 55.4% 5.8million 44.6% Employed Unemployed 5.3million 91.4% 0.5million 8.6% Men Women 0.3million 61% 0.2million 39% Wage-employed Self-employed Employer Unpaid worker 3.3million 63.2% 1.5million 28.8% 0.2million 4.2% 0.2million 3.8% Private sector Public sector 1.9million 57% 1.4million 43% Source: Central Bureau of Statistics Labor Force Survey 2010. THE TOLL OF WAR 11 Poverty and Inequality During the 2000s, poverty rates registered small increases, but overall remained within regional norms. Table 1.1 summarizes poverty measurements and Gini index over the three periods of surveys covered in UNDP (2005) and UNDP (2011). In 1996–97, the national extreme poverty rate was 14.3 percent and the overall poverty rate was 33.2 percent, with a Gini coefficient of 33.7. In 2003–04, the extreme poverty and overall poverty rates had declined to 11.4 and 30.1 percent, respectively, while the Gini index increased to 37.4. Thus, growth was not pro-poor, as the expenditure distribution became more unequal, but it was more than offset by increased per capita expenditure. In 2006–07, the poverty rates rose slightly (12.3 percent extreme poverty and 33.6 percent overall poverty), reversing the gains between 1997 and 2004, whereas inequality fell to a Gini index of 32.6. According to UNDP (2014), multidimensional poverty in Syria (5.5%) stood above those in Tunisia (1.2 percent) and Jordan (2.4 percent), but below that of Egypt (6.0 percent) in 2010. The inequality measure in Syria was well within the range observed in other countries during the 2000s; the Gini index varied from 41 percent in Tunisia in 2000 to 31 percent in Egypt in 2008. Table 1.1: Incidence of the Abject, Extreme, and Overall Poverty Rates and the Gini Coefficient, 1997–2007 Abject poverty Extreme poverty Overall poverty Gini coefficient rate (%) rate (%) rate (%) 1997 14.3 33.2 33.7 2004 2.2 11.4 30.1 37.4 2007 1.2 12.3 33.6 32.6 Source: UNDP 2011. Note: The estimates reflect UNDP’s cost-of-basic-needs methodology-based poverty assess- ments. Abject poverty defines the share of population whose expenditure lies below the food poverty line. Extreme poverty defines the share of population whose per capita expenditure is less than the cost of food plus expenditure on absolute-minimum essential nonfood goods. Overall poverty defines the share of population whose per capita expenditure is less than the cost of food and a reasonable minimum expenditure on nonfood items. UNDP = United Nations Development Programme. Poverty in Syria has been concentrated in the northeast region, especially in rural areas. Rural areas in Syria have consistently been poorer than urban centers, and the urban-rural divide has grown since 1997 (table 1.2). The poorest part of the country in 2007 was the rural northeast, with a headcount ratio of 19.7 percent, seven percentage points higher than the next area (12.8 percent in rural areas of the southern region). While 44 percent of the total population lived in the northeast, it accounted for 56 percent of the extreme poor. The rural northeast was also the only area that experienced an increase in 12 Syria Before the Conflict poverty between 1997 and 2004. The least poor area in 2007 was the coastal region, with an extreme poverty rate of 7.7 percent. The southern and central regions enjoyed a huge drop in the poverty rate between 1997 and 2004, which drove poverty reduction for overall Syria in 2003–04. Comparing the urban-rural discrepancies in the headcount ratio in Syria with those in comparator countries, the differences were not particularly pronounced in Syria. In 2007, the difference in urban versus rural poverty rate was 6 percentage points. This was higher than the spatial inequity in Jordan, but far more moderate than the situation in Egypt in 2010, where the headcount ratio in rural areas was 17 percentage points higher than in urban areas (32 versus 15 percent). Table 1.2: Incidence of Extreme Poverty, by Region, 1997, 2004, and 2007 1997 2004 2007 Urban 10.69 5.82 10.6 Southern region Rural 15.19 10.67 12.8 Total 11.2 Urban 13.94 11.16 10.8 North-East region Rural 15.24 17.91 19.7 Total 15.4 Urban 14.80 9.02 7.76 Central region Rural 22.58 11.10 9.05 Total 8.4 Urban 11.34 9.34 5.6 Coastal region Rural 9.52 9.70 9.09 Total 7.7 Urban 12.6 8.7 9.9 Syria Rural 15.9 14.2 15.1 Total 14.3 11.4 12.3 Source: UNDP 2011. Note: For ease of exposition, the table focuses on the extreme poverty rate for the discussion of regional differences. The interpretations are qualitatively the same for the overall poverty rate. Regions comprise the following governorates: South – Damascus, Rural Damascus, Deraa, El Suaida, and El Quneitra; Northeast – Idleb, Aleppo, Al Raqqa, Deir Ezzor, and Hassakeh; Central – Homs and Hama; Coastal – Tartous and Latakia. THE TOLL OF WAR 13 Several external factors reinforced the deterioration of poverty in the south- ern and northeast regions. Syria and the greater Fertile Crescent region experi- enced moderate-to-severe drought from 1998 to 2009, with the winter of 2007/08 being the driest since records were kept starting in 1931 (Kelley et al. 2015). This caused large-scale failure of crops and livestock deaths, displacing as many as 1.5 million people from rural farming communities to urban areas in the south, in search of alternative jobs and livelihoods. This displacement coincided with the arrival of 1.2 million to 1.5 million Iraqi refugees between 2003 and 2007, who settled primarily in the urban regions in the south. The large inflow of internal migrants and Iraqi refugees could depress wages in the urban south, especially in the informal sector, where many of the displaced would be employed or seek work, and drought-driven increases in food prices could further depress the real incomes of the poor (UNDP 2011). Governance Unlike the economic dimensions discussed so far, Syria consistently ranked low in ratings on basic indicators of governance, including institutional checks and balances and political pluralism (figure 1.3). In governance indicators, the country hovered at the bottom of the scale or saw a visible deterioration during the 2000s. Similarly, the country scored substantially below Organisation for Economic Co-operation and Development (OECD) and middle-income MENA comparators, including Egypt, Jordan, and Tunisia.9 These rankings reflect the closed nature of Syria’s governance system and considerable concentration of power in the executive, even by regional standards. The country also performed poorly on administrative and policy implementation functions in absolute terms and relative to other middle-income MENA states and the OECD, although after 2008 its absolute rating showed signs of modest improvement. Studies from that period reveal that the quality of regulatory functions and services to citizens and businesses were low and stagnant (International Crisis Group 2009). The government’s Mid-Term Evaluation Report of the Tenth Five-Year Plan documented poor implementation of planned institutional reforms and insufficient progress on administrative reform measures (Syrian Center for Policy Research 2013). The lack of civil liberties and freedom was more prominent in Syria than in the other Arab Spring countries, based on low ratings on several governance and institutional indicators. Syria’s ratings on measures of freedom of association and assembly, and freedom of expression and belief, are low not only relative to young OECD members (for example, Estonia), but also relative to other MENA comparators. Relevant global indicators suggest that, during the 2000s, Syria’s coercive capacity was in line with that of other MENA 9 The Bertelsmann Transformation Index data (note shown) suggest that Syria also significantly underperformed on freedom of association relative to comparators. This is supported by many scholarly accounts from the pre-2011 period, documenting Syria’s extremely constrained spaces for independent civil society (Hinnebusch 1995; Baroutt 2011). 14 Syria Before the Conflict comparators, which is particularly noteworthy, given Syria’s geographic proximity to countries that were experiencing or recently emerging from violent conflict (for example, West Bank and Gaza, Lebanon, and Iraq). Another important feature of Syria’s governance landscape prior to 2011 was the country’s high levels of perceived corruption. According to global indexes, Syria’s rankings on control of corruption and control of economic monopolies sharply declined after 2005. Although the country had already trailed MENA middle-income peers on both fronts in 2005, the gap widened substantially by 2010. These trends also undermined citizen trust in public institutions. Gallup surveys from 2009–10 show that the percentage of Syrians expressing trust in key public institutions, such as local police and the judicial system, was lower than comparators. For instance, in 2010, only 48 percent of Syrians reported trust in local police; in comparison, about 87 percent of Jordanians responded favorably.10 Overall, the inability to match the aggregate economic developments with further economic and political inclusion, transparency, and civil liberties constituted major risk factors. The analysis so far has shown that the strong growth performance in the decade before the conflict was not accompanied by an equally strong performance in several areas. Although Syria did not compare unfavorably with other regional economies in many areas, including aggregate inequality and rural-urban disparity of poverty, in certain areas it did. These included deterioration in labor market conditions, especially for women. These various factors, in turn, constituted enabling conditions for the onset of the conflict before the Arab Spring events kindled the fire.11 10 In Syria, the Gallup survey did not field questions on trust in the executive or government more generally. Data limitations inhibited a comprehensive comparison of Syria with MENA peers other than Tunisia. 11 The main causes of the onset of the conflict in 2011 should be separated from the main drivers of its duration and intensity. Whereas the former is largely determined by domestic factors and the Arab Spring effect, the latter are purely determined by the characteristics of the Syrian political economy. For an analysis on this, see Fearon (2017); for a broader assessment of the socioeconomic drivers of the Arab Spring, see Ghanem (2016). THE TOLL OF WAR 15 Figure 1.3: Governance Indicators in Syria SEPARATION OF POWERS POLITICAL PLURALISM & PARTICIPATION 12 16 Estonia Separtation of powers (1-10 max) 14 Political pluralism & participation 10 Estonia 12 8 10 6 8 Egypt 6 4 Jordan Jordan Tunisia 4 Egypt Syria Tunisia 2 2 0 0 Syria 2006 2008 2010 2006 2007 2008 2009 2010 Source: Bertelsmann Transformation Index Source: Freedom in the World, Freedom House FREEDOM OF ASSOCIATION & ASSEMBLY FREEDOM OF EXPRESSION & BELIEF 12 16 Estonia Freedom of association/assembly Estonia 14 10 Freedom of expression & belief 12 8 10 6 8 Jordan Egypt Egypt Jordan 6 4 Tunisia 4 Tunisia 2 Syria Syria 2 0 0 2006 2008 2010 2006 2007 2008 2009 2010 Source: Bertelsmann Transformation Index Source: Freedom in the World, Freedom House CONTROL OF CORRUPTION (PERCENTILE RANK) CONTROL OF MONOPOLIES 1.5 12 Control of monopolies (0-10 best) Control of corruption (-1.5 to 1.5 best) 1.0 Estonia 10 Estonia 0.5 8 0.0 Jordan Jordan 6 Egypt Tunisia Tunisia -0.5 Egypt 4 -1.0 Syria 2 Syria -1.5 0 2005 2006 2007 2008 2009 2010 2006 2008 2010 Source: World Governance Indicators Source: Bertelsmann Transformation Index 2 Musa, a 25-year-old Kurdish marksman, stands atop a building as he looks at the destroyed Syrian town of Kobane, also known as Ain al-Arab, on January 30, 2015. Kurdish forces recaptured the town on the Turkish frontier on January 26 in a symbolic blow to the jihadists who have seized large swathes of territory in their onslaught across Syria and Iraq. Photo Credit: Bulent Kilic/AFP/Getty Images. CHAPTER THE WRATH OF CONFLICT The conflict has inflicted significant costs in terms of physical damage, loss of lives, and demographic dispersion, as well as worsening economic and human development outcomes in Syria. Across the 10 cities studied in this report, 27 percent of the housing stock has been destroyed or partially damaged. The effects on human lives and displacement have been even more dramatic. Conflict-related deaths are estimated between 312,000 and 470,000. Many more are estimated to have died as a result of the indirect effects of the conflict, including reduced access to health care, medicine, and nutrition, and over half of the country’s pre-conflict population has now been forcibly displaced. Overall, Syria’s 2016 GDP was estimated at only about a third of its 2010 value, and today two of every three Syrians live in extreme poverty. As the conflict remains active, these outcomes deteriorate further on a daily basis. 18 The Wrath of Conflict C H A P T E R 2 : THE WRATH OF CONFLICT Once triggered, the conflict quickly escalated to affect Syrians’ lives in extensive and devastating ways. Conflicts destroy tangible and intangible assets, and leave deep marks on a country’s culture and the minds of its people. Syria exemplifies this. What began as peaceful protests in 2011 escalated into a full-scale armed conflict among many parties. The conflict has drawn in other countries in a direct or indirect manner, and it has led to spillovers in neighboring countries and destinations in Europe. The intensity and impact of the conflict have not been the same everywhere in Syria. Although the ongoing conflict has affected the entire Syrian landscape and demography one way or another, it has taken different forms of varying magnitude across the country’s vastly heterogeneous geography. Map 2.1 shows the variation in exposure to conflict in different areas. A large portion of Syria has experienced extensive low-intensity conflict (defined as more than 70 percent of populated areas within the district experiencing low-intensity conflict). The particularly destructive forms of conflict, artillery shelling and airstrikes, have not been equally distributed. Raqqa, for instance, has been mostly exposed to extensive airstrikes (defined as more than 2,000 strikes from 2011 to 2017), but it has not experienced heavy artillery shelling or extensive low-intensity conflict. Aleppo (city and district) has experienced the highest level of conflict incidence in all categories. This chapter takes stock, in a nonexhaustive way, of outcomes associated with the conflict. It was not possible for this study to capture all the consequences of the conflict in a comprehensive manner. Some of those, including the political, social, security-related, and institutional effects, are not quantifiable. The bulk of these effects will be manifested in the future, or extend into the future. In the absence of better tools to capture those, the analysis in this chapter focuses on effects that are observed quantitatively, and supports them with qualitative observations. More specifically, we discuss the effects of the conflict by using the following classification: (i) physical damage, (ii) loss of lives and demographic dispersion, and (iii) economic and human development outcomes. The findings of this chapter will then be used in chapter 3, where the analysis will turn to assessing the overall economic impacts of the conflict. Physical Damage The conflict has inflicted extensive damages on Syria’s physical infrastructure. Cities like Homs, Aleppo, and Damascus, and many smaller towns, have served as battlegrounds for government and rebel offensives, with tragic humanitarian consequences for their inhabitants. Over time, the conflict has caused the breakdown of partial or entire urban systems in many cities, by destroying houses and public service–related infrastructure, like roads, schools, and hospitals, leading to the collapse of economic activities in many areas. As bridges, water resources, grain silos, and other economically significant assets became strategic targets, the bill due to physical damage increased. THE TOLL OF WAR 19 Map 2.1: Incidence of Conflict, 2011–17 Aleppo Raqqa Idlib Aleppo LOW INTENSITY CONFLICT Raqqa Idlib By District (ADM2) Relative Prevalence Hama Dayr Az-Zawr Extensive Homs Dayr Az-Zawr Moderate Hama Rif Dimashq Limited to none Dar’a Homs Rif Dimashq Dar’a Aleppo Raqqa Idlib HEAVY ARTILLERY SHELLING IN Aleppo Idlib Raqqa EACH GOVERNORATE Hama Dayr Az-Zawr By District (ADM2) Relative Intensity Homs Dayr Az-Zawr Extensive Hama Moderate Rif Dimashq Dar’a Limited to none Homs Rif Dimashq Dar’a 20 The Wrath of Conflict Aleppo Raqqa Idlib Aleppo AIRSTRIKES IN EACH GOVERNORATE Idlib Raqqa By District (ADM2) Relative Intensity Dayr Az-Zawr Hama Extensive (Recent) Dayr Az-Zawr Homs Extensive Hama Moderate Rif Dimashq Dar’a Limited to none Homs Rif Dimashq Dar’a Source: World Bank staff calculations. This section assesses the degree of physical damage by using remote-sensing techniques in conjunction with other available information. In the absence of comprehensive access to on-the-ground information, the analysis relies on remote-sensing-based calculations, which are then cross verified by using various public information channels, including local news and other media outlets and social media analytics (for still or motion pictures of destruction). The analysis relies on three components: • City-level assessments. These provide damage estimates as of January 2017 for 10 cities, including Aleppo, Dayr az-Zawr, Raqqa, Dar’a, Idlib, Douma, Kobane, Tadmur (Palmyra), Homs, and Hama. The specific sectors of focus for each city are the following: housing, water and sanitation (WASH-public network), energy (public network), health (public facilities), education (public and facilities), and transportation (primary road network). • Governorate-level assessments. These extrapolate the findings to provide damage analysis for the country’s eight most conflict-affected governorates: Aleppo, Homs, Hama, Dayr az-Zawr, Raqqa, Dar’a, Idlib, and Rif Dimashq. The extrapolation relies on city-level damage; however, it utilizes additional information like intensity and composition of conflict in cities and corresponding governorates, as well as the baseline differences between initial stocks of infrastructure. • Targeted assessments. These provide assessments in a narrow category that may not be comprehensive as a result of data limitations, for example, damage in industrial zones, steel and cement factories, and a pilot assessment of environmental damage in the cities of Aleppo and Hama, with a special focus on conflict-driven debris. THE TOLL OF WAR 21 Housing Syria is a highly urbanized country. Prior to the conflict, Syria was rapidly urbanizing, with 56 percent of the population living in urban areas, most of which are in rain-fed agricultural regions, including the basin of the Euphrates River, or along interior trade routes. The two largest cities, Damascus and Aleppo, accounted for nearly 37 percent of the urban population and 20 percent of the total population. According to the 2010 Syrian census data, there were 4,128,941 conventional dwelling units across Syria’s 14 governorates. Residential housing stock in the eight examined governorates in the same year amounted to 2,881,392 housing units distributed among different housing typologies (Syria Central Bureau of Statistics). The conflict further intensified the rural-urban migration. Starting from the mid-to-late 1990s, Syria experienced rural-to-urban migration spurred by drought and environmental degradation. Around 40,000 to 60,000 families migrated from rural to urban areas each year because of droughts. The conflict provided a major boost to migration. Today, around 72.6 percent of thE population (13.7 million people) is living in urban areas in Syria, and these urban areas have undergone dramatic demographic change. Those directly impacted by the conflict, like Homs and eastern Aleppo, have been experiencing an exodus of residents that have fled violence, destruction, and the collapse of basic services. Other areas, promising relative security and consistency of services, have experienced large influxes, with the net result being a significant concentration of the population (internally displaced persons (IDPs) and hosts) in urban areas. Table 2.1: Housing Damage in Cities City Units in 2010 Damaged units in 2017 Damaged units in 2017 (% of 2010 value) Partial damage Destroyed Partial damage Destroyed Aleppo 662,323 154,465 49,830 23.3 7.5 Raqqa 55,746 7,214 2,418 12.9 4.3 Dar'a 30,532 3,678 922 12.0 3.0 Douma 18,786 3,476 884 18.5 4.7 Dayr az-Zawr 48,430 15,107 4,849 31.2 10.0 Homs 216,191 37,350 12,526 17.3 5.8 Hama 101,902 2,531 3,020 2.5 3.0 Idlib 46,014 1,164 2,916 25.3 6.3 Tadmur 3,178 1,042 272 32.8 8.6 Kobani 20,801 2,226 814 10.7 3.9 Total 1,181,813 238,311 78,339 19.8 6.5 Source: World Bank staff calculations. 22 The Wrath of Conflict In conflict-affected cities, physical destruction has been extensive. The conflict in Syria has severely impacted the housing sector, where a huge number of residential units have been partially damaged or fully destroyed, housing investment disrupted, quality of living space deteriorated, and land and property rights severely challenged. At the city level, an estimated total of 316,649 housing units were exposed to impact, with Aleppo bearing the largest share at 64 percent of the impacted urban housing, followed by Homs at 16 percent. The damage across the examined cities constitutes an estimated 78,339 residential units destroyed and 238,311 units partially damaged (table 2.1). Across the 10 cities, 27 percent of the housing stock has been impacted, with 7 percent destroyed and 20 percent partially damaged. The percentage vary across cities, with destroyed building percentage ranging between 3 and 10 percent of the baseline housing stock, while the percentage of partially damaged units ranges from 2.5 to 33. Similarly, damage varies across neighborhoods within cities as well (maps 2.2 and 2.3). Governorate-level assessments provide a broader sense of the damages in the housing sector across Syria. The damage across the eight governorates translates into an estimated 220,826 residential units destroyed and 649,449 partially damaged.12 In total, about 32 percent of the housing stock in the eight governorates has been impacted, with 9 percent destroyed and 23 percent partially damaged (table 2.2). The percentages vary among governorates, with the percentage of destroyed units ranging between 3 and 10 percent of the baseline housing stock, while the percentage of partially damaged units ranges from 3 to 31 percent. Map 2.2: Neighborhood-Level Housing Damage Aleppo Dar’a 12 The physical status of each unit was classified in three categories: destroyed (> 40 percent damage), partially damaged (< 40 percent damage), and no damage. THE TOLL OF WAR 23 Dayr Az Zawr Douma Hama Homs Idlib Kobani Raqqa Tadmur Source: World Bank staff calculations. 24 The Wrath of Conflict Map 2.3: Governorate-Level Estimates of Housing Damage ESTIMATED DAMAGE TO SYRIA’S HOUSING SECTOR (within the 8 surveyed governorates) KOBANI 22% of all housing partially damaged Jarablous 8% of all housing destroyed ‘Afrin ‘Ayn al-’Arab ‘Azaz (Kobane) Tel Abyad al-Bab ALEPPO Harem Central Raqqa Mt. Simeon (Raqqa City) IDLIB (Aleppo City) Jisr RAQQA ash-Shughour Central Idlib as-Safira Manbij Ariha (Idlib City) ath-Thawra Central Dayr as-Zawr (Dayr az-Zawr City) Ma’arat an-Nu’man DEIR-EZ-ZOR Estimated number of damaged housing units 3,679-7,057 al-Mayadiin 7,058-12,281 12,282-17,240 Albukamal 17,241-30,749 30,750-184,847 ALEPPO RAQQA IDLIB DAYR AZ-ZAWR Governorate Est. apt. units Est. other types Aleppo 319,209 461,750 Idlib 46,396 182,922 Day az-Zawr 28,128 122,537 Raqqa 12,669 114,911 Governorates not included in coverage 5.67% 1.62% 11.36% 1.56% 0.55% 5.23% 1.49% 3.25% 5.57% 17.05 22.33 22.77 15.27 64% 65.65% 75.27% 64% 6.38% 4.47% 6.51% IDLIB ALEPPO RAQQA DAYR AZ-ZAWR Partially damaged apartment units Partially damaged other types No damage to housing Destroyed apartment units Destroyed other types THE TOLL OF WAR 25 Estimated number of damaged housing uits 2,007-4,895 4,896-10,805 as-Suqaylabiyya 10,806-21,255 Central Hama Mhardeh (Hama City) as-Salmiyya 21,256-38,399 Masyaf HAMA 39,400-78,984 ar-Rastan al-Makhrem Telkalakh HOMS al-Qusayr TADMUR Tadmur Central Homs (Palmyra) (Homs City) an-Nabk Yabroud al-Qutayfa az-Zabadani at-Tal DOUMA Darayya Douma Qatana Central Rif Dimashq as-Sanamayn Izra’ DAR’A Central Dar’a (Dar’a City) HAMA Governorate Est. apt. units Est. other types HOMS Rif Dimashq 273,785 253,610 RIF DIMASHQ Homs 122,048 190,882 Hama 65,342 202,779 DAR’A Dar’a 18,080 120,080 Governorates not included in coverage 0.87% 2.35% 3.11% 3.66% 1.05% 3.47% 5.15% 10.87% 9.06% 6.67% 24.34% 8.21% 16.59% 84.96 64.64% 64% 76.03% 4.78% 6.95% 3.23% HAMA HOMS DAR’A RIF DIMASHQ Partially damaged apartment units Partially damaged other types No damage to housing Destroyed apartment units Destroyed other types Source: World Bank staff calculations. 26 The Wrath of Conflict Table 2.2: Housing Damage Estimates across Governorates Governorate Units in 2010 Damaged units in 2017 Damaged units in 2017 (% of 2010 value) Partial damage Destroyed Partial damage Destroyed Aleppo 889,884 236,947 68,743 23.3 7.5 Raqqa 145,374 27,054 8,899 12.9 4.3 Dar'a 157,430 44,081 12,596 12.0 3.0 Rif Dimashq 600,955 103,794 40,259 18.5 4.7 Dayr az-Zawr 171,679 48,070 13,734 31.2 10.0 Homs 356,577 97,936 28,144 17.3 5.8 Hama 305,518 18,402 27,547 2.5 3.0 Idlib 261,302 73,165 20,904 25.3 6.3 Total 2,888,719 649,449 220,826 22.8 8.6 Source: World Bank staff calculations. The damage reported here should be interpreted as lower bound estimates. First, the damage assessments are based on actual physical conditions, as seen from January 2017. Therefore, the actual damage inflicted by the conflict may be higher if there has been some reconstruction in the area. Although it was not possible for this study to estimate the damage without reconstruction, rough estimates of reconstruction are summarized in map 2.4. Kobane (‘Ayn al-‘Arab) district in northeastern Aleppo governorate has witnessed one of the most extensive reconstruction campaigns anywhere in Syria over the course of the conflict, from the time when predominately Kurdish militias drove Islamic State fighters from the area in 2015. Second, the analysis presented here does not capture variations in the quality of the housing units. UN-OCHA (2016) reports that with massive migration to safer areas, adequate housing is simply not available in many locations, resulting in people sheltering in substandard housing (UN-OCHA 2016). The pressure on rental demand in these perceived safe urban areas has increased rental rates at an estimated range of between 500 and 1,000 percent. Families often inhabit (sometimes even rent) shells of houses (apartments without finishing, windows, or doors) after sealing the open windows and doors with plastic sheets and curtains. These should unambiguously have an impact on the actual values of damage. Third, the damage assessment does not capture the effects of conflict on housing investments that would take place in the absence of conflict. A background note to this report (Akbar 2017) performs back-of-the-envelope calculations to come up with a ballpark measure of these foregone investments. Accordingly, the loss of new housing stock is estimated at 330,784 unbuilt residential units across the eight governorates. THE TOLL OF WAR 27 The destruction in Syrian cities has also created another problem: debris. The conflict-related damage has led to large-scale debris accumulation in Syrian cities, which may require massive amounts of clearance, transportation, and processing or disposal. In a pilot study conducted for this report, remote-sensing-based analysis was used to analyze the extent and concentration of debris accumulation in Aleppo and Homs (map 2.5). The results show that about 14.9 and 5.3 million tons of debris have accumulated in these cities, respectively. To put these figures in perspective, in Aleppo, it would take about six years of continuous work and 26 million truck-kilometers to clear the debris. In comparison, in Homs, it would take about 2.5 years and 2.3 million truck-kilometers. It is also important to note that the debris removal and processing poses significant organizational challenges: it is costly; it takes time to clear; and the operations impose further environmental risks, including dust, carbon dioxide emissions, and water resource pollution. Thus, locating crushing sites often faces resistance from locals. Map 2.4: Reconstruction, by Governorate and District Aleppo Raqqa Idlib Aleppo Raqqa Idlib Hama Dayr Az-Zawr Homs Dayr Az-Zawr Hama Rif Dimashq Dar’a Homs Rif Dimashq Dar’a GOVERNORATE RECONSTRUCTION By District (ADM2) Relative Reconstruction Level Extensive Moderate Limited to none 28 The Wrath of Conflict Map 2.5: Calculated Debris Accumulation, as of February 2017 Aleppo Homs Source: World Bank staff calculations. THE TOLL OF WAR 29 Water and Sanitation Infrastructure The Syrian economy became increasingly vulnerable to water scarcity in the decades preceding the conflict. Syria is technically not a water scarce country like some of its neighbors.13 However, with policies that do not encourage the sound management of resources, it has been perceived as water scarce (Barnes 2009). The combination of drought, poor governance, and highly inefficient flood irrigation, combined with that around half of the irrigation water comes from groundwater pumping, has led to a situation of rising production costs as well as depleting groundwater tables (Gleick 2014). Many households drill their own wells in rural and urban areas; control over drilling of wells is limited. Around 78 percent of all groundwater withdrawals are deemed unsustainable in Syria (Wada, van Beck, and Bierkens 2012). Water shortages were common before the conflict. Potable water coverage was almost universal prior to the conflict; however, most regions suffered from shortages, particularly during the summer when supply was rationed to only for a few hours a day. Demand management has received much less attention. Illegal connections in new settlements located in urban areas were common before the crisis, and tariffs were at below cost-recovery rates. More than two-thirds of subscribers paid the lowest tariff, equal to 6 U.S. cents per cubic meter, as reported in 2009. The conflict has significantly deepened the problems in access to water. This study assessed the physical damage in 457 water supply and sanitation infrastructure assets across eight governorates (table 2.3). Overall, nearly two-thirds of the water treatment plants, half of the pumping stations, a third of the water towers, a quarter of the sewage treatment plants, and a sixth of the wells have been destroyed or partially damaged across Syria. However, the damage exhibits significant variation across regions. For instance, in Raqqa, almost all the infrastructure remained damaged at the time of this assessment. In contrast, almost all assets were intact in Kobani, reflecting the extensive reconstruction efforts after the liberation of the city from the ISIS siege. Physical damage does not fully reflect the mounting complexity of delivering water and sanitation services. The remote-sensing-based damage assessments presented here do not identify directly the level of damage to piped networks for water distribution, drainage, or sewerage, because of difficulties in assessing the functionality of subsurface assets. In addition, because water is pumped by using electricity, the functionality of water services is determined by the status of the electric grid and the availability of 13 Syria is characterized by an arid to semi-arid climate and generally has a high level of climatic variability, that is, part of the country is very green and receives ample rainfall; other parts are very dry. Most of the rainfall occurs during the winter months. The coastal region receives high levels of rainfall; other parts of the country receive significantly less. Surface water is estimated at 10 billion cubic meters and groundwater at 6 billion cubic meters annually. Irrigation consumes around 87 percent of the available water resources, versus 9 percent for domestic use and 4 percent for industry. 30 The Wrath of Conflict generators and fuel. Lack of electricity was the main reason for service outages, with damage to pumping stations and piping networks being the second most common problem. Table 2.3: Damage in Water and Sanitation Infrastructure (Eight Governorates) Asset type Baseline Destroyed Partially Total Percentage of coverage damaged damaged assets with damage Well 234 3 29 32 14% Water Tower/Tank 176 18 34 52 30% Water Treatment Plant 8 0 5 5 63% Sewage Treatment Plant 4 0 1 1 25% Dam 6 0 0 0 0% Other Drainage Str. 3 1 1 2 67% Pumping Station 21 0 12 12 57% Storage Reservoir 2 0 0 0 0% Water/Sanitation Office 3 2 1 3 100% Source: World Bank staff calculations. Water shortages have prompted various coping mechanisms. Syrians have reverted to alternative mechanisms to restore water access, especially during long episodes of public water network outages. According to the United Nations, in Hama, 60 new wells were developed by the water authority, to maintain pipeline supply. In Aleppo, more than 300 old and new wells provided much-needed water supplies, including private and mosque wells. There were more than 65 public water trucks delivering water from more than 550 holding tanks that ranged in size between 5 and 95 cubic meters, which were then refilled by water from wells. The city is working on installing 140 more holding tanks in eastern Aleppo, and there are more than 1,500 private water trucks providing water to residents. In addition to shortages, there are also many complaints about the quality of water and the soaring price of water trucks ($12 for 1,000 liters of water). Power Infrastructure The electricity sector was already weak and in need of reform before the conflict. Demand for electricity was increasing steadily and at a faster rate than expansion in generation capacity, leading to outages and load shedding. Load shedding in 2009 was 671 gigawatt-hours (GWh), compared with 391 GWh in 2008, and was expected to continue in the following years. Significant technical and nontechnical losses and deteriorating infrastructure resulted in poor reliability and insufficient supply of electricity, affecting economic and THE TOLL OF WAR 31 social development. Technical losses of the transmission and distribution network amounted to 15.9 percent in 2010. Nontechnical losses (caused by theft, errors in meter reading and billing, and so forth) were 9.9 percent of the total system demand in 2010. The demand-supply gap and the high level of losses reflected insufficient investment and poor sector governance. New investment was difficult to attract because of low domestic electricity tariffs, which averaged to 4.42 U.S. cents/kilowatt-hour (kWh), compared with 12.9 U.S. cents/kWh for electricity exports. Physical damage to electricity infrastructure from the conflict has been severe, but the functionality of the transmission grid has so far been largely maintained despite that it spans different areas of control. According to media reports and the damage assessment, secondary infrastructure (power lines and substations) appears at least partially functional. However, although the national grid remains largely functional (map 2.6), with major transmission links still in place, service quality is very poor in most of the country, and some areas have been disconnected from public supply for years. The three hydroelectric dams and five of the 13 major power plants for which information was available remain fully operational, and two other power plants are partially functioning (map 2.6). The two major plants that were fully destroyed by fighting are the Zeyzoun Power Plant in Idlib governorate (544 megawatts (MW) installed capacity) and Aleppo Thermal Power Station in Aleppo governorate (1,065 MW installed capacity). Al-Teem Power Station, a minor electricity plant near the city of Deir ez-Zor (100 MW installed capacity), is also fully destroyed. Although the institutional arrangement is unclear, electricity is distributed outside the boundaries of government controlled areas. Anecdotal evidence shows that the government implicitly and explicitly coordinates with the different rebel groups to ensure that power plants are serviced even if those plants are under rebel control. There is no further information on how this arrangement works. Disruptions to service delivery have been overwhelming, as fuel shortages and conflict-driven constraints to operation and maintenance have led to a sharp drop in public power supply. Power generation declined to 16,208 GWh in 2015, compared with 43,164 GWh in 2010, a drop of 62.5 percent. Much of this decline appears to be due to fuel shortages, as available generation capacity declined by only about 30 percent in the same period. The fuel shortages appear to apply to heavy fuel oil and natural gas. The consumption of heavy fuel oil and natural gas for power generation fell by 70 and 54 percent, respectively, between 2010 and 2015. Demand has also declined due to the conflict, with annual peak load falling from 9,595 MW in 2012 to 5,471 MW in 2015. Nevertheless, the fact that the drop in generation is much sharper than in peak demand suggests that the main constraint is on the supply side. 32 The Wrath of Conflict Map 2.6: Estimated Damage to Major Power Sector Assets 2 Destroyed power plants Both power plants are non-functioning 6 Partially damaged power plants 4 Power plants are non-functioning 2 Power plants are partially functioning 9 Power plants and 3 dams with no damage 8 Power plants are dams are functional 1 Power plant’s functionality is unknown Source: World Bank staff calculations. Table 2.4: Status of the Electricity Grid and Major Power Infrastructure in Seven Cities in Syria (%) City Physical status Operational status No damage Partial damage Destroyed Functioning Partial function No function Aleppo 89.39 3.81 6.80 0.33 99.00 0.66 Idlib 94.29 1.43 4.29 92.86 2.86 4.29 Hama 99.24 0.51 0.25 0.25 99.75 0.00 Kobani 66.67 33.33 0.00 66.67 0.00 33.33 Tadmur 30.00 70.00 0.00 10.00 90.00 0.00 Douma 88.24 5.88 5.88 23.53 64.71 11.76 Raqqa 94.59 0.00 2.70 13.51 86.49 0.00 Source: World Bank staff calculations. THE TOLL OF WAR 33 The decline in electricity supply has caused major disruptions in Syrians’ lives (table 2.4). The majority of cities receive only a few hours of electricity a day. The government has applied a rationing policy via load shedding throughout the country. Electricity availability varies considerably among districts within a single city, as well as between urban and rural areas. Schedules adjust irregularly and with little warning, primarily in accordance with fuel availability. Strategic diversion of electricity is common, for example, for harvests, irrigation, or the need to maintain critical infrastructure, such as hospitals. Electricity shortages have induced various coping mechanisms. Because of the crisis, black markets for fuels and electricity-generating supplies have flourished. Fuel prices vary significantly across cities, depending on accessibility and different government pricing schemes across regions. Commercial and private generators as well as car batteries (charged from the grid or other sources) serve as the main source of electricity for households in many conflict-affected cities. Surveys conducted in Aleppo, Homs, and Idlib show that only a small percentage of the respondents relied solely on the public grid for power generation (figure 2.1). Many households converted small electrical appliances to work with 12-volt car batteries. The Syrian government continues to subsidize electricity, but it is unclear to what extent. Electricity subsidies continue to pose a burden on government resources. Although we do not have data on the actual amount of electricity subsidies, local media sources report targeted increases in electricity prices in an attempt to free government resources. Electricity tariffs vary based on the type of consumer, volume of electricity used, and voltage. Most electricity subsidies go to household consumption at lower levels of monthly usage. Figure 2.1: Sources of Electricity Supply in Three Conflict-Affected Cities 4.7% 4.7% 4.7% 1.9% 5.5% 0.9% 18.7% 21.3% 31.8% 29.1% 49.5% 49.1% 60.7% 64.8% 67.3% ALEPPO IDLIB HOMS Always from the grid Car batteries No alternative Private generator Purchase from private providers Source: United Nations. 34 The Wrath of Conflict Transportation Infrastructure Syria's roadway network expanded rapidly before the conflict, to keep pace with the fast-growing economy and population. In the decade preceding the conflict, the road network grew by 10 percent, of which 70 percent was asphalted. This increase was more than matched by the increase in the number of vehicles. Six years of economic reforms, including a reduction in import taxes from more than 250 to 50 percent, and the introduction of bank credits for purchasing cars, created an unprecedented surge in the number of vehicles on the road. According the Syrian Central Bureau of Statistics, the total number of registered vehicles registered in Syria increased from 1.2 million in 2006 to 2.1 million in 2010. Syria’s well-developed network of motorways is located mainly in the western half of the country (map 2.7). The Motorway System is a divided, multi-lane highway along the Damascus–Homs–Hama–Aleppo corridor, with extensions to the Mediterranean Ports of Tartous and Lattakia, and to Jordan to the south and Lebanon to the west. The secondary road network was generally in good condition and additional road construction was underway to extend the major highway network to include a link from Lattakia to Aleppo. Motorways and secondary roads have weight-load limits, and weigh stations operate to limit truck overloading. The eastern part of the country is only connected through two-lane roads due to the sparsity of the population. In 2010, the total distance of the road network in Syria was 69,837 kilometers (km), of which 45,345 km was paved (including 1,103 km of expressways). M5 is the most important motorway in the country. At 474 km in length, it functions as the backbone of the national network. It connects the border with Jordan in the south with Damascus, the capital, and continues further north to Aleppo, the country's second largest city. Other cities connected by this motorway are Daraa, Al Nabk, Homs, and Hamah. The conflict has damaged the transport infrastructure significantly. The physical impact of the conflict can be classified in two major categories: (i) damage due to exploding bombs and ordinance, causing craters and surface depression; and (ii) collateral damage due to falling debris from destroyed buildings, road cuts, and checkpoints, making large sections of the road inaccessible and restricting the movement of people and vehicles. The data analysis provides an estimate of the total linear length of roads and bridges damaged and undamaged in the 10 surveyed cities. The estimate extends beyond the surveyed roads within a city, to include other roads as well. Interpretation of this type of information comes from comparing satellite imagery maps with the data. THE TOLL OF WAR 35 Map 2.7: Major Roads and Railways Source: Courage Services. Figure 2.2: Damages to Intra-City Roads 5.0 4.5 4.0 Kilometers (Thousand) 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0.0 ALEPPO HOMS RAQQA HAMA DAYR TADMUR DAR’A IDLIB KOBANI DUMA AZ-ZAWR Not damaged Damaged Source: World Bank staff calculations. 36 The Wrath of Conflict Primary roads and trunk roads have received the most damage within cities.14 Throughout the 10 cities, a total of 30 km of roads and nine bridges were damaged or destroyed. The majority of the road damage in the 10 cities is conflict-related, mainly attributed to bombardment, collateral damage, lack of maintenance, and man-made barricades. Aleppo is the most impacted city, accounting for the majority of the damages in the 10 cities, in absolute and relative terms (figure 2.2). With the latest escalation of the conflict, damage to Aleppo’s road transportation sector doubled from September 2016, reaching 44 percent damage in February 2017. However, the impact on road functionality extends far beyond direct road damage. Large segments of roads are often rendered inaccessible due to direct damage to short segments that has substantial impact on overall mobility. Airports, railways, and ports have also sustained damage. Syria has 26 airports with paved runways, and five airports with runways of more than 3,000 meters. Prior to the conflict, Syria’s three international airports, at Damascus, Aleppo, and Latakia, together handled an annual average of 2.5 million passengers on 13,000 flights. Currently, the Damascus Airport is the only airport operating international flights. The Syrian railway system (2,423 km in total length) is nonoperational due to the conflict. The conflict has not only led to physical destruction of the road network, but also disrupted mobility and eliminated connectivity within the country and the region. Road-based travel distances across the country have increased as conditions related to conflict, weather, and other factors have forced travelers to take less direct, alternative routes to reach their destinations. Figure 2.3 shows the minimum road-based travel distance between the most prominent cities in each governorate before the conflict. Next, using publicly available reporting and satellite imagery on damage and other restrictions, this analysis assessed the operational status of each route segment (as of March 2017), to calculate the minimum distance that needs to be traveled by civilians to go from one city to another. In nominal terms, the connectivity of Dayr az-Zawr was the most diminished. Compared with pre-conflict travel, it took an additional 155 km to reach Damascus, 170 km to Dar’a, and 187 km to Hasaka. The latter was the largest proportional increase in travel distance (42.6 percent) among all the cities analyzed here. Health Care Facilities Before the conflict, physical infrastructure capacity in health care was below the averages in the Middle East and varied significantly by region. From 2005 to 2010, hospital bed ratios increased from 1.4 to 1.5 per 1,000 people, 14 The damage assessment for the road transport sector in Syria covers the following road asset classes: motorway, trunk road, primary road, secondary road, tertiary road, track, residential road, service road, and bridges. The assessment only covers physical damage to roads and bridges, and does not include additional inter-urban infrastructure assets, such as traffic lights, culverts, road and highway inventories, private cars, fuel facilities, and public buses. THE TOLL OF WAR 37 Figure 2.3: Changes in Connectivity: Pre-Conflict versus De Facto Distances, March 2017 PRE CONFLICT GOVERNORATE CAPITAL, TOTAL TRAVEL LENGTH IN KILOMETERS Damascus Dar'a Dayr az-Zawr Aleppo Idlib al-Qunaytara Homs Hama Hasaka Raqqa Tartous Latakia Sweida Damascus Dar'a 101 Dayr az-Zawr 448 544 Aleppo 350 450 321 Idlib 306 407 384 62 al-Qunaytara 64 70 512 413 370 Homs 162 263 359 188 145 226 Hama 211 311 380 142 99 274 49 Hasaka 618 718 253 364 425 681 458 442 Raqqa 432 532 137 197 260 495 272 256 189 Tartous 253 354 455 233 182 317 97 110 548 362 Latakia 320 420 496 175 116 383 158 140 538 372 140 Sweida 101 54 517 438 394 95 250 299 706 520 341 408 CURRENT GOVERNORATE CAPITAL, TOTAL TRAVEL LENGTH IN KILOMETERS Damascus Dar'a Dayr az-Zawr Aleppo Idlib al-Qunaytara Homs Hama Hasaka Raqqa Tartous Latakia Sweida Damascus Dar'a 111 Dayr az-Zawr 603 714 Aleppo 391 502 354 Idlib 359 470 428 74 al-Qunaytara 68 78 670 458 426 Homs 170 281 442 221 190 237 Hama 225 336 421 167 138 292 55 Hasaka 738 849 441 416 488 805 577 556 Raqqa 460 570 163 211 285 527 299 278 278 Tartous 259 370 528 274 206 326 97 110 663 385 Latakia 328 438 548 213 140 394 164 140 627 405 140 Sweida 102 83 694 482 450 108 262 316 829 551 350 419 CURRENT GOVERNORATE CAPITAL, TOTAL TRAVEL LENGTH % INCREASE Damascus Dar'a Dayr az-Zawr Aleppo Idlib al-Qunaytara Homs Hama Hasaka Raqqa Tartous Latakia Sweida Damascus Dar'a 8.2% Dayr az-Zawr 25.7% 23.7% Aleppo 10.6% 10.2% 9.3% Idlib 14.8% 13.4% 10.4% 15.3% al-Qunaytara 4.9% 9.9% 23.6% 9.7% 13.2% Homs 4.8% 6.5% 18.7% 15% 23.7% 4.7% Hama 6.3% 7.2% 9.7% 14.7% 28.6% 5.9% 10.9% Hasaka 16.3% 15.4% 42.6% 12.5% 13% 15.3% 20.6% 20.5% Raqqa 6.1% 6.7% 15.7% 6.6% 8.9% 5.9% 8.9% 7.8% 32.1% Tartous 2.2% 4.3% 13.7% 14.9% 11.7% 2.7% 0% 0.1% 17.4% 6% Latakia 2.5% 4.1% 9.4% 17.6% 17.7% 2.8% 3.2% 0% 14.3% 8% 0% Sweida 0.8% 34.4% 25.5% 9.2% 12.5% 12.1% 4.3% 5.5% 14.9% 5.7% 2.5% 2.7% Source: World Bank staff calculations. 38 The Wrath of Conflict hovering slightly below the MENA average of 1.6 beds/1,000 people in 2009 (WDI 2017). According to the Ministry of Health, there were 498 hospitals in Syria in 2010 (MOH 2011). Bed capacity also showed large geographic variation: Idlib, with a population of roughly 1.5 million, accounted for only 3.9 percent of total beds, while Aleppo, with a population of about 2.4 million, had 20.1 percent of total beds (Syria Central Bureau of Statistics 2011). Although this study did not have full access to relevant data, it is likely the governorates in the northeast, including Hasaka and Raqqa, had significantly inferior access to health care facilities. Hospitals have not been immune from attacks in the conflict; they have been specifically targeted and used for military purposes. Physicians for Human Rights has documented attacks on 400 facilities (some multiple times) and the deaths of 768 health personnel in Syria between March 2011 and the end of July 2016. These figures are corroborated by the Syrian American Medical Society, which documents a nearly 90 percent increase in number of attacks on medical facilities and personnel in Syria following the May 3, 2016 passage of United Nations Security Council Resolution 2286, which condemns attacks on medical facilities and personnel in conflict zones (Syrian American Medical Society 2017). Unfortunately, since these data were collected in July 2016, the attacks on health facilities have continued, with further bombing of key hospitals in Aleppo in April, May, and December 2016. It is now estimated that an attack takes place on medical facilities or personnel every 48 hours (Syrian American Medical Society 2017). Six in 10 health facilities in the 10 surveyed cities, including hospitals, medical centers, medical points, and polyclinics, have faced some form of damage as of February 2017. The health infrastructure is in grave condition. In eight of the 10 cities surveyed, more than half of all types of health facilities were impacted by the conflict. Overall, 16 percent of all health facilities were completely destroyed and 42 percent were damaged partially. The degree of physical damage suggests that local communities’ access to the quality health care services they need is severally limited by the civil conflict in all 10 cities. This situation affects not only those who need emergency medical care, but also patients who require care for noncommunicable diseases, chronic conditions, maternal and child health, and mental health services. Activities related to disease surveillance are also hampered, exposing the population to the risks of disease outbreaks. The degree of damage to health infrastructure varies considerably across the 10 cities. In Aleppo, 14 hospitals and 57 medical centers were destroyed as of December 2016. This corresponds to 35 percent complete destruction of pre-conflict health infrastructure in Aleppo, one of the highest across the 10 cities (figure 2.4). The damage in Raqqa and Dayz az-Zawr has also been extensive. As of February 2017, in both cities, almost nine in 10 health facilities suffered some form of damage. The share of health facilities that were partially damaged reached 80 percent in Raqqa and 79 percent in Dayz az-Zawr. The THE TOLL OF WAR 39 Figure 2.4: Damage to Total Health Facilities, by City 120% 100% Percentage % 80% 60% 25% 28% 60% 79% 80% 100% 40% 69% 52% 20% 50% 35% 33% 2% 20% 6% 10% 8% 0% ALEPPO HAMAH IDLIB DARAH HOMS DAYR TADMUR RAQQA DOUMA KOBANI AZ-ZAWR Completely destroyed Partially damaged Source: World Bank staff calculations. health infrastructure in Idlib also suffered severe destruction. As of February 2017, eight in 10 health facilities experienced some degree of damage. Damage also varied by facility type and ownership. As of February 2017, 54 percent of all hospitals in the 10 cities received some form of damage. The lower levels of care were also not shielded from the conflict. As of February 2017, approximately 68 percent of all medical centers, 84 percent of specialized medical centers, 75 percent of medical points, and 51 percent of all polyclinics experienced some form of physical damage (figure 2.5). Among damaged hospitals, there was considerable variation between public and private hospitals with respect to the degree of damage. As of February 2017, an estimated 60 percent of all public hospitals and 48 percent of private hospitals were completely destroyed or partially damaged in the 10 surveyed cities. More specifically, about 11 percent of all public hospitals were completely destroyed and almost 50 percent of them were partially damaged. In the private sector, the proportion of hospitals that were destroyed reached 9 percent, and almost 40 percent experienced some level of damage. The targeting of health facilities and medical personnel has led to provider flight and a hollowing out of the entire health system. Data on the number of physicians, nurses, and other health workers who fled are limited, but efforts are ongoing to map the size of this workforce exodus and the location. According to Physicians for Human Rights, at least 15,000 of Syria’s 30,000 physicians have left. Particularly hard-hit areas have seen even greater losses. Of Aleppo’s 6,000 physicians at the start of the conflict, fewer than 70 were still in the city as of 2014 (PHR 2015). 40 The Wrath of Conflict United Nations organizations and nongovernmental organizations working inside Syria report an acute shortage of health professionals. Often medical residents or, in several cases, veterinarians are the only providers in towns under siege. There is an especially acute shortage of female doctors (especially obstetricians and gynecologists), which has severely affected access to health care for half the population. Due to the unpredictable security situation, women are increasingly scheduling caesarian sections (as opposed to having a normal vaginal delivery, which is difficult to predict) and leaving as soon as they come out of anesthesia due to fear of their hospital being bombed. This procedure is considered a major surgery and usually requires an inpatient stay of four to five days to ensure safe recovery. Health care providers are often forced to use innovative methods like providing diagnosis via Whatsapp or other video-over-Internet applications with specialists abroad, due to physician shortages; training future cadres of health professionals through online courses held in bunkers; and developing cross-border pharmaceutical supply chains and payment systems to allow for medicines and salaries to reach besieged towns. Figure 2.5: Damage to Health Facilities, by Facility Type, February 2017 64% 38% 79% 49% 75% 38% 39% 42% 20% 30% 9% 11% 11% 5% 13% PRIVATE PUBLIC HOSPITAL HOSPITALS MEDICAL SPECIALIZED MEDICAL POLYCLINIC HOSPITAL HOSPITAL (UNKNOWN) (ALL) CENTER MEDICAL POINT (ALL) CENTER Completely destroyed (%) Partially damaged (%) Source: World Bank staff calculations. Physical damage reflects only a subset of the effects of the conflict on the health sector (map 2.8). The physical damage discussed here does not capture many direct inputs to the provision of health care service, including machinery, tools, and other medical supplies. Moreover, direct destruction of infrastructure in other sectors, including transportation, energy, and water, surely further cripples access to health care and service delivery, including maintenance of critical drug supply chains and referral networks. More people may have been killed in Syria due to a breakdown of the health system than due to direct fatalities from the fighting. The destruction of the existing health infrastructure has led to a deterioration of health status across the country, with a reemergence of communicable diseases, heightened impact of noncommunicable diseases, and increasing burden of injuries. For THE TOLL OF WAR 41 instance, polio, long thought to be eliminated, has reemerged and spread across the country, due to the inability of routine immunizations to take place (Sparrow 2014). As summarized by Zaher Sahloul, the head of the Syrian American Medical Society, some 200,000 people are believed to have died in Syria because they did not have access to routine medical care as of February 2014 (Baker 2014). These are in effect a “secondary death toll” that is even higher than those killed by bombs and firearms (Baker 2014). “These are the women who died in labor because there was no one to do a C-section, or the men who have a heart attack and can’t find a physician, or have complications from diabetes. People are dying of chronic diseases that three years ago would have been completely manageable” (Baker 2014). Mental health conditions are worsening, with an entire generation of children showing signs of “toxic stress.” The Syrian Public Health Network estimated an increase in anxiety, depression, and post-traumatic stress disease as a direct result of the conflict (Syria Public Health Network 2016). This is especially acute in children, who are being classified as having levels of “toxic” stress, with more than 70 percent of children recently surveyed reporting they were experiencing bed wetting, a sign of post-traumatic stress disorder (Save the Children 2017). It is not difficult to see why: two-thirds of children are said to have lost a loved one, or had their house bombed or shelled, or suffered conflict-related injuries. 42 The Wrath of Conflict Map 2.8: Governorate-Level Damage to the Health Sector ESTIMATED DAMAGE TO SYRIA’S HEALTH SECTOR (within the 8 surveyed governorates) KOBANI Jarablous ‘Afrin ‘Ayn al-’Arab ‘Azaz (Kobane) Tel Abyad al-Bab ALEPPO Harem Central Raqqa Mt. Simeon (Raqqa City) IDLIB (Aleppo City) Jisr RAQQA ash-Shughour Central Idlib as-Safira Manbij Ariha (Idlib City) ath-Thawra Central Dayr as-Zawr (Dayr az-Zawr City) Ma’arat an-Nu’man DEIR-EZ-ZOR Estimated number of damaged medical facilities 0-7 al-Mayadiin 8-18 19-28 Albukamal 29-62 ALEPPO RAQQA 63-172 IDLIB DAYR AZ-ZAWR 50% of all medical facilities partially damaged 16% of all medical facilities destroyed Governorates not included in coverage 22 71 30 23 61 156 39 87 23 17 77 5 5 5 10 13 3 3 3 5 IDLIB ALEPPO RAQQA DAYR AZ-ZAWR Private hospitals partially damaged/destroyed Public hospitals partially damaged/destroyed Specialized medical center partially damaged/destroyed Medical center partially damaged/destroyed Health facilities with no damage THE TOLL OF WAR 43 Estimated number of damaged health facilities 0-6 7-16 17-28 as-Suqaylabiyya Mhardeh Central Hama (Hama City) 29-53 as-Salmiyya Masyaf 54-111 HAMA ar-Rastan al-Makhrem Telkalakh HOMS al-Qusayr TADMUR Tadmur Central Homs (Palmyra) (Homs City) an-Nabk Yabroud al-Qutayfa az-Zabadani at-Tal DOUMA Darayya Douma Qatana Central Rif Dimashq as-Sanamayn Izra’ DAR’A Central Dar’a (Dar’a City) HAMA HOMS RIF DIMASHQ Governorates not included in coverage DAR’A 7 5 26 43 23 91 85 23 170 119 67 8 26 1 12 12 3 10 HAMA HOMS DAR’A RIF DIMASHQ Private hospitals partially damaged/destroyed Public hospitals partially damaged/destroyed Specialized medical center partially damaged/destroyed Medical center partially damaged/destroyed Health facilities with no damage Source: World Bank staff calculations. 44 The Wrath of Conflict Education Facilities Education outcomes were improving until 2010 in Syria, almost reaching the education Millennium Development Goal. In 2009, Syria’s primary net enrollment rate (NER) of 93 percent was equal to the average for the MENA region and higher than the average for middle-income countries (90 percent). The NER for secondary school was 67 percent for the same year, above the MENA average of 60 percent. The gender gap in enrollment was negligible, with an NER gender parity index of 0.98 in primary and 1.01 in secondary in 2009.15 Student learning outcomes in Syria were below international averages but higher than most MENA countries. In the 2007 Trends in Mathematics and Science Study, Syria ranked 32 and 39 of 49 countries for eighth grade science and mathematics, respectively. Compared with other countries in the MENA region, Syria’s eighth-grade students ranked third for their performance on the science assessment and fifth on the mathematics assessment. Education facilities have become military bases and targets during the conflict. Schools have been used as military quarters and informal shelters for displaced community members. Vocational institutes, secondary schools, and education offices were the highest targeted education institutions, with respective percentages of fully damaged facilities of 14.5, 14.7, and 33.3. Primary schools and universities remained almost intact, at 6.8 and 5.8 percent, respectively, for full damage. Education facilities in Aleppo have the highest occurrence of damage, with a total of 49 education facilities destroyed (map 2.9). Da’ra follows this with 21 education facilities destroyed, and Raqqa and Tadmur with seven facilities destroyed. Other governorates such as Kobani, Homs, Douma, and Daya Az-Zawr, have witnessed five or fewer facilities fully damaged, while Hama and Idlib had no education facilities that were fully destroyed. An estimated 57 percent of education facilities are still functioning in Syria. Many schools have limited access to electricity and water and sanitation facilities. In some cases, schools are being over utilized, for instance in Idlib, where the number of students exceeds the capacity of functional schools to absorb students, versus other governorates in the country where schools are highly underutilized due to population displacement and movement. 15 World Bank EdStats database. THE TOLL OF WAR 45 Map 2.9: Governorate-Level Damage to Syria’s Education Sector Facilities ESTIMATED DAMAGE TO SYRIA’S EDUCATIONAL SECTOR (within the 8 surveyed governorates) KOBANI Jarablous ‘Afrin ‘Ayn al-’Arab ‘Azaz (Kobane) Tel Abyad al-Bab ALEPPO Harem Central Raqqa Mt. Simeon (Raqqa City) IDLIB (Aleppo City) Jisr RAQQA ash-Shughour Central Idlib as-Safira Manbij Ariha (Idlib City) ath-Thawra Central Dayr as-Zawr (Dayr az-Zawr City) Ma’arat an-Nu’man DEIR-EZ-ZOR Estimated number of damaged educational facilities 37-110 al-Mayadiin 111-202 203-311 Albukamal 312-523 ALEPPO RAQQA 524-1,831 IDLIB DAYR AZ-ZAWR 53% of all educational facilities partially damaged 10% of all educational facilities destroyed Governorates not included in coverage 7.48% 5.65% 2.84% 7.98% 23.85% 26.62% 24.1% 42.11% 68.67% 67.73% 55.05% 67.92% IDLIB ALEPPO RAQQA DAYR AZ-ZAWR Primary school partially damaged/destroyed Secondary school partially damaged/destroyed Educational facilities with no damage 46 The Wrath of Conflict Estimated number of damaged education facilities 1-25 26-73 74-160 as-Suqaylabiyya Central Hama Mhardeh (Hama City) 161-364 as-Salmiyya Masyaf 365-804 HAMA ar-Rastan al-Makhrem Telkalakh HOMS al-Qusayr TADMUR Tadmur Central Homs (Palmyra) (Homs City) an-Nabk Yabroud al-Qutayfa az-Zabadani at-Tal DOUMA Darayya Douma Qatana Central Rif Dimashq as-Sanamayn Izra’ DAR’ACentral Dar’a (Dar’a City) HAMA HOMS RIF DIMASHQ Governorates not included in coverage DAR’A 1.71% 13.2% 24.97% 47.64 47.64% 45.47% 66.2% 67.21% 8.83% 6.89% 85.08% 6.89% HAMA HOMS DAR’A RIF DIMASHQ Primary school partially damaged/destroyed Secondary school partially damaged/destroyed Educational facilities with no damage Source: World Bank staff calculations. THE TOLL OF WAR 47 Demographic Dispersion and Loss of Lives and Opportunities Among all the consequences of the conflict, the effects on human lives and displacement have been the most dramatic. The pre-conflict population of Syria was estimated at 20.7 million in 2010 (World Development Indicators). Since 2011, the conflict has created a complex set of pressures on the country’s population. The most recent calculations by the International Organization for Migration (IOM) put the population within Syria at 18.8 million as of November 2016 (table 2.5). Data limitations render a precise and comprehensive decomposition of demographic changes impossible: conflict affects fertility rates and life expectancy alike. In addition, an important portion of demographic movements takes place informally: some refugees remain unregistered and, in certain cases, migrants do not factor into in-country population or refugee totals. There is significant uncertainty over conflict-related casualties in the country. The casualties that are directly related to the conflict are estimated to be between 400,000 (UN, as of April 2016) and 470,000 (Syrian Center for Policy Research, as of February 2016). This report relies on the IOM demographic indicators, because IOM provides aggregate numbers and an accompanying track of population status within the country at monthly intervals. Syria has become the largest forced displacement crisis in the world since World War II (box 2.1). Over half of the country’s pre-conflict population has been forcibly displaced. According to the United Nations High Commissioner for Refugees (UNHCR), the total number of Syrians presently registered as refugees outside the country in Lebanon, Turkey, Jordan, Iraq, Egypt, and North Africa is 4.9 million (table 2.6). In addition, more than 800,000 Syrian nationals are estimated to have sought asylum in Europe in 2015 and 2016. Many of these individuals have moved more than once, and have not been removed from registration lists in their first country of refuge. These numbers also do not include an estimated 0.4 million to 1.1 million unregistered Syrian refugees in Lebanon, Jordan, Turkey, and Iraq. The Syrians who fled to reach safety and reasonable living conditions endured very harsh conditions to get there. Departures were often quite rapid, with little time to consolidate assets or sell property—averaging fewer than seven days. Syrians without passports commonly traveled through Syria, a dangerous and expensive endeavor for those who live in the south of Syria, Homs, Hama, or Aleppo, then crossed the border into Turkey illegally. According to the REACH Initiative, the most commonly used crossing point was near the village of Khirbat al-Joz in Idlib governorate. In nearly all cases, money for bribes was reported to add to the cost of the journey through Syria. Once in Turkey, (as of late 2015) Syrians would typically head toward the Port of Bodrum or Izmir before taking a boat to Greece. Smugglers typically arrange boats for an average of US$1,200 per person to the Island of Lesbos. Other boats travel to Chios, Samos, and other islands close to the Turkish coast. 48 The Wrath of Conflict Depending on the weather and the condition of the boat, the journey was reported to take from 45 minutes to several hours. Smugglers did not commonly travel in the boats; instead, a passenger was nominated to navigate the vessel. Once in Greece, new arrivals often reported walking to points where they were registered by the Greek authorities and then boarded boats to mainland Greece. In late 2015, preferred onward destinations included Athens and Thessaloniki, before traveling by bus toward Evzonoi and the border with the Former Yugoslav Republic of Macedonia. From there, Syrians and others typically travel northward (sometimes by train) through Serbia (see map 2.10). Table 2.5: Key Demographic Indicators Indicator Number Source/Comment Syria: Pre-war population (2010) 20,720,602 World Bank (WDI figures) based on UN Population Division data and national census reports. Estimated population within 18,792,029 IOM. UN Population Division projections at Syria as of December 2016 18,564,000 for 2016. Estimated total number of 400,000 – UN (400,000 as of April 2016); Syrian Center for deaths as a result of conflict 470,000 Policy Research (470,000 as of February 2016). since 2011 Internally displaced persons 5,715,168 IOM (Jan 2017). UNOCHA lists 6.3 million IDPs in since 2011* Dec 2016, including the 513,833 persons “affected by shelter damage” and displaced within their own community as displaced. IOM does not. Registered Syrian refugees 4,857,617 UNHCR (December 2016, since 2011) Cumulative Syrian Asylum 884,461 UNHCR (April 2011-Oct 2016). *Note that those Applications arriving in Europe may remain registered in other countries. Note: IDP =internally displaced persons; IOM = International Organization for Migration; UN = United Nations; UNHCR = United Nations High Commissioner for Refugees; WDI = World Development Indicators. THE TOLL OF WAR 49 Box 2.1: The Terminology of Forced Displacement Forced Displacement “Forced displacement” refers to the coerced and involuntary movement of a person or persons away from their home or home region due to a well-founded fear of persecution. “Migrants,” as opposed to the “forcibly displaced,” choose to move not because of a direct threat of persecution or death, but primarily to improve their lives by finding better employment, education, or family reunion opportunities. It is an important distinction, although many population movements (like those to Europe from host countries outside Syria) are “mixed migration” flows, motivated by a combination of these factors. Refugee The 1951 United Nations Refugee Convention defines a “refugee” as any person who, “owing to well-founded fear of being persecuted for reasons of race, religion, nationality, membership of a particular social group or political opinion, is outside the country of his nationality and is unable or, owing to such fear, is unwilling to avail himself of the protection of that country.” The Convention defines a universal right of protection from persecution as well as from refoulement, and a right to access work, education, and health care. The Convention is nonbinding, although it is ratified by 145 countries. With the exception of Turkey (which retains a geographical limitation), the countries hosting the majority of Syrians have declined to join the Convention or the subsequent 1967 Protocol. Asylum Seeker An "asylum seeker" is a person who has applied for asylum under the 1951 Refugee Convention and the person remains an asylum seeker for as long as their application (or appeal against refusal of an application) is pending. For Syrians, The United Nations High Commissioner for Refugees generally approves all asylum seekers for refugee status. In its broader context, "refugee" means a person fleeing the diverse impacts of conflict but not necessarily fearing direct and personal persecution as defined by the 1951 Refugee Convention. Map 2.10: Primary Reported Routes through Turkey and the "Balkan Corridor" (as of September 2015) Source: REACH Initiative. 50 The Wrath of Conflict Table 2.6: Total Registered Syrian Refugee Population, by Country/Region, 2012–16 Country Jan 2012 Dec 2013 Dec 2014 Dec 2015 Dec 2016 Lebanon 6,916 138,213 884,017 1,159,396 1,011,366 Jordan 4,013 116,778 576,354 623,338 655,496 Turkey 9,500 174,598 560,129 1,622,839 2,814,631 Iraq 180 66,920 213,223 233,625 230,836 Egypt. Arab Rep. 924 12,836 131,707 138,212 116,013 North Africa n/a n/a n/a n/a 29,275 Total 21,533 509,345 2,365,430 3,777,410 4,857,617 Source: United Nations High Commissioner for Refugees. Unlike refugees, most IDPs remained near their place of origin. Among those who have fled within Syria, an average of 56 percent has chosen to remain within their governorate of origin. As a rule, this is typical of internally displaced populations—especially those with fewer resources and among IDPs who believe their displacement will be short-lived. The greatest percentage of intra-governorate displacement was in As-Sweida, due to its geographic isolation and the lines of conflict that encircle the province (figure 2.6). The greatest numbers of IDPs within their provinces were in Aleppo and Rural Damascus governorates. Most of those who fled their residence remained in Syria, with the exception of those from Idleb, Al-Hasakeh, and Latakia. The governorate of Aleppo reported the highest rate of displacement for flight within and outside Syria, however, with almost equal percentages of the displaced population fleeing within and outside the country. In contrast, the governorates of Idleb, Al-Hasakeh, and Lattakia reported more individuals fleeing out of Syria than within Syria (figure 2.7). Migration decisions are determined by shelter, security, and food at source and destination. A 2015 World Bank study of push and pull factors driving forced displacement and return in the cities of Aleppo, Homs, and Dar’a is illustrative of some of the factors informing the calculus of decision making, prompting movement (table 2.7). In each location, the dominant push factors driving populations from their homes were the lack of shelter, physical security concerns, and ultimatums by authorities. Ethnic division and partisan factors also drove residents from their home areas. Dominant pull factors included access to shelter, availability of food (bakeries/bread in particular), security, and kinship/partisan affiliations. THE TOLL OF WAR 51 Figure 2.6: Destination and Origin Governorates of IDPs Top governorates of destination for An averagre of 56% of IDPs are displaced 5,542,996 people fleeing their locations within their governorate of origin... Intra-governorate displacement in: 29% Quneitra 2% Tartous 95% As-Sweida 5% Dar’a 6% Damascus 89% Aleppo 7% Homs 7% Ar-Raqqa 78% Dar’a 14% Idlib 22% rural Damascus 70% Idlib 29% in Aleppo 66% Al-Hasakeh 62% Ar-Raqqa 61% Rural Damascus 59% Homs 51% Quneitra Source: International Organization for Migration, as of December 2016. Note: IDP = internally displaced person. Figure 2.7: Direction of Flight for the Forcibly Displaced, by Governorate FLED INSIDE SYRIA FLED FROM: FLED OUTSIDE SYRIA 1,792,581 Aleppo 1,632,838 958,782 Rural Damascus 500,367 718,420 Damascus 386,006 607,943 Homs 445,223 499,878 Hama 218,288 328,308 Deir-ez-Zor 189,987 201,016 Dar’a 174,805 142,688 Idlib 179,037 90,294 Al-Hasakeh 296,942 40,372 Lattakia 178,357 102,755 Ar-Raqqa 75,597 44,550 Quneitra 17,470 13,138 As-Sweida 13,684 2,271 Tartous 7,384 Source: International Organization for Migration, as of November 2016. 52 The Wrath of Conflict Table 2.7: Push/Pull Factor Matrix for Aleppo, Homs, and Dar'a (1: Primary Driver, 2: Secondary Driver, 3: Tertiary Driver) Theme Factor Description Aleppo Homs Dar’a Push Pull Push Pull Push Pull Political Security The perception that 1 1 1 1 1 1 Stability departure/ return is related to physical safety Forced Formal /informal 1 1 1 1 1 relocation authorities in a given area create conditions that displace populations Available Sustained access or no 1 1 1 1 1 1 shelter access to safe living space Accessibility Infrastructure The extent of infrastructure 2 2 2 2 2 damage, irrespective of service provision Access to The absence/presence of 2 2 2 2 2 services service provision, through public or private networks Food Access to affordable food 2 2 3 3 1 1 availability through either local markets or aid organizations Affiliation Economic The prospect of generating 3 3 3 2 3 opportunity income Kinship ties Ethno-sectarian, religious, 2 1 1 3 tribal, and/or family connections Partisanship Political beliefs that serve 2 1 1 2 as a reason for departure, discrimination or return Source: World Bank 2015. Note: 1 = primary driver; 2 = secondary driver; 3 = tertiary driver. THE TOLL OF WAR 53 Although some return migration has already taken place, it remains small compared with the total numbers for displacement. So far, 566,000 returns have been reported within the country. The greatest proportion of returns occurred in Aleppo (332,000), followed by Hama (61,000). Nearly all returnees are staying in their own homes upon return (525,000 or 93 percent of the total number of returnees). The second most common shelter type for returnees is abandoned buildings, with about 20,000 returnees living in this type of shelter. Of the returnees living with host families, 16,000 live in abandoned buildings. More than 4,400 returnees report that they are paying for accommodation in their location of origin, with another 1,000 returnees in camps and formal collective centers in their home areas. The return decision is a more complex process than departure. Displaced people often choose not to return, even with security and a political settlement in place, for a variety of reasons. In some cases, they cannot afford return travel. In other cases, especially after protracted displacement, IDPs prefer to stay in urbanized destinations for the livelihood opportunities they may offer. In still other cases, the memory of traumatic events in home areas precludes return; demographic realities have changed; shelter is unavailable or lost; or scarce resources or competing claims to land or livelihood assets make return untenable. Generally, across case studies of the criteria that regulate return decisions, the most influential factors are security, shelter, access to land and livelihood assets, restoration of basic services, and political as well as kinship affiliation considerations. Overall, the conflict has led to significant losses in Syrian human capital in addition to imposing unsurmountable pain and suffering on Syrian people. Preliminary calculations by Hamilton and Nguyen (2017) show that the combined effects of casualties, forced dispersion, and reduced investments in human capital formation could add up to a 30 percent permanent loss in the country’s human capital stock (compared with the 2010 stock).16 To put this figure in perspective, according to the Wealth of Nations data set at the World Bank, a typical lower-middle-income country such as Syria would have about 16 These calculations rely on the discounted present value of the lifetime earnings of the currently employed population cohort, assumed to include individuals ages 15–65 years. Earnings within a given age and gender cohort are in turn a function of years of schooling and years of work experience. In addition to wage employment, total human capital estimates also include the earnings of the self-employed. This approach ensures that the asset value of human capital is consistent with other asset values in the System of National Accounts. The preliminary estimates are conditional on future GDP growth, which is taken at a constant long-term growth rate of 2.5 percent, and reflects efficiency gains, after a brief period of post-conflict recovery starting from 2018. The effective discount rate for earnings is therefore the difference between an assumed discount rate and the assumed GDP growth rate. The effective discount rate used in the new World Bank estimates of human capital is 1.5%, which is consistent with a discount rate of 4 percent and real growth in per capita GDP of 2.5 percent per year. This methodology (also called the Joregensen-Fraumeni methodology) for valuing human capital includes not only the earnings of older workers (ages 25–65), but also the expected earnings of those ages 15–24, based on school enrollment rates by age and gender. The issue for the 15–24 cohort is that many people are still in school, which has two effects on their lifetime earnings: (i) their lifetime earnings accumulation is delayed, but (ii) their earnings once employed will be higher as a result of their greater knowledge and skills. Gross enrollment rates are therefore an important input to human capital valuation, as are employment rates (by age and gender) and survival rates (also by age and gender). More details and data sources are provided in Hamilton and Nguyen (2017). 54 The Wrath of Conflict 18 percent of its wealth in natural capital, 25 percent in produced capital, and 57 percent in human capital. Thus, human capital is the largest source of wealth in all regions and countries, with the exception of the large petroleum exporters, and significant losses in human capital are also significant losses in overall wealth. Unsurprisingly, labor earnings constitute at least 50-60 percent of GDP in virtually all countries. In what follows, the report will discuss the economic implications of losses in human capital as well as physical capital in more detail. Effects of the Conflict on Economic Activity The destruction of physical capital, casualties, forced migration, and breaking up of economic networks has had devastating consequences for Syria’s economic activity. After growing by 3.4 percent in 2010, Syria’s GDP was estimated by this report to have contracted by 61 percent between 2011 and 2015 and the economy was estimated to contract by an additional 2 percent in 2016 – a 63 percent decline compared to 2010 GDP.17 The oil GDP declined by 93 percent during the same period, while the non-oil economy contracted by 52 percent due to the severe destruction of infrastructure, reduced access to fuel and electricity, low business confidence, and disruption in trade. Hydrocarbon production plunged from 383,000 bpd in 2010 to 10,000 barrels per day (bpd)18 in 2015 and 2016, due to ISIS control of a considerable portion of the country’s oil production. The most severe contraction of the economy took place in 2012 and 2013, when the economy shrank by 29 and 32 percent, respectively, as economic activity plummeted and fighting intensified and spread across the country. In nominal terms, the total GDP, which in 2010 stood at US$60 billion, was estimated to be at US$20 billion and US$15 billion in 2015 and 2016, respectively. Syria’s energy sector has deteriorated since March 2011 with oil and natural gas production declining dramatically. Syria's energy sector has encountered several challenges as a result of the conflict and subsequent sanctions imposed by the United States and the European Union. The energy infrastructure—including oil and natural gas pipelines and electricity transmission networks—has been damaged, which has hindered the exploration, development, production, and transport of the country's energy resources. Prior to the conflict, Syria's oil and gas sector accounted for approximately one-fourth of government revenue. Since the onset of the conflict, crude oil production in government-controlled areas has fallen sharply (97 percent decline), from 386,000 bpd in 2010 to 10,000 bpd19 in 2015 and 2016, according to the United Nations Economic and Social Commission for Western Asia (UNESCWA) (figure 2.8). Total oil production, 17 The World Bank reference year for real GDP is 2010. 18 This refers to oil in areas under government control, but with the fields under rebel control, oil production is estimated at 40,000 bpd. 19 ISIS controls 80 percent of the fields; the government controls 8 percent of the fields; and Kurdish forces dominate the remaining 12 percent (Almohamad and Dittmann 2016). THE TOLL OF WAR 55 Figure 2.8: Economic Activity during the Years of the Conflict ECONOMIC GROWTH SINCE 2011 20 0 -20 Percentage -40 -60 -80 -100 2010 2011 2012 2013 2014 2015 2016 Overall GDP Growth Non-Oil GDP Growth Oil GDP Growth GOVERNMENT CONTROLLED PRODUCTION OF OIL 500 400 383 340 Thousands bpd 300 200 145 100 58 28 10 10 0 2010 2011 2012 2013 2014 2015 2016 Source: World Bank staff estimates; United Nations Economic and Social Commission for Western Asia. 56 The Wrath of Conflict including from fields under opposition and ISIS control, is estimated to be up to 40,000 bpd, but estimates vary widely (box 2.2). In May 2016, the European Union extended its embargo on Syria’s oil exports to Europe until June 2017, which accounted for about 95 percent of exports. Syria's official crude oil exports are assumed to have almost ceased, but evidence points toward significant illicit exports, especially from non-government-controlled areas. The capacity of its two state-owned refineries has fallen to half of pre-conflict capacity (Energy Information Agency), exacerbated by the imposed sanctions. According to the Energy Information Agency, the Islamic Republic of Iran continues to supply Syria with approximately 60,000 bpd of crude oil, but this level is still insufficient to meet demand. Until recently, the conflict has had a less severe impact on the natural gas sector than on oil, because some of the main gas fields and processing facilities have remained under government control. Syria’s gas production stabilized around four to five billion cubic meters in 2013–16, down from a record 8.7 billion cubic meters in 2011. However, there is evidence that gas infrastructure has suffered significant further damage since mid-2016. In the summer of 2016, ISIS captured and destroyed Al-Shaer, one of the Syrian government’s last two remaining large gas fields. In January 2017, the Al-Maher and Jihar gas field and processing plant were destroyed. The processing facility at Ebla, although currently idle, was struck by ISIS rockets in February 2017. Now only one of eight major gas processing plants remains in operation and in government control. Agricultural production also registered significant losses. The conflict has led to widespread damage to irrigation systems and shortages of labor and inputs such as seeds, fertilizers, and fuel. Syria had been a net producer of wheat in the prior 20 years (excluding drought years) but production of this crop is currently estimated to be at 1.5 million tons, approximately 55 percent lower than the pre-conflict average of 3.4 million tons (2007–11) (FAO and WFP 2016). However, some crops, such as barley,20 have seen an increase in production of 40 percent above the average of the past 10 years, to 0.9 million tons, following better rainfall. Syria’s livestock economy, which accounted for about 40 percent of total agricultural production, was also severely affected by the conflict. According to the Food and Agriculture Organization (FAO), currently there is 30 percent fewer cattle, 40 percent fewer sheep and goats, and 60 percent less poultry (FAO and WFP 2016). Overall, agriculture GDP contracted by 41 percent between 2011 and 2015, and this followed a 10 percent decline in 2010. Conflict-driven disruptions of irrigation systems and labor shortages have played important roles in the loss of agricultural output. The FAO Food Security Study reported that because of the damage to pumping stations (both for surface and groundwater irrigation), irrigation structures, and equipment, and the poor availability of electricity and fuel, many farmers have 20 A rainfed and more resistant crop under adverse weather and input conditions and not dependent on irrigation and fertilizers like wheat is. THE TOLL OF WAR 57 abandoned irrigation and turned to predominantly rainfed crops such as barley, coriander, and other herbs. A pilot study conducted for this report in four agricultural zones in Syria supports these observations (box 2.3). The availability of farm labor during the 2015/16 season varied greatly across governorates, based on security conditions and availability of alternative employment. According to respondents to the NAPC survey, in Hasakeh, the low availability of farm labor contributed to the collapse of cotton production, which is now reduced to 4,000 hectares. Box 2.2: Oil Production by Non-State Groups In areas outside government control, non-state groups have continued to extract and process oil and gas. Estimates from November 2014 suggest that ISIS was producing, refining, and selling as much as 50,000 barrels of crude oil per day. Another estimate from October 2015 puts the number at between 34,000 and 40,000 barrels a day. The coalitions led by the United States and the Russian Federation against the Islamic State (ISIS) consider oil production a “war-sustaining activity,” and have therefore targeted extraction, refining, and transport infrastructure across ISIS-held territory. By late September 2016, the United States–led coalition claimed to have struck 2,638 “oil infrastructure” targets, primarily in Syria, leaving much of the country’s extraction and processing capacity destroyed (though much had reportedly already been damaged by previous fighting). See map B2.2.1 and table B2.2.1. Map B2.2.1: Oil Economy during Conflict TURKEY Isis control Isis support Rebel-held Sinjar Syrian regime Mosul KURDISTAN REGIONAL GOVERNMENT Kurds Aleppo Rebel-held/Nusra* Raqqa *Jabhat al-Nusra, al-Qaeda’s Syrian affiliate Deir Ezzor Oil fields Mobile refinery Oil market Primary oil routes al-Qaim Smuggling route LEBANON SYRIA Damascus IRAQ Baghdad IRAN ISRAEL JORDAN SAUDI ARABIA 100KM Source: Erika Solomon, Robin Kwong and Steven Bernard, at Financial Times, February 29, 2016. Notes: Military map changed substantially since the time of this analysis: ISIS lost ground including the al-Jabseh field to the Syrian Democratic Forces. 58 The Wrath of Conflict Box 2.2 (cont.): Oil Production by Non-State Groups Table B2.2.1: Sample of ISIS Crude Oil Production Oilfield Est. production Price (barrels per day) ($/barrel) al-Tanak 11,000-12,000 $40 al-Omar 6,000-9,000 $45 al-Tabqa 1,500-1,800 $20 al-Kharata 1,000 $30 al-Shoula 650-800 $30 Deiro 600-1,000 $30 al-Taim 400-600 $40 al-Rashid 200-300 $25 Source: Erika Solomon, Robin Kwong and Steven Bernard, at Financial Times, February 29, 2016. By 2015, the Assad government retained control of only 8 percent of Syria’s oil resources (Ar Rasm, Bilas Sharifah, al-Sha’ir, Elba, and Jazal fields). The oil fields held by the Islamic State group in the summer of 2015 had a pre-conflict output of 250,000 barrels per day (bpd)—or roughly 65 percent of the country’s overall capacity. Syria’s Kurdish-led factions also control oil fields, most notably the Rmailan field in Hasakeh province. The field, which at its height produced up to 170,000 bpd, in late 2015 operated at about 15,000 bpd. Syria’s two state-owned refineries in Baniyas and Homs are also only partly operational: by early 2015, the nominal capacity was just below 240,000 bpd or 75 percent of the country’s pre-conflict demand. Syrian officials claimed that infrastructure damage has further decreased production capacity to about 50 percent. Under conflict conditions, civilians and armed groups have also set up private makeshift oil refining sites across the country. For a 2016 report entitled “Scorched Earth and Charred Lives,” the Dutch NGO PAX used satellite data to identify as many as 37 separate makeshift oil refining sites across Syria, each with hundreds or even thousands of very rudimentary individual structures to process oil that reportedly has been primarily purchased by armed groups in the areas. The central government reportedly continues to purchase oil from such areas. There has been considerable destruction of major industrial zones in Syria, most notably in Aleppo. Many Syrian cities had clearly demarcated industrial areas and, in some cases, industrial zones a short distance from the city. In Aleppo, which was the manufacturing center of Syria and a hub for sectors including pharmaceuticals, textiles and garments, chemicals, and THE TOLL OF WAR 59 agroprocessing, many of the industrial zones were the scene of substantial fighting and have suffered major damage. As table 2.8 shows, the worst hit zones in Aleppo have seen 70 percent or more of their buildings damaged, according to recent satellite imagery. Furthermore, this imagery does not capture lateral damage to the sides of buildings and looting. Manufacturers were forced to close their businesses or move them elsewhere. In 2013, the government published a directive to encourage factories to move to safe cities and simplified the paperwork required to do so. In 2014, they took a decision to establish two new industrial zones in Latakia, with sizes of 350,000 and 225,000 square meters, respectively. In interviews, private Syrian banks reported that lending for factory relocation and setting up in safe cities like Latakia and Tartus is a significant line of business. However, other manufacturers have moved their production elsewhere in the region, such as Turkey, Egypt, and Jordan. Egypt in particular has attracted the garment and textile business from Aleppo. But not all sectors have been equally able to move outside the country. Machines from heavier industries, and even some machines in textile production, were simply too cumbersome to move and had to be abandoned. The Syrian government initially reacted to businesses moving outside the country by banning them from selling their products back into Syria. However, the government is now attempting to attract back businesses that relocated overseas. Sanctions have also had a significant impact on the manufacturing sector. This impact has been most severe in sectors that export, require a substantial share of imported inputs, or were closely linked to external partners who have cut ties. For exporters, none of the interviewed private banks reported that they were providing any foreign currency services or letters of credit, due to the sanctions. State-owned banks are similarly affected and in many cases directly sanctioned for their links to the government. As a result, accessing import and export finance is one of the biggest challenges facing manufacturers, according to interviews with manufacturing conglomerates. Many of these conglomerates suffered physical damage to facilities as well, but today are more worried about the ongoing impact of sanctions. Trade financing is occurring largely through informal methods, but this undoubtedly limits its accessibility. Box 2.3: Effects of the Conflict on Syrian Agriculture: A Pilot Study A pilot study, conducted jointly by the World Bank and the European Space Agency EO4SD for this report, used Satellite Earth Observation to assess the impact of the conflict on agricultural production in Syria. The study mapped temporal and spatial changes in land cover and agricultural productivity, by using a vegetation index, which is a measure of the greenness of vegetation. Three areas and respective irrigation schemes in Syria, located near Dara’a, Aleppo, and Ar Raqqah, were compared with similar schemes in nearby zones in Turkey and Jordan. The status of these agricultural areas is assessed and compared between 2011 (pre-conflict) and 2016 at field level (10x10 meter to 30x30 meter resolution). (See maps B2.3.1 and B2.3.2.) 60 The Wrath of Conflict Box 2.3 (cont.): Effects of the Conflict on Syrian Agriculture: A Pilot Study Map B2.3.1: Study Areas Included in Map B2.3.2: Change in Vegetation the Demonstration Cover for February in March (Al Hassakeh) from Pre-Conflict in 2011 to Conflict in 2016 Source: World Bank and European Space Agency staff calculations. The main conclusion is that the conflict has affected agricultural output through land utilization, cropping intensity, and crop yields. Cultivated land has decreased by about 50 percent in the Syrian Al Eis irrigation scheme; formerly wet (potentially irrigated) areas in the dry season diminished by 36 percent; and productivity has declined by 36 percent in winter and 47 percent in summer. In Ar Raqqah, productivity also decreased in spring and summer, by on average 15 percent. In Al Hassakeh, productivity decreased by 15 percent in summer but increased by 60 percent in spring. The results include targeting criteria and an agricultural activity analysis, and can assist in planning for reestablishing agriculture input supply chains. Agricultural production capacity is influenced by three variables: land utilization (extent of cultivated areas), cropping intensity (number of crop cycles in a year), and crop yields (productivity). Earth observation can provide information to support all three variables. However, in an Earth observation context, the word productivity is a combined measure of all three effects, derived by a vegetation index, here the Normalized Difference Vegetation Index. The extent of cultivated areas significantly decreased in the Al Eis irrigation scheme (45,672 hectares) between 2011 and 2016. Although pre-conflict there were 34,327 hectares under cultivation, less than half of the area was still in use in 2016 (12,308 hectares). A detailed classification of the cultivated area in Al Eis shows that half of the El Ais area was fallow or abandoned in 2016, compared with the cultivated area extent in 2011. The detailed classification of the cultivated area in Al Eis shows that in 2016 the remaining active cultivations were mainly winter crops (11.5 percent), followed by spring crops (10.9 percent). Summer crops, which typically need irrigation, are only found in 3.9 percent of the Al Eis area. Two cropping cycles are rarely found in 2016. Earth observation showed that vegetation activity was not present or in general significantly lower in the Syrian irrigated area compared with an intact irrigation scheme in Turkey in 2016. Furthermore, the crop cycle and in particular the start of the season was very similar to the surrounding natural, rainfed vegetation in the Syrian Al Eis irrigated area, indicating a breakdown of the irrigation infrastructure. THE TOLL OF WAR 61 Table 2.8: Damage by industrial area in nine cities City Industrial area Damage (%) Aleppo Shaykh Najjar 81 Aleppo Ramouseh 73 Aleppo Ash-Shaqif 70 Aleppo Billeramoun 67 Deir ez-Zor Deir ez-Zor 54 Tadmur Tadmur 43 Kobani Kobani 39 Idlib As-Sina’iyya 34 Aleppo Jibreen 18 Dar’a Dar’a 8 Raqqa Raqqa 4 Douma Douma 3 Hama Hama 2 3 Hama Hama 1 1 Source: World Bank staff estimates. On the expenditure side, the conflict has led to a collapse in investments (figure 2.9). Public and private investment in Syria contracted greatly after the start of the crisis, due to the heightened level of risk and uncertainty and the severe damage to infrastructure, equipment, electricity, water systems, and communications. There was a 24 percent cut in public investment in 2011, and a contraction of over 31 percent in 2015 in real terms. Investors exited Syria due to the insecurity and poor business environment, which led the private investment’s share of GDP to decline from 12 percent in 2010 to 4 percent in 2015. Public investment’s share of GDP was also severely hindered by declining revenues and shifting priorities toward military consumption and public wages. Overall, the period from 2011 to 2015 witnessed a substantial fall in public investment, from 9 percent of GDP in 2010 to only 1 percent of GDP in 2015. International trade, especially exports, also declined severely (figure 2.9). Total exports fell from US$7.9 billion in 2011 to US$631 million in 2015, a decline of 92 percent in just four years. Imports declined from US$17,348 million in 2011 to US$4,654 million in 2015—a 73 percent fall. The steeper decline in exports than imports (and GDP) is potentially driven by three factors: tradable sectors, for example, oil, could be more exposed to changes in income than non-tradable sectors; sanctions on Syrian exports could be effectively blocking activity; and some export industries may have substituted away from formal exports to production for domestic consumption, while 62 The Wrath of Conflict Figure 2.9: Investment and Trade Dynamics PRIVATE AND PUBLIC INVESTMENT 30% 20% Percent of GDP 14% 12% 10% 9% 8% 8% 7% 6% 4% 4% 2% 2% 1% 0% 2010 2011 2012 2013 2014 2015 Private fixed capital formation Public fixed capital formation Total fixed capital formation EXPORTS AND IMPORTS IN SYRIA BETWEEN 2011 AND 2015, USD MILLIONS 17,348 7,946 7,633 6,468 5,945 4,654 2,147 1,469 1,093 631 2011 2012 2013 2014 2015 Exports Imports Source: World Bank staff estimates. THE TOLL OF WAR 63 others may have engaged, directly or indirectly, in illicit trade. The mineral sector’s exports suffered the most dramatic declines (from US$4.7 billion in 2011 to close to zero in 2015). On the import side, the decline in manufactured goods imports (from US$11.5 billion in 2011 to US$2.8 billion in 2015) explained two-thirds of the total reduction in imports. Foreign exchange reserves are almost depleted. With conflict and international trade sanctions, the current account deficit was estimated to reach 28 percent of GDP in 2016, up from 0.7 percent of GDP in 2010. This was due mostly to the collapse of oil exports, loss of tourism revenues, and sanctions. The government has been increasingly financing its growing external deficit through withdrawals from the Central Bank reserves. It is estimated that the amount of foreign reserves declined severely, from nearly US$21 billion in 2010 to less than US$1 billion in 2015. Public Finance The conflict has led to an escalating budget deficit, especially after 2012, when state revenues began to plummet. Lower revenues and high military spending sharply widened the fiscal deficit to 9 percent of GDP in 2016, from 2 percent of GDP in 2010 (UNESCWA). However, underlying fiscal developments were much more adverse than suggested by the changes in the budget deficit: total public revenue dropped from 23 percent of GDP in 2010 to less than 3 percent of GDP in 2015. This was mainly due to the losses in oil and tax revenues, the collapse of international trade due to sanctions, a growing informal economy, and weak administrative collection capacity. In response to this shortfall, government spending was cut back, but these measures were not enough to offset the fall in revenues (figure 2.10). The conflict has led to a shift from capital expenditure toward current expenditure. Capital expenditures collapsed from about 9 percent of GDP in 2010 to an estimated 1 percent of GDP in 2015. Current expenditure, as a percent of GDP, was much higher, at 18 percent in 2015, compared with 16 percent of GDP prior to the crisis, reflecting increased spending on public sector wages, military, and imports of essential goods. Wages made up the largest share of current expenditure, followed by subsidies and transfers, goods and services, and interest payments. Current expenditure on wages held steady and public debt increased to an unsustainable level. Although on the decline in recent years, current expenditures have continued to account for the largest share of the budget despite the collapse in revenue over the course of the conflict. Wages made up the largest share of current expenditure, followed by subsidies and transfers, goods and services, and interest payments. As a result, public debt has increased to unsustainable levels, driven by the collapse in revenue and increase in expenditure, particularly for military purposes. Public debt is estimated to have increased by 150 percent of GDP in 2015 (US$21 billion), up from only 30 percent of GDP in 2010 (US$18 billion) (figure 2.10). 64 The Wrath of Conflict Figure 2.10: Fiscal Deficit and Debt FISCAL BALANCES 60 40 Percent of GDP 20 0 -20 -40 2010 2011 2012 2013 2014 2015 2016 Total revenue Total expenditure Budget deficit GROSS PUBLIC DEBT 160 120 Percent of GDP 80 40 0 2010 2011 2012 2013 2014 2015 Source: United Nations Economic and Social Commission for Western Asia and International Monetary Fund estimates. THE TOLL OF WAR 65 Money and Banking The nominal exchange rate of the Syrian pound sharply depreciated. Prior to March 2011, the Syrian pound was stable and traded at LS 47 against the U.S. dollar. With depressed export revenues and declining international reserves, between 2010 and 2015, the official exchange rate of the Syrian pound depreciated by 459 percent and reached 260 (LS/US$). The Central Bank of Syria (Syria Central Bureau of Statistics) took several measures, including limits on import licenses, to curtail demand for foreign currency. Other measures included forcing exporters to surrender their foreign currency earnings in April 2015. However, the ability of the Syria Central Bureau of Statistics to intervene effectively has proven limited, considering the continuous depletion of foreign currency reserves. The Syrian currency continued its depreciation against the U.S. dollar during 2016, but at a much sharper pace. Foreign currency shortage and ongoing devaluation led to increasing dollarization in government-controlled and opposition areas. The conflict also led to a steep increase in prices. The consumer price index (CPI) in Syria has increased rapidly throughout the period of the crisis. According to UNESCWA, the prices for products such as rice, subsidized bread, and wheat flour rose dramatically, by 723 percent, 418 percent, and 388 percent, respectively, between March 2011 and November 2015. In October 2014, the government raised the subsidized price of diesel from 36 to 48 cents per liter. The price of heating oil increased from 73 to 85 cents per liter. Higher costs for rent and utilities put additional pressure on purchasing power. In 2016 only, CPI inflation was estimated to have reached to 58 percent, due to ongoing disruption of trade, large shortages in supply, and the sharp depreciation of the Syrian pound (figure 2.11). However, the actual inflation rate varies significantly across the country, reflecting the fragmentation of Syria’s economic space, with significantly higher price hikes in areas with more intense conflict. Although prices are less volatile in government-controlled areas (Latakia, Tartous, and Damascus), they are particularly rampant in areas with high security risks or besieged areas (such as Deir-Ez-Zor and Ar Raqqa), where the trade disruptions and security risks increased the value of goods associated with high transportation costs. Syria’s financial sector has been hit hard by a shrinking economy, tightening international sanctions, lack of public and private investments, and reduced business opportunities in a conflict environment. The banks have endured physical destruction of branches, but all 20 banks are still operational. Private sector banking and payments systems appear to be largely functional in the key government-held areas. There, deposit and withdrawal services have continued, with limited lending services. Automated teller machine (ATM) and branch networks have deteriorated, but are still functional in government-held areas, although card payment transactions are not available. Communication failures periodically affect inter-branch connectivity and ATM services. Moreover, the security of staff and branches has become a major 66 The Wrath of Conflict Figure 2.11: Exchange Rates and the Consumer Price Index OFFICIAL AND PARALLEL EXCHANGE RATES IN SYRIA 600 500 400 SYP/US$ 300 200 100 0 2010 2011 2012 2013 2014 2015 2016 Official exchange rate Parallel exchange rate INFLATION AS MEASURED BY CONSUMER PRICE INDEX 100% 80% Period average, y-o-y 60% 40% 20% 0% 2010 2011 2012 2013 2014 2015 2016 Source: International Monetary Fund and United Nations Economic and Social Commission for Western Asia estimates. THE TOLL OF WAR 67 concern for the banks. Branches have been subject to armed robberies, and the ones in the conflict regions have had to close.21 As early as December 2013, it was estimated that 30 percent of the 539 bank branches (243 private and 296 public bank branches) were destroyed or not operational (Badra 2015). In nongovernment areas, access to and functioning of basic banking and payments systems is much more limited. Private banks are unable to operate where the central bank does not have authority. As such, it appears that no private bank branches are open in rebel-held areas. However, several private banks noted that in the northeastern city of Al-Qamishly, which is partly held by Kurdish groups, branches of private Syrian banks remain operational, thanks to some form of agreement between Kurdish groups and the government. Outside of this, rebel, Kurdish, and ISIS-held areas have none of the formal banking services they previously enjoyed. The economy in these areas appears to be almost entirely cash based. In this challenging environment, banks have had difficulty finding lending opportunities within acceptable risk parameters. According to International Monetary Fund data,22 lending to the private sector constituted around 45 percent of bank lending in 2010. Thus, the collapse of the private sector could have substantial impact on bank income levels. World Bank23 data indicate that the net interest margin of the banking system declined from 2.7 to 1.8 percent from 2011 to 2013, while the share of noninterest income in total income rose from 44.5 to 84.1 percent. In the absence of credible banking sector data, particularly from the predominant public sector banks, it is very difficult to assess accurately the impact of the conflict on the banking sector. International sanctions have compounded the effect of the conflict on the banking and financial sector. Between 2011 and 2015, international sanctions cut off the Syrian Central Bank (SCB) and state banks (representing 75 percent of banking assets) from international debt markets and payment and settlement systems. Bank for International Settlements data show that the exposure of foreign banks to Syrian banks declined from around US$550 million to US$90 million between 2011 and 2015. Despite budgetary problems and the need to attract foreign direct investments, Syria failed to attract any private financial flows from abroad. The only official financial assistance was provided by the Iranian government in the form of credit lines, totaling US$1.9 billion in 2013, US$3 billion in 2014, and US$0.97 billion in 2015 (IMF 2016). Assuming that the Iranian credit was the only financial aid (with no assistance from the Russian Federation or China) the remaining budgetary deficit was probably funded by the SCB and commercial banks, thereby negatively affecting the banking system. 21 In the absence of credible financial soundness indicator data, the team is unable to provide a more robust assessment of the impact and damage to the financial infrastructure of the banking sector. 22 International Monetary Fund International Financial Statistics database. 23 World Bank Global Financial Development Database. 68 The Wrath of Conflict The nonbank financial sector was hit hard, but enterprises largely continued to operate. The Syrian stock market (DSE) is still operational, with a few investors, and has had no de-listings from the exchange. As of 2015 Q4, DSE had 23 listed companies with US$7 billion in total assets, almost all from the banking and insurance sector. As of 2015, six brokerage firms of 13, and four advising firms have discontinued their operations due to the downturn in DSE’s performance and lack of proper income opportunities (Badra 2015). The insurance sector suffered from increased claims. The conflict saw an increase in claims, which worsened the financial position of insurance companies, despite the rejection of many claims on the basis of noncoverage. The insurers have also had difficulty finding reinsurers, due to sanctions and high country risk, while rising reinsurance costs were partly passed onto clients. Moreover, lower incomes of the population and high costs of insurance services hit the demand for insurance products. This is also visible in the DSE data, which show that the assets of the listed insurance companies are down from US$269 million in 2011 Q1 to US$95 million in 2015 Q4. Despite the conflict and having to write off part of their loans, certain Micro Finance Institutions (MFIs) are continuing operations: the Aga Khan Foundation has continued microfinance operations in Syria, although scaled down to rudimentary levels. It also collaborated with local organizations on aid distribution, such as cash assistance through savings accounts. Ibdaa MFB resumed activities in Syria and is working on opening new branches. The United Nations Relief and Works Agency provided 30 percent of total microfinancing in Syria in 2012 and granted 9,334 loans worth US$2.6 million in 2015.24 By 2015, it had to write off 8,648 loans with an uncollected balance of US$318,145.25 Human Development Outcomes Economic deterioration hit employment. Since the onset of the conflict, jobs were destroyed at an estimated rate of approximately 538,000 per year on average26 between 2010 and 2015, adding 482,000 people to the unemployment pool every year. The working age population has also decreased by 2 percent annually, corresponding to an estimated decrease of 264,000 individuals net per year in that period. As a result, the labor force decreased by 1 percent, about 56,000 individuals net per year, annually (figure 2.12). Overall, the unemployment rate increased from 8.6 percent in 2010 to a disastrous 52.9 percent in 2015, a 44.3 percentage point change. Idleness and informality have become the norm. More than three in four Syrians of working age (77 percent, or nine million individuals) are not involved in any economic value generation: 2.9 million unemployed and 6.1 million 24 www.UNRWA.ORG. 25 www.UNRWA.ORG. 26 The difference between the number employed in 2015 and 2010 would determine the number of net jobs created (or destroyed in this case). This number is “net” as it includes jobs created and destroyed. THE TOLL OF WAR 69 inactive (figure 2.13). Unemployment among youth reached 78 percent in 2015 (UNESCWA 2016). The current inactivity in the working age population will have long-term consequences, including rapidly depreciating skills and human capital. In the short term, however, the consequences are dire: individuals, especially young men, are increasingly joining the military or getting involved in violence to survive (SCPR 2014). Those who are lucky enough to find a job turn to informal job opportunities (as a wage earner or self-employed) or become involved in the “war economy” as a coping mechanism in an effort to earn an income for their families (World Vision 2016). These jobs are mostly labor intensive, unskilled, or low skilled and poorly paid. Even those employed in relatively stable jobs or in the private sector have seen their living conditions decline steeply due to the rise in the prices of basic commodities and rapidly depreciating currency. Demographic mobility helped to adjust the distribution of economic activity. The conflict pushed people and economic activities away from conflict zones into relatively more stable and safer ones. With millions of IDPs moving from high-conflict intensity areas, like Deir Ezzor and Dara’a, to low-conflict intensity areas, like Tartous and Lattakia, the former experienced labor shortage and the latter labor surpluses. With saturation in the labor market and the associated depression in wages (FAO and WFP 2015), the relatively safe areas became attractive to businesses and investments. According to Dun & Bradstreet (2016), citing data from 2015 from the Syrian Investment Agency, the regions of Suwayda and Tartus emerged as the most dynamic investment areas in 2015, as they attracted almost 70 percent of all new investments for 2015. These two regions were not considered particularly attractive before the onset of the conflict in 2010; they attracted only 11 percent of overall investments. By contrast, households in Aleppo and Dara governorates have lost more jobs than any other parts of the country (UNESCWA 2016). The emigration of people abroad, due to the security situation, lack of job opportunities, and steep decline in the standard of living have led to an important skills shortage in Syria. Since the onset of the conflict, the government has scaled down its social security programs to cope with mounting fiscal pressures. With declining revenues, the government decided to reduce its expenditures, first by dramatically decreasing spending on subsidies. This led to a significant increase in the prices of basic food and fuel. In October 2013, manufacturers started paying market prices for fuel oil and “mazout” or gasoil, effectively ending a decades-old policy of subsidies that followed the arrival to power of the Baath party. Key staple products (rice and sugar), fuel oil, and “mazout” are still subsidized for households, but prices have risen significantly. Prices of “mazout” increased about nine-fold and that of fuel oil increased 10-fold from 2011 to 2015. As for rice and sugar, prices increased by 2.3-fold in the same period. In an effort to reduce expenses for households, those key staple foods are still subsidized under a coupon scheme; however, the loss of jobs, prevalence of violence, and reduction in subsidies have led to poverty descending rapidly on the Syrian people. 70 The Wrath of Conflict Figure 2.12: Labor Market Indicators: Pre-Conflict versus Conflict CHANGE IN THE SIZE OF THE LABOR FORCE, EMPLOYED AND UNEMPLOYED 1,500 Conflict period 500 400 1,000 300 500 200 Thousands Thousands 100 - - (500) (100) (200) (1000) (300) (1,500) (400) 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 Change in the number of unemployed Change in the number of employed Change in the size of the labour force (right-axis) Source: World Bank staff estimations. ANNUAL GROWTH AND PERCENTAGE POINT (PP) CHANGE IN KEY LABOR MARKET INDICATORS 44.3% Unemployment rate (pp change) 0.5% -2.2% Total population (annual growth) 2.7% -2.1% Working age population (annual growth) 2.6% 49.2% Unemployed (annual growth) 5.1% -12.8% Employed (annual growth) 2.0% Post-crisis (2010-2015) Pre-crisis (2005-2010) Source: World Bank staff calculations based on data from the Syrian Center for Policy Research. THE TOLL OF WAR 71 Figure 2.13: Structure of the Working-Age Population in Syria, 2015 Working age population 11.6million 62.9% Inactive Active 6.1million 52.8% 5.5million 47.2% Employed Unemployed 2.6million 47.1% 2.9million 52.9% Agriculture Industry Building & Construction 0.4million 16% 0.02million 1% 0.4million 16% Trade & Tourism Transportation & Other services Communications 0.5million 19% 0.2million 7% 1.1million 41% Source: World Bank staff estimations. Estimates suggest that about six in 10 Syrians live in extreme poverty today. Micro-data on per capita expenditure at the household level and the poverty line were not available for this exercise. Nevertheless, a simple extrapolation of the growth elasticity of extreme poverty (from 2004–07) suggests drastic increases in extreme poverty in Syria (figure 2.14). Assuming that the estimated growth elasticity remains constant at -2.911 throughout the crisis period and that inequality does not increase (scenario 1), the poverty rate in 2016 is estimated to be 66.5 percent. When we allow the income elasticity to fall as the poverty rate increases, it moderates the impact of falling income on poverty; the estimated poverty rate under scenario 2 is 54.5 percent. Finally, when we assume that elasticity is decreasing and the Gini index is increasing at the rate of 1 percent per year (scenario 3), the predicted extreme poverty rate in 2016 is 62.7 percent. Aid distribution statistics support the assessment of a rapid increase in poverty rates. The United Nations Office for the Coordination of Humanitarian Affairs estimates that about nine million people are in need of food assistance in Syria today (UNOCHA 2016). About seven million people, or one in three individuals, are food insecure, and an additional two million people are at risk of falling into food insecurity. The entire population of besieged locations, and over half of all IDPs, are considered food insecure. According to OCHA, 74 percent of the sub-districts in Syria reported members adopting, on average, more than five types of negative coping mechanisms, to access food. Some of the negative coping mechanisms reported were early marriage, begging, borrowing, child labor, and child recruitment in the conflict. These negative coping mechanisms are disproportionally affecting women and children (UNOCHA 2016). 72 The Wrath of Conflict Figure 2.14: Simulated Poverty Rate With and Without the Conflict, 2007–16 (%) 70 66.51 62.72 60 54.54 50 40 30 20 12.3 10 8.88 0 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 Counterfactual Scenario 1 Scenario 2 Scenario 3 Food security is the biggest item in aid. Since the onset of the conflict in early 2011, US$3 billion, representing 50 percent of earmarked official aid flows or 32 percent of all official aid flows to Syria have been assigned to the food security sector (figure 2.15). Of all sectors, food security is the biggest recipient of aid, followed by the health and the “emergency shelter and nonfood items” sectors (US$774 million and US$591 million, respectively). The amount of official aid labeled as going to food security might be underestimated, as aid flows marked as multi-sector and for which a sector has not yet been specified include food security. Aid earmarked for the food security sector as a share of yearly aid flows has been stable, at around 33 percent on average between 2012 and 2016. Over 60 percent of the abovementioned food assistance is channeled through the World Food Program (WFP). Foreign aid to Syria grew at an annual growth rate of 131 percent between 2011 and 2016, with the majority of the earmarked sources originating from the Government of the United States. Funds to Syria grew from US$39 million in 2011 to US$2,584 million in 2016 (figure 2.16). Of those funds, 23 percent originated from the Government of the United States, 21 percent of the Government of Germany, and 12 percent each from the European Commission and the Government of the United Kingdom, in 2016. Other donors included the Turkey Humanitarian Fund (accounting for 4 percent of total funds in 2016), Government of Kuwait (accounting for 5 percent in 2015 and 2016), Government of the United Arab Emirates (accounting for 9 percent THE TOLL OF WAR 73 in 2014), Government of the Kingdom of Saudi Arabia (accounting for 33 percent in 2013), as well as the Governments of Sweden and Switzerland (accounting for 10 percent each in 2011). Figure 2.15: Food and Non-Food Foreign Aid Flows in Syria TRENDS IN FOOD ASSISTANCE TO SYRIA, 2011-2016 1000 800 US$ million 600 400 200 0 2011 2012 2013 2014 2015 2016 World food programme Food security sector SECTORAL DISTRIBUTION OF OFFICIAL AID FLOWS, 2011-2016 (SHARE OF EARMARKED OFFICIAL AID FLOWS) Mine action Early recovery Protection Education Coordination and support services Water sanitation hygiene Multi-sector Emergency shelter and NFI Health Food security 0% 10% 20% 30% 40% 50% US$ million Source: United Nations Office for the Coordination of Humanitarian Affairs, FTS database. 74 The Wrath of Conflict As of December 2016, 5.8 million individuals received in-kind food assistance. Most beneficiaries are in Aleppo, Rural Damascus, and Idlib, while areas such as Der-ez Zor and Ar Raqqa received minor assistance relative to their need (51,000 and 6,000 beneficiaries, respectively). Overall, 1.6 million beneficiaries were in hard-to-reach and besieged areas. The national average cost of a standard food basket is LS 32,332 (US$64) (December 2016), which is 487.5 percent higher compared with 2012. The basket is most expansive in the governorates of Rural Damascus, Homs, and Damascus, and most affordable in As-Sweida, Dar’a, and Quneitra (map 2.11). WFP alone distributes more than four million food baskets per month. The food baskets distributed by WFP provide beneficiaries approximately 1,700 kilocalories per day and consist of the following commodities: wheat flour, bulgur, rice, red split lentils, whole lentils, white beans, oil, salt, and sugar, while their ready-to-eat rations provide beneficiaries approximately 2,100 kilocalories per day and consist of the following commodities: canned fava beans, canned humus, canned vegetables, canned chicken, olive oil, biscuits, halawa, and juice (orange/apple/pineapple or grape, five bottles, 200 milliliters each). The majority of those commodities is imported from Turkey (pulses, wheat flour, bulgur wheat, sugar, and oil); rice is imported from India or Vietnam. The only items bought locally are salt and ready-to-eat rations. Figure 2.16: Trends in Reported Funding, by Source, 20017–16 (US$, millions) 3,000 Thousands bpd 2,000 1,000 - 2011 2012 2013 2014 2015 2016 All other funding Government of UK Government of USA Government of Kuwait European Commission Government of Kingdom of Saudi Arabia Turkey Humaitarian Fund Government of Switzerland Government of UAE Government of Sweden Government of Germany Source: United Nations Office for the Coordination of Humanitarian Affairs, FTS database. THE TOLL OF WAR 75 Map 2.11: Number of Beneficiaries, by Sub-District (December 2016) Source: United Nations 2016. A small cash-based transfer program targets pregnant and lactating women. In 2014, WFP introduced cash-based monthly transfers for pregnant and lactating women. The program now targets 20,000 women in the Homs and Lattakia governorates. The voucher is considered complementary to the food basket program, and aims at improving the dietary diversity and micronutrient intake of mothers and mothers-to-be. The vouchers can be used to purchase fresh fruits and vegetables, fresh dairy products, and fresh animal protein (meat, chicken, eggs, and fish) from designated WFP contracted shops. The reference for the value of the voucher is a basic fresh food basket. The prices of these items are monitored monthly and the value of the vouchers is adjusted approximately every three months. 3 A Syrian man sells fuel in the northern embattled Syrian city of Aleppo on February 9, 2016. Shop shelves are bare and prices have doubled in Syria's Aleppo as the threat of a blockade looms after advancing regime forces cut off the city's second main supply route the previous week. Photo Credit: Karam Al-Masri/AFP/Getty Images. DISENTANGLING THE ECONOMIC CHAPTER IMPACT OF THE CONFLICT The economic impact of the conflict is manifested through multiple channels, and each channel has its own bite. The analysis in this chapter shows that physical destruction, by itself, explains only a limited share of the aggregate economic impact. The bulk of conflict-driven economic losses is generated not by destruction of productive factors, but by low utilization of these factors. Disruptions in economic organization are about 20 times costlier than capital destruction alone in the first six years of the Syrian conflict, when measured by the cumulated losses in gross domestic product. Reduced connectivity, higher transportation costs, disruptions in supply chains and networks, increased rent seeking, and erosion of social trust render production daunting in an active conflict environment. In comparison, conflict-related casualties generate the most catastrophic effects for the subjective well-being of Syrians. In addition, they are also the primary drivers of forced displacement, and the economic impacts generated by casualties are much more persistent than other channels. Finally, the analysis shows that the longer the conflict lasts, the slower the post-conflict recovery will be. 78 Disentangling the Economic Impact of the Conflict C H A P T E R 3 : DISENTANGLING THE ECONOMIC IMPACT OF THE CONFLICT What is the economic impact of the conflict on the Syrian economy? Answering this question turns out to be surprisingly difficult. The previous section showed that the conflict has been associated with changes in many areas that can affect the well-being of Syrian citizens. These include casualties, forced displacement, physical destruction, limitations on the mobility of goods and people, and a vast amount of foregone opportunities in the normal course of economic and social life. These multidimensional effects, however, need to be converted into impact assessments before they can fully inform policies like prioritization of mitigating actions and design of reconstruction and reconciliation plans. The conversion to impact assessments entails classifying the individual effects into similar economic categories, comparing them with reliable benchmarks (for example, counterfactuals), and accounting for the overlapping parts to ensure an accurate assessment. However, undertaking such an effort is complicated by several factors: • Data are sparse and, when available, vary in quality across different components. Active conflicts often lead to severe shortages of data. This study did not have the opportunity to collect first-hand data by implementing surveys and interviews in Syria. Thus, only remote-sensing-based assessments and secondary sources of data were used. • There is no perfect indicator capturing the full extent of the impact of the conflict. Aggregating various impacts is desirable to provide the “big picture” and avoid double counting, but requires a conversion of sorts, as different impacts are measured in different units. For example, the destruction of physical assets can be evaluated as rebuilding costs, but not the loss of human lives. The latter entails assigning a monetary value to human lives, which may be controversial, not to mention difficult. Certain aggregate indicators, such as the impact on income (or GDP), could help with some of these issues in the short term; however, they do not reflect the true reality and welfare costs encountered by people in conflict, and may even present an anomalous picture. The U.S. GDP, for example, grew by 11 percent annually during World War II, a record that remains today. • The absence of reliable counterfactuals could hamper the assessment of impact. In any given area of interest, assessing the impact of the conflict entails a good understanding of what would happen had there been no conflict. For “exogenous” events like natural disasters, these are relatively straightforward. The “business as usual” scenario could provide a relatively reliable counterfactual. In the case of a conflict, however, the conflict itself is an outcome of various factors, including income distribution, quality of institutions, and mobilization costs. Thus, it is not always clear what it means to have no conflict when underlying conditions dictate it. The regional THE TOLL OF WAR 79 economies that could provide a comparison point are affected by the factors that triggered the conflict and the conflict itself. Therefore, it is difficult to estimate what Syria would look like now in the absence of conflict. • The conflict is an active conflict, and its effects will continue to materialize long after it ends. The effects of casualties, destruction of physical assets, and, perhaps most importantly, degradation of institutions will have long-lasting impacts on the well-being of the surviving Syrians. A large portion of these effects has not yet been captured by current estimates of indexes like the poverty rate or income; thus, a forward-looking analysis is in fact a crucial aspect of impact assessment. In the absence of alternatives, second-best solutions are utilized. As described in the previous chapter, the data problem was tackled by a pragmatic approach that uses verified remote-sensing information and data from secondary sources. First, without a reliable counterfactual, a statistical method was utilized to generate imperfect counterfactuals for aggregate economic variables like GDP. The GDP series that are calculated by statistical methods based on pre-conflict data series are compared with the national accounts–based GDP estimates, to produce an assessment of GDP impact since the beginning of the conflict. Second, an integrative framework that is based on a dynamic general equilibrium model was introduced to deliver different indicators of impact in a consistent manner. This framework is used to run several inquiries that are not directly observable from data. These include assessing the relative importance of difference channels (physical destruction, casualties, and economic disorganization) in the manifestation of the conflict’s impact. In addition, the role played by demographic mobility is also discussed, by using quantitative implications. Finally, the propagation of the conflict’s post-conflict impact is analyzed, by using various conflict end-time scenarios. Aggregate Impact A first step in assessing the impact of the conflict on Syrians’ well-being is to generate an estimate for the aggregate impact. The effects of the conflict in different fields may have overlapping components that have to be counted only once, and not twice. For instance, the delivery of water often uses electricity as an intermediate input; the impact assessments should add up the changes in value added in these sectors, not total nominal values of outputs. GDP calculations help to do this. But, like all indicators, GDP-based calculations fail to capture some other dimensions, which will be discussed. Impact on GDP Overall, economic outcomes deteriorated significantly during the conflict; but that is not the whole story. The previous chapter summarized the 80 Disentangling the Economic Impact of the Conflict estimates of economic performance in Syria since the beginning of the conflict. Simply comparing these estimates with those prior to the conflict, however, does not provide the “impact.” To estimate such impact, the actual economic outcomes need to be compared with where Syria would have been today had there been no conflict (figure 3.1). Unfortunately, it is impossible to know that with precision. With significant changes in external conditions, including the drastic change in oil prices over the past few years and the outbreak of the Arab Spring, it is difficult to accurately estimate how the Syrian economy would have evolved. This study employs various statistical approaches to estimate a counterfactual path for the Syrian GDP. In the absence of a reliable counterfactual, the options are limited. Simple benchmarking is not possible, because the set of comparable countries are influenced by the conflict itself, or they have not been fully exposed to the changes Syria would have experienced even in the absence of the conflict. This study employs synthetic control methods (SCM), which control for several factors (institutions, human capital, physical capital stock, population, and so forth) that cannot be controlled by simple benchmarking to produce counterfactual estimates for Syrian GDP (box 3.1). In addition, simple benchmarking and vector autoregression–based estimates are also computed for further checks. Figure 3.1: Conceptualization of Pre-Shock, Post-Shock, and Counterfactual Relationship Counterfactual Impact if stock variable Pre-shock Impact if flow variable Post-shock 2010 2016 THE TOLL OF WAR 81 The estimates under various techniques show similar outcomes: Syrian GDP was forecast to continue growing without the conflict, albeit at a slower rate than during the pre-conflict era. Figure 3.2 depicts two synthetically constructed Syrian real GDP per capita estimates, which use different weights obtained from the two different donor pools using SCM. First, it is important to note that the in-sample estimates provided a good match with the actual series before the conflict. SCM1 is preferred to SCM2, as the former better reflects the impact of declining oil prices in growth in 2015 and 2016, as the donor pool is restricted to countries with known similar structural features (including oil exporters). Overall, however, both techniques estimate a GDP per capita of around about $6,000 (in 2010 dollars) in 2016, compared with about $5,700 in 2010. Figure 3.2: Comparing the Counterfactual Real GDP per Capita by Different Methodologies with Actual Estimates US$ Syria US$ 8,000 (Real GDP per capita $) 8,000 7,000 7,000 6,000 6,000 5,000 5,000 4,000 4,000 3,000 3,000 2,000 2,000 1,000 1,000 0 0 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 Syria Counterfactual Counterfactual Counterfactual (SCM1) (SCM2) (VAR Model) Box 3.1: Counterfactual GDP Calculations The Synthetic Control Methods (SCM) was developed by Abadie and Gardeazabal (2003) and Abadie et al. (2010). SCM allows the creation of a control group without needing to find one country or case study. The approach also avoids the arbitrariness of the selection of the control group, by relying on a combination of comparator countries that best approximate the characteristics of the country that is exposed to the shock. A combination of comparator countries is likely to produce a better control/comparison group for the country exposed to the shock than any one comparison unit alone. 82 Disentangling the Economic Impact of the Conflict Box 3.1 (cont.): Counterfactual GDP Calculations By checking the pre-shock differences between the outcome variable for Syria and the synthetic Syria, we are able to judge the quality of the control group. We can then create the counterfactual outcomes for Syria, by controlling for the effect of the conflict through identification of the comparator countries that best predict the historical development of Syria prior to the conflict, by using a set of predictors. This produces a set of weights for each country (most end up being 0, as they are not relevant) and these weights are used to construct a synthetic Syria series based on the actual outcomes for the comparator countries during the period of the conflict, to produce a counterfactual Syria. Finally, this approach allows us to conduct robustness checks by way of p-values, by conducting the same approach on every other country in the donor pool to produce a distribution of effects, and seeing the significance of Syria’s post-treatment difference compared with all other countries. This allows us to build a confidence level for our inference that confirms that the variation in the data is due to the impact of the treatment (conflict). In this exercise, this will be frivolous, as the impact of the conflict is so severe and pervasive that there will necessarily be a very large variation between the counterfactual outcomes and the actual outcomes. Specification and Data Following the endogenous growth theory literature, the vector of predictors that was selected to explain economic growth included the stock of physical capital, stock of human capital, democracy proxy, trade openness, and population growth. This is a common specification for cross-country growth simulations and was most recently utilized by the International Monetary Fund (IMF) in its SCM estimation of reform effects (Adhikari et al. 2016). The data rely on a cross-country panel data set for the period between 1990 and 2011. The outcome of interest is real gross domestic product per capita, but we also employ the technique for population growth. The majority of the data was sourced from the Penn World Tables until 2014, then data from the IMF World Economic Outlook data were used to extend the data until 2016. These variables capture, in a broad sense, the impact of institutions, demography, and macroeconomic conditions in addition to traditional growth accounting variables, such as the stock of physical and human capital. Before conducting the SCM to construct Syria’s macroeconomic counterfactual scenario, we had to ensure that none of the comparators that were selected violated the exogeneity criterion. Explicitly, we needed to avoid countries that were involved or significantly affected by the conflict. To take this a step further, there is also an argument to restrict the donor pool to countries that have structural similarities to Syria. For example, if the SCM is conducted and finds that Tuvalu, a small island country in the Pacific Ocean with 10,000 people, shared a similar growth path to Syria historically, there may be sufficient grounds to imagine this was a spurious correlation. Accordingly, we conduct the SCM using two donor pools of countries. The first donor pool is a restricted group based on the countries identified as comparator countries. The second donor pool is all countries in the world excluding those that were involved in or significantly affected by the conflict. THE TOLL OF WAR 83 Figure 3.3: Actual and Counterfactual GDPs and Cumulative GDP Losses Real GDP (US$ 2010) 80 250 70 200 60 US$ billions US$ billions 50 150 40 100 30 20 50 10 0 0 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 Counterfactual Real GDP Cumulative difference (SCM1) (US$ 2010) (Real GDP US$ 2010) (RHS) Source: World Bank staff estimates. What is the impact of the conflict on Syrian GDP? The impact is calculated by using the SCM1 counterfactual (which is broadly similar to that of SCM2) and the estimates for the actual GDP per capita of Syria presented in the previous chapter. Actual GDP fell US$51 billion (in 2010 prices) short of the counterfactual GDP in 2016. From 2011 to 2016, the accumulated impact on real GDP is estimated to be US$226 billion (in 2010 prices, Figure 3.3), nearly two times the GDP in 2010. This is mostly driven by the steep drop-off in actual GDP estimates, leading to a 63 percent reduction in the size of the GDP compared with 2010. Once the conflict ends and Syria begins reconstruction, based on previous experiences in post-conflict economies, actual GDP will quickly rebound and start closing the gap with the counterfactual. Nevertheless, given the intensity and protracted nature of the conflict, the losses will have permanent effects on the well-being of many generations. Beyond GDP GDP is an imperfect measure of well-being even in peaceful times, and more so during conflict. By denoting the value of aggregate final goods and services produced in a year, including the production of military assets that are “consumed” by the public, GDP, by itself, does not fully capture the complex effects of the conflict on Syrians’ well-being. For instance, an important aspect of the conflict’s impact is on the life expectancy of Syrians, which is estimated to have fallen from 74.4 years in 2010 to 69.5 years in 2016 (World Health Organization), both measured at birth. This five-year drop in life expectancy 84 Disentangling the Economic Impact of the Conflict has set Syria back about three decades (the last time the Syrian life expectancy at birth was recorded at 69.5 years was in 1988). Even when the welfare of an average Syrian citizen is simply measured by her income and longevity only, it is obvious that focusing only on income changes over several years is insufficient (figure 3.4). Yet, because income and life expectancy are measured in different units (dollars and life-years, respectively), measuring the deterioration in welfare requires some sort of a conversion between the two. Figure 3.4: Indifference Map for Income and Life Expectancy in Syria 80 75 2010 70 2016 Life expectancy (years) 65 60 55 50 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 5500 Income (US$) Source: Onder, Ponthiere, and Pestieau (2017). A simple calculation shows that the loss due to reduced life expectancy could be as high as 45 percent of pre-conflict income. Onder, Ponthiere, and Pestieau (2017) use an “equivalent income approach” to quantify the domestic welfare loss associated with decreasing life expectancy in a simple way. This involves comparing the lifetime welfares of an average Syrian citizen (welfare is defined as based on income and longevity) before the conflict and at the sixth year of the conflict. Since longevity and income are lower in the sixth year of the conflict than before it, welfare is also lower, unambiguously THE TOLL OF WAR 85 (figure 3.4). The next step is to decompose the welfare effects determined by income loss and longevity loss. To do this, the income in the sixth year of the conflict is increased hypothetically until it reaches a level (equivalent income) where the person reaches her pre-conflict welfare, while holding the short longevity fixed. The results show that this hypothetical income level could exceed the actual pre-conflict income by 45 percent. With actual income decreasing by about half because of the conflict in this simple exercise, the welfare loss because of decreasing longevity is significant.27 The next section extends this analysis by introducing an extended set of impact indicators. To provide a more comprehensive assessment of impact, which takes into consideration the medium and long-term effects of the conflict, the analysis next turns to combining the damage and loss assessments developed so far and producing indicators of impact that capture different dimensions in a consistent manner. Integrative Framework This section of the report focuses on the prominent underlying factors that shaped the impact of the conflict on Syrians. To assess the unmeasured dimensions of the impact of the conflict, this section will employ a simulation model (Artuc and Onder 2017). Figure 3.5 illustrates the main operating principles of the simulation model; box 3.2 provides a summary of the model’s characteristics, and the appendix to this report contains the formal definitions. The first objective is to assess the relative importance of each channel through which the conflict might have manifested its impact. The analysis focuses on the following three major conflict-related mechanisms to explain the overall economic impact. Table 3.1 shows the baseline and alternative specifications of shocks and other factors, followed by the analytical approach. • Capital destruction. This encompasses the damage to all economically relevant physical capital, including infrastructure, machinery, and buildings. The extent of destruction is determined by using the physical damage assessments in the housing sector across governorates. • Casualties. This focuses on directly conflict-related casualties, which rely on the midpoint of the casualty estimates provided in the previous chapter. Aside from the obvious distress and welfare impacts on families, the economic impact of casualties includes reduction in productive labor and forcible displacement of Syrians. 27 The roles played by income and longevity are relatively robust to alternative parametrizations of the Constant Relative Risk Aversion period utility function used in this exercise. However, the role of longevity varies between alternative measurement methods when losses in social welfare are calculated. An ex ante approach measures social welfare by taking a uniform life expectancy at birth for all Syrians; thus, reduction in life expectancy affects the lifetime welfare of all individuals in the same manner. An ex post approach, however, takes into account the variation in actual life spans among individuals. The ex post approach penalizes early mortalities, especially when the social welfare function reflects aversion to inequality, and thus, the estimated welfare impact of the conflict is higher in this case. 86 Disentangling the Economic Impact of the Conflict • Economic disorganization. This comprises all factors that lead to reduced productivity or idleness in factors of production, including, but not limited to, losses in network access and connectivity (which may reduce access to input and marketing opportunities), reduced participation in economic activity arising from security concerns, and any other intangible capital that may have been destroyed. Methodologically, this component provides a residual category in calibrations: given capital destruction and casualties, the level of economic disorganization was chosen to target observable outcomes (changes in GDP and interregional mobility). Next, the analysis focuses on the role of mobility in shaping the impact of the conflict on Syrians’ well-being. The demographic analysis in the previous section showed that one of the most striking consequences of the conflict has been the large-scale demographic displacement of Syrians. On the one hand, the exodus of people from conflict-prone regions may improve welfare by reducing casualties. On the other hand, demographic mobility could further reduce economic activity, as displaced people do not immediately integrate into the labor markets at destination. To assess the relative importance of these effects, the analysis will compare the baseline results with alternative scenarios where moving out of the country or all mobility (internal and external) is restricted. The impact of the conflict will propagate past the immediate end of the conflict, whenever it happens, as shown by the use of different conflict end-date scenarios. The recovery in investment and capital stock may take place relatively rapidly, but the recovery of human capital and economic organization will take years, if not decades. The persistence of these effects is particularly significant when conflict is protracted. The longer is the conflict, the deeper are the effects. Thus, to put the baseline impact estimates into perspective, the analysis will consider alternative end-dates to the conflict. The baseline scenario assumed that the conflict ends within the current year (2017). The next scenario will extend this date to 2020. The final scenario will assume that the conflict does not end in the near future, and this will provide a worst-case benchmark. The analysis assumes that, when the conflict continues, the additional years of conflict have the same damage ratio and geographical composition as previous years. Although this assumption is not perfect, it is less arbitrary than other potential assumptions for the future path the conflict may take. In all cases, various indicators are used to compare the effects of the underlying factors. The analysis compares different scenarios on the basis of their outcomes in various indicators, including flow variables in annual terms like GDP and wages. In addition, these effects are accumulated over various periods (for example, from the beginning to the end of the conflict as well as benchmark years after the end of the conflict). To put these numbers in perspective, we compare them with the corresponding annual pre-conflict values. Finally, the present values of future wages are also reported to provide a more continuous observation of the effects on material well-being. THE TOLL OF WAR 87 Figure 3.5: Simulation Diagram BEFORE CONFLICT OTHER REGIONS CAPITAL NETWORKS LABOR Depreciates at a normal rate, Link active capital and active In each period, decides to move replenished by investment workers at a normal rate (allows to another region/sector or not equilibrium unemployment) in a “smart” way (by taking all future outcomes into consideration) AFTER CONFLICT OTHER REGIONS CAPITAL NETWORKS LABOR Depreciates at a high rate Link active capital and active Given higher probability of death (calibrated by using workers at a lower rate and low livelihood opportunities, remote-sensing damage (calibrated by targeting actual emigrates at a faster rate assessments), replenishment by GDP and migration patterns) investment slows down 88 Disentangling the Economic Impact of the Conflict Box 3.2: Simulation Model The simulation model builds on the model in Alix-Garcia, Artuc, and Onder (2017) and extends it in various aspects, including the characterization of a conflict environment and a built-in matching mechanism. The model comprises 14 regions (governorates) in Syria and a stylized outside region to account for outmigration. Agents, who are identical otherwise, choose a region in which to live and work every period. Moving to a different region entails a one-time cost, which has a fixed and a random component. Agents An agent can be employed (active) or unemployed (inactive). Active agents are paid wages that are determined by market conditions, while inactive workers get a (small) fixed income, for example, a transfer. When a worker moves to a new region, she is initially inactive. She searches for a job and is matched to a vacancy in the region with some probability, which increases when the size of the pool of searchers (new arrivals, new separations, and unmatched workers from previous periods) is small. The number of matches is determined by a Cobb-Douglas matching technology. For simplicity, the separation probability is assumed to be equal to the capital depreciation rate. The matched worker-employer pair constitutes a factor of production, that is, a job. This factor produces an output together with the capital factor via a Cobb-Douglas production function, and it receives its marginal product. The employer and employees share this marginal product equally. The welfare of agents is characterized by a Constant Relative Risk Aversion (CRRA) utility function with various degrees of risk aversion. An important implication of this is that utility is assumed to be zero in the case of a death, which in principle may be disputed. A different assumption, for example, a negative utility in the case of death, would have quantitative effects on the computations of welfare (Onder, Ponthiere, and Pestieau (2017) employs a range of negative utilities from death in calculating equivalent incomes). The probability of death is specific to regions, and determined by the intensity of the conflict. For simplicity, we assume that the agents who migrate outside Syria (refugees) receive a fixed, exogenously determined utility in every period with certainty. Capital The potential supply of capital is defined over a fixed continuum in each region, for example, land. Similar to workers, there are two types of capital: active and inactive. In each period, some of the active capital depreciates and becomes inactive. With the conflict, this depreciation rate increases and varies across different regions. After becoming inactive, capital can be activated by investment, which requires an investment cost with fixed and random components. If the investment cost is larger than the value difference of active and inactive capital, the capital stays inactive. If not, the capital becomes active after paying the cost. This decision is repeated for every unit of inactive capital every period. Capital receives a rent equal to the marginal product. The value of capital is calculated using the expected rents, probability of depreciation, and distribution of investment shocks. (It is similar to the case of the problem for workers, but simpler.) THE TOLL OF WAR 89 Box 3.2 (cont.): Simulation Model Shocks Shocks are stochastic, for example, they are unpredicted by agents. With economic disorganization shock, productivity decreases and the separation probability increases. Capital shock increases the capital depreciation rate significantly, which is also reflected in investment decisions, as the depreciation rate changes the net returns to investment as well. A casualty shock not only leads to deaths in the current period, but also increases the probability of death in future periods; thus, when there is a casualty shock at some location, people perceive this as a security threat in the coming periods. Calibration The model’s parameters are calibrated with widely used standard approaches from the economics literature (risk aversion parameter in CRRA utility 0.9, annual capital depreciation in non-conflict years of 2.5 percent, and so forth). The shocks are calibrated to match the physical damages, cross-governorate and cross-border migration, and gross domestic product outcomes by the end of the six-year conflict, as summarized in the previous section. We did not aim to match the year-by-year changes within that six-year period. This framework does not capture many other effects. The simulations that are presented here do not capture several important outcomes, including environmental, cultural, political, and security-related effects that are posed by the conflict. These left-out effects are nevertheless crucial determinants of Syrians’ well-being. In addition, the results are presented in comparison with pre-conflict levels, leaving aside the discussion on counterfactuals. Table 3.1: Structure of the Simulation-Based Analysis Baseline Channels of impact Mobility scenarios End scenarios SHOCKS (1) (2) (3) (4) (5) (6) (7) (8) (9) Capital destruction Casualty Economic disorganization MOBILITY Income-based mobility YES Welfare-based mobility YES YES YES YES YES YES YES YES Internal mobility allowed YES YES YES YES NO YES YES YES YES External mobility allowed YES YES YES YES NO NO YES YES YES END OF CONFLICT Conflict end year 2017 2017 2017 2017 2017 2017 2017 2021 n/a 90 Disentangling the Economic Impact of the Conflict Channels of Impact The economic impact of the conflict is manifested through multiple channels. The simulation results show that the negative effects of the conflict are not only generated by destruction of capital and human lives, but also by behavioral responses, like reduction in investment in the baseline scenario. With a combined shock, investments drop to about 80 percent lower than their pre-conflict levels and remain there throughout the duration of the conflict (table 3.1). This reduction is primarily driven by low expected returns to investment: a decrease in overall productivity, a high destruction rate of physical capital, and a decrease in available labor (as more than 3.5 million Syrians leave the country in gross terms over the course of the conflict) bring down the payoff to investing. As a result, by the end of the conflict (which is the end of the sixth year in the baseline), GDP remains at 65 percent lower than its pre-conflict level. To put this in perspective, this reduction means a full erosion of the economic achievements of the decade preceding the conflict. Capital destruction, by itself, would have a relatively subdued effect. To assess the implications of physical capital destruction only, the simulations turn to scenario 2 (capital destruction only), as specified in Table 3.1. This includes using the rate of destruction observed in physical damage assessments as of February 2017 to calibrate the capital destruction shock. The results show that if the conflict only destroyed capital (at the same rate it has done so far), without other effects, its impacts on income and welfare would be relatively limited. To see why, note that when only capital is destroyed, investments decrease by about 22 percent from the initial level relatively quickly and remain there throughout the years of the conflict. In comparison, they decrease by 80 percent when all shocks (capital destruction, casualties, and economic disorganization) are introduced in the baseline scenario.28 In the absence of other effects on labor and productivity, an increase in capital depreciation/destruction does not reduce profitability as much. This mechanism highlights a remarkable difference between natural disasters and conflicts. When a natural disaster destroys built capital in a well-functioning market economy with strong institutions, the recovery of capital stock is rapid, and impacts are not persistent. Conflicts are different: casualties, demographic movements, and lower returns to investment compound the damage to physical capital. As a result, the effective losses due to physical capital damages are more pronounced, and they can linger into the future at a much higher rate than in natural disasters. 28 To see why investments can decrease in the case of the capital-destruction-only scenario, there are two distinct mechanisms that affect the incentives to invest in our model. First, given a fixed continuum of capital in different cities (for example, land), which prevents an excessive agglomeration after a shock, there is an upper bound to investment, which becomes more binding as demographic mobility concentrates population in conflict-free areas. Second, in a given location, the investment decision is affected by two factors: the return to capital in a given period and the probability of further physical damages in future periods. Thus, even when capital destruction increases the marginal productivity of capital, and thus return to investment in a given period, a sustained attrition in physical capital may still deter investments. THE TOLL OF WAR 91 Stability and a conducive business environment are crucial for investments to play a major role during recovery. The result above shows that destruction of capital hurts most when it is combined with factors restricting the expected returns to investment. These include human capital loss, demographic dispersion, institutional degradation, and escalation in rent-seeking behavior. All these are prominent in the current conflict. In the absence of such complications, investment could jump back to pre-conflict levels quickly, and help the economy recover rapidly along with recovery and reconstruction efforts. Simulations show that investments would be back to their pre-conflict level four years after the end of the conflict in a capital-destruction-only scenario. Historical experience in countries like post-war Germany and Japan, where conflict did not recur, also support this view. For instance, Davis and Weinstein (2002) show that the Japanese cities that were bombed during World War II reached their pre-conflict relative sizes vis-à-vis other cities in about 20 years after the end of the war. In comparison, post-conflict recovery is much slower in countries with a high risk of resurgence in violence. Mueller, Piomentese, and Tapsoba (2017) show that this is because, in peaceful years, post-conflict countries with a high likelihood of conflict recurrence grow no faster than countries with no conflict history and substantially slower during the resurgence of violence. Casualties are the primary drivers of outmigration. When the conflict brings about casualties only (scenario 3 in table 3.1), the impact on GDP is found to be comparable to that of capital destruction. In both cases, the current GDP effects remain less than 5 percent, compared with the pre-conflict levels during the years of conflict. Behind this similarity, however, lies a major difference between the two. Conflict-related casualties lead to a massive exodus of Syrians, whereas capital destruction does not. Simulations show that about 2.5 million out-migrants of 3.5 million in the baseline can be attributed to casualty-related factors alone. This reduces GDP, but not at a drastic level, because the decrease in investment remains relatively modest in this case. Investment decreases by about 22 percent in the capital-destruction-only scenario and only about 14 percent in the casualty-only scenario. The impact of casualties is much more persistent than that of the other shocks. The accumulated losses in GDP reach 93 percent of the annual pre-conflict GDP by the 20th year after the beginning of the conflict (table 3.2 and figure 3.6). Only a sixth of this effect takes place during the years of the conflict. In comparison, the accumulated loss is 26 percent of GDP in the capital destruction scenario. About two-thirds of this effect is manifested during the conflict. 92 Disentangling the Economic Impact of the Conflict Table 3.2: Combined and Isolated Impacts of Shocks (Outmigration in millions; other variables in comparison to their pre-conflict values in percentage terms) Conflict Post-conflict 1st Year 2nd Year 3rd Year 4th Year 5th Year 6th Year 10th Year 15th Year 20th Year BASELINE (COMBINED SHOCKS) Investment -83.18 -81.77 -80.61 -79.58 -78.66 -77.80 -13.77 -7.31 -6.50 Outmigration 0.81 0.70 0.63 0.58 0.53 0.49 0.00 0.00 0.00 Wages -54.99 -54.81 -54.70 -54.64 -54.60 -54.57 -23.53 10.31 8.55 GDP -58.34 -60.60 -62.20 -63.42 -64.38 -65.15 -42.38 -13.16 -11.12 CAPITAL DESTRUCTION ONLY Investment -23.07 -22.85 -22.73 -22.65 -22.60 -22.57 0.01 -0.03 -0.04 Outmigration 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Wages -1.54 -2.31 -2.70 -2.92 -3.04 -3.10 -0.13 0.02 0.04 GDP -1.62 -2.44 -2.89 -3.16 -3.34 -3.46 -0.59 -0.31 -0.23 CASUALTY ONLY Investment -13.25 -13.41 -13.71 -14.04 -14.38 -14.71 -5.76 -5.57 -5.18 Outmigration 0.71 0.51 0.40 0.34 0.29 0.26 0.00 0.00 0.00 Wages 0.84 1.66 2.41 3.08 3.69 4.26 6.55 6.57 6.11 GDP -1.88 -2.33 -2.64 -3.00 -3.42 -3.89 -5.13 -5.82 -5.87 DISORGANIZATION ONLY Investment -75.46 -75.04 -74.70 -74.41 -74.15 -73.91 -9.09 -2.62 -2.21 Outmigration 0.13 0.14 0.15 0.15 0.16 0.16 0.00 0.00 0.00 Wages -54.52 -53.89 -53.34 -52.86 -52.42 -52.02 -27.22 4.09 2.97 GDP -55.97 -57.08 -57.95 -58.66 -59.26 -59.76 -37.05 -6.63 -4.98 Note: Outmigration is reported in millions and in gross terms. wages increase. Focusing exclusively on a single indicator, for example, current wages, could imply that conflict is good for Syrians. However, this is obviously misleading, because it does not take into consideration the well-being of the deceased. Although the survivors may experience an increase in GDP, the Syrian population incurs significant welfare losses in aggregate. As the effects of the conflict extend well beyond casualties, the casualty-only scenario is purely illustrative of the impact that the casualty channel has on key economic variables. The bulk of the impact of the conflict is associated with disruptions in economic organization. Using capital damage and the casualty number estimates from the damage assessments, our simulations show that the bulk of the conflict’s economic impact must be associated not with the net decrease in productive factors of production, but with how the remaining productive factors are put to use. The conflict not only destroys productive THE TOLL OF WAR 93 Figure 3.6: Dynamics of Impact under Various Shock Scenarios CUMULATIVE CHANGE IN GDP End of conflict End of conflict End of conflict End of conflict +3 years +5 years +10 years 0 -100 Percent of initial GDP -200 -300 -400 -500 -600 -700 -800 CUMULATIVE CHANGE IN WAGES End of conflict End of conflict End of conflict End of conflict +3 years +5 years +10 years 200 100 Percent of initial wage 0 -100 -200 -300 -400 -500 -600 Capital destruction Casualty Economic disorganization Combined shock PRESENT VALUE OF FUTURE WAGES 15 10 Deviation from initial value %5) 5 0 -5 -10 -15 -20 -25 0 5 10 15 20 Years Capital destruction Casualty Economic disorganization Combined shock 94 Disentangling the Economic Impact of the Conflict factors, it also increases rent seeking and cronyism, prevents the connectivity of people, reduces their incentive to pursue productive activities, and breaks economic networks and supply chains. Simulations show that these factors, which are captured by changes in total productivity and matching mechanisms built into the model, by themselves would lead to reduction of GDP by more than half over the course of the conflict. In cumulative terms, losses in GDP are equivalent to six times the annual pre-conflict GDP after 20 years. In practice, there are many mechanisms that drive the disruptions in economic organization. The simulation model in this report captures the effect on economic organization under two aggregate categories: a diminished propensity of matching between active capital and labor and a lower total factor productivity. In practice, there are many factors that feed into these categories, including, but not limited to, higher trade costs, disruptions in supply chains and other economic networks, and intensified rent seeking. Unfortunately, many of these are not quantifiable with the current level of access to information in Syria. The previous chapter showed estimates for increased distance of travel among major cities, which can proxy for higher trade costs; but that is only a small component of the problem. Recently, the Syria Center for Policy Research, documented a significant decrease in Syrian social capital, defined over three broad categories: social networks and community participation, trust, and shared values and attitudes (SCPR 2017). The study’s social capital index, which relies on a community-based survey encompassing all governorates, registered a 30 percent decrease from the beginning of the conflict until mid-2014 when the survey was conducted.29 This decline is largely explained by a sharp fall in the community confidence and networks components. Geographically speaking, the composite index registered the highest decreases in Raqqa (80 percent), Hasake (52 percent), and Idleb (47 percent). Such institutional degradation must have had significant effects on the economy’s capability of putting productive factors into use; however, characterizing the exact channels of transmission and identifying the optimal offsetting mechanisms are beyond the scope of this report and will require more targeted studies. Post-conflict planning should extend attention beyond physical destruction. What would it take to ensure a rapid and sustainable economic recovery in the post-conflict Syria? Our simulations and anecdotal evidence suggest that even reinstating the lost capital, by itself, would not be sufficient to bring the economy back to its pre-conflict level if institutional and organizational challenges are not tackled simultaneously. Regardless of the source of financing, boosting public investments without a comprehensive approach 29 The “Population Status Survey” was conducted with key informants in 698 residential areas across Syria. The social capital index is a simple average of three normalized components: social relations and networks (measured by participation in decision making, volunteering, problem solving cooperation, and women's participation); community trust (measured by trust between individuals and the sense of security); and common values, attitudes, or understandings (measured by the degree of consensus among the members of the community on a shared vision of the region, and the status of women in society). THE TOLL OF WAR 95 would trigger further rent seeking and cronyism, which were important enabling factors in the conflict’s onset, as described in chapter 1. Thus, recovery and reconstruction are by no means an engineering issue; our results show that the issue is primarily an economic and social one, where the incentives of Syrian citizens are at the core. Role of Mobility Mobility has important consequences for economic outcomes. But is it good or bad for Syria? In a textbook economic framework with no externalities, the flexibility to move across regions or jobs always improves the efficiency of the economy. However, does this conclusion hold in the case of a catastrophic event like the conflict where about half of the country’s population is forced to move? In this section, we will try to shed light on the role played by displacement in shaping the impact of the conflict on the Syrian economy. Why have Syrians migrated? The exodus of millions of Syrians triggered a debate about the motivation behind this movement. The conflict has led to a massive outflow of Syrians from their homeland. In December 2016, the total number of registered Syrian refugees was nearing five million (UNHCR). It is important to identify the push and pull factors behind such movement in a careful and detailed manner. Nevertheless, the analysis developed for this study can shed some light on the relative importance of security and economic factors in explaining the mobility of Syrians. Security considerations dominate economic concerns in explaining demographic mobility. Our simulations do not support the view that refugees are mostly economic migrants. To shed light on the motivations to move, the analysis considers a case where individuals are risk-neutral as opposed to being risk-averse, which was assumed in the baseline (a widely accepted moderate degree of risk aversion, σ=0.9, was chosen from the economics literature). Overall, risk-neutral and risk-averse agents migrate by comparing expected wages in different locations; however, risk-averse agents “overreact” and move away disproportionately when the probability of conflict-related deaths increases. The results are shown in table 3.3. Holding everything else constant (including the shocks), when agents are risk-neutral, gross outflows are significantly smaller than that in the baseline (about 10,000 refugees a year during the conflict as opposed to more than half a million under the baseline), and this happens although GDP and wages fall in a comparable manner in these scenarios. This finding shows that a high emphasis on the escalated probability of death that comes with the conflict is the primary driver of mobility. 96 Disentangling the Economic Impact of the Conflict Table 3.3: Economic Outcomes under Risk Neutrality (Migration in millions; other variables in comparison to their pre-conflict values in percentage terms) Conflict Post-conflict 1st Year 2nd Year 3rd Year 4th Year 5th Year 6th Year 10th Year 15th Year 20th Year Investment -84.72 -83.99 -83.30 -82.65 -82.02 -81.44 -8.47 -1.64 -0.91 Outmigration 0.01 0.01 0.01 0.01 0.01 0.01 0.00 0.00 0.00 Wages -55.01 -54.62 -54.33 -54.11 -53.94 -53.82 -25.89 4.49 2.25 GDP -59.37 -62.77 -65.38 -67.42 -69.04 -70.32 -47.11 -13.09 -7.79 PV of Wages -23.77 -25.40 -26.53 -27.24 -27.60 -27.66 -16.99 -7.21 -3.84 What would happen without mobility? The most direct implication of mobility restrictions would be a greater casualty rate. The cumulative conflict-related casualty in Syria was estimated at about 1.5 to 2 percent of the 2010 population as of December 2016. This ratio could easily climb rapidly if Syrians could not migrate. Migration restrictions involve accessibility barriers (like the difficulties of crossing the Mediterranean Sea) and policy barriers (like a destination country sealing off its borders against refugee arrivals). To account for the consequences of these, this analysis considers two additional scenarios. First, we assume that only outmigration is restricted; thus, no Syrians can leave the country, but they can become IDPs. Second, we restrict all mobility across the country; thus, nobody can become an IDP or a refugee. From an economic point of view, a restricted migration regime would increase the Syrian GDP but decrease per person income. Figure 3.7 shows the outcomes of the alternative scenarios. Both migration restriction scenarios show a relatively smaller GDP loss compared with the baseline. For instance, the GDP loss by the 20th year is about 670 percent of the annual pre-conflict GDP when all migration is restricted, and about 760 percent when mobility is free. In contrast, the average aggregate real wage losses amount to 535 percent of the average annual pre-conflict wage in the former, and only 433 percent in the baseline. This contrast is intuitive in this simple framework. Migration restrictions keep the labor force larger despite the additional casualties. This also increases labor per unit of capital, reducing marginal productivity and wages. As a result, each worker receives a smaller wage, but in aggregate, production increases because there are more workers. In a more complex, less stylized framework where, for example, trapped labor might be demotivated or actively resist supporting economic activity (for example, through strikes or major productivity slowdowns), it is not clear that we could not observe both a fall in GDP and in wages. THE TOLL OF WAR 97 Figure 3.7: Dynamics of Impact under Different Migration Restriction Scenarios (Baseline: Free Movement) CUMULATIVE CHANGE IN GDP End of conflict End of conflict End of conflict End of conflict +3 years +5 years +10 years -300 Percent of initial GDP -400 -500 -600 -700 -800 CUMULATIVE CHANGE IN WAGES End of conflict End of conflict End of conflict End of conflict +3 years +5 years +10 years -300 -350 Percent of initial GDP -400 -450 -500 -550 -600 All migration restricted Outmigration restricted Baseline PRESENT VALUE OF FUTURE WAGES 5 0 Deviation from initial value (%) -5 -10 -15 -20 -25 -30 0 5 10 15 20 Years All migration Outmigration restricted Baseline 98 Disentangling the Economic Impact of the Conflict What If…? The Lingering Impact of a Protracted Conflict How will the conflict’s impact on the Syrian economy evolve if the conflict does not stop in 2017? The baseline simulations assumed that the conflict ends in 2017. Alternative conflict-end scenarios are considered to provide a benchmark for the possible evolution of the conflict’s impact on the Syrian economy. Thus, two alternative scenarios are shown: the conflict ends in its 10th year (2021) and the conflict does not end in the near future. In both cases, shocks and parameters are the same as in the baseline specification. With protracted conflict, economic outcomes continue to deteriorate. In the baseline, where conflict ends by the end of the sixth year, GDP recovers by about 20 percentage points (as a share of pre-conflict GDP) by the 10th year (table 3.4). In comparison, it continues to deteriorate in the alternative conflict-end scenarios. In both alternative scenarios, GDP reaches a level that is about 67 percent less than the pre-conflict level by the 10th year. In the case of perpetual conflict, it reaches about 69 percent less than pre-conflict GDP by the 20th year. These scenarios show that as the conflict persists, its impact on current variables increases over time, but at a decreasing rate. several factors drive this slowdown. As conflict-related damage takes place in certain geographic areas and the additional years of the conflict are assumed to have the same pattern, the economic base affected by the conflict shrinks over time. Migration drains the labor force in conflict-affected regions and the capital stock reaches a stable level with new damage and investment ratios. Therefore, the economy converges to a new equilibrium, and adjustment slows, compared with the initial years when the adjustment is rapid. The longer the conflict persists, the slower the recovery would be once the conflict ends. Simulations show that despite the moderation in the rate at which the conflict affects economic outcomes, its effects become more persistent. When the conflict ends in its sixth year, GDP recoups about 41 percent of the remaining gap with its pre-conflict level within the next four years. In comparison, it recoups only 28 percent of the gap in four years if it ends in its 10th year. The total loss in GDP reaches 760 percent of the annual pre-conflict GDP by the 20th year in the baseline (figure 3.8). With a perpetual conflict, this loss stands at 1,320 percent. When the conflict ends in the 10th year, the present value of future wages recovers back to pre-conflict levels by the 20th year. In the case of a perpetual conflict, it never recovers. Simulations also show that outmigration doubles between the sixth year of the conflict and the 20th year. Thus, the supply of migrants will not cease as long as there is conflict. THE TOLL OF WAR 99 Table 3.4: Economic Outcomes under Various Conflict End-Date Scenarios (Migration in millions; other variables in comparison to their pre-conflict values in percentage terms) Conflict 1st Year 2nd Year 3rd Year 4th Year 5th Year 6th Year 10th Year 15th Year 20th Year BASELINE (CONFLICT ENDS IN 6TH YEAR) Investment -83.18 -81.77 -80.61 -79.58 -78.66 -77.80 -13.77 -7.31 -6.50 Outmigration 0.81 0.70 0.63 0.58 0.53 0.49 0.00 0.00 0.00 Wages -54.99 -54.81 -54.70 -54.64 -54.60 -54.57 -23.53 10.31 8.55 GDP -58.34 -60.60 -62.20 -63.42 -64.38 -65.15 -42.38 -13.16 -11.12 CONFLICT ENDS IN 10TH YEAR Investment -83.18 -81.77 -80.61 -79.58 -78.66 -77.80 -74.84 -14.83 -7.80 Outmigration 0.81 0.70 0.63 0.58 0.53 0.49 0.00 0.00 0.00 Wages -54.99 -54.81 -54.70 -54.64 -54.60 -54.57 -54.49 -17.22 11.38 GDP -58.34 -60.60 -62.20 -63.42 -64.38 -65.15 -67.10 -40.00 -14.37 PERPETUAL CONFLICT Investment -83.18 -81.77 -80.61 -79.58 -78.66 -77.80 -74.84 -71.82 -69.36 Outmigration 0.81 0.70 0.63 0.58 0.53 0.49 0.37 0.26 0.19 Wages -54.99 -54.81 -54.70 -54.64 -54.60 -54.57 -54.49 -54.39 -54.31 GDP -58.34 -60.60 -62.20 -63.42 -64.38 -65.15 -67.10 -68.23 -68.75 100 Disentangling the Economic Impact of the Conflict Figure 3.8: Dynamics of Impact under Various Conflict End-Date Scenarios CUMULATIVE CHANGE IN GDP 7th year 10th year 15th year 20th year -300 -500 Percent of initial GDP -700 -900 -1100 -1300 -1500 CUMULATIVE CHANGE IN WAGES 7th year 10th year 15th year 20th year -300 -400 Percent of initial GDP -500 -600 -700 -800 -900 -1000 -1100 -1200 Perpetual conflict Conflict ends in 10th year Baseline (conflict ends in 7th year) PRESENT VALUE OF FUTURE WAGES 10 0 Deviation from initial value (%) -10 -20 -30 -40 -50 -60 0 5 10 15 20 Years Perpetual conflict Conflict ends in 10th year Baseline (conflict ends in 7th year) A Syrian refugee family, from Aleppo, staying under a shelter during a rainy day on March 8, 2014, at Uskudar in Istanbul. Photo Credit: Bulent Kilic/AFP/Getty Images. CONCLUDING REMARKS 102 Concluding Remarks CONCLUDING REMARKS As this report was being prepared, the conflict was still active. By the end of March 2017, conflict was still active in large number of fronts, and despite much efforts brokered by many parties, including those of the international community, a stabilization and resolution path out of the conflict remained elusive. The economic and social impacts of the conflict will continue to unfold, albeit with different magnitude and composition. The brief account outlined in this report stressed the combination of several risk factors that contributed to the onset, duration, and intensity of the conflict. These included a mass popular mobilization spurred by the spread of the Arab Spring to Syria in 2011, a larger regional context in which armed rebellion was made feasible and was kept going by the support of various powers, and the increasingly sectarian character of the conflict, which made credible commitments to a reasonable resolution impossible. Whereas the first two factors may be deemed passé as of now, the last factor is not. More importantly, the economic and social impacts of the conflict may in turn determine the future path of the conflict. The importance of economic factors in shaping the outcomes on the ground has increased. An important question that arises from the analysis thus far is the following: in what direction could the economic impact of the conflict affect peace and stability? It is obviously very difficult to answer this question. However, it is unambiguous that the conflict has increased regional inequalities, war economy, and rent seeking, and further depressed civil liberties and polarized the Syrian society. Although these factors played only an enabling role for the onset of the current conflict, with much worsened outcomes now, they could play a more prominent role in further complicating a reasonable stabilization. The economic and social impacts of the conflict also amount to immense practical challenges. Syria will eventually have to overcome a multitude of urgent economic and social challenges to promote peace and stabilization. The conflict has cast a wide swath of the population into poverty, and reoriented several million toward livelihood sectors that will not be sustainable in a post-war economy. A whole generation of children has received inadequate education. This, coupled with a significant brain-drain, has caused a dramatic decline in Syria’s human capital. The immense spectrum of needs arising from the conflict will require a strident response. Planning, coordination, and implementation arrangements will need to reflect the dual challenge of providing immediate peace dividends and humanitarian support to those in need, while also strengthening the national systems and capacities that are necessary to eliminate poverty, promote development objectives, and help to implement a new vision for Syria. 104 Appendix A P P E N D I X A . METHODOLOGY AND ASSESSMENTS Methodology of Physical Damage Assessments Physical damage assessments were conducted in two stages: a) City-level assessments. These provided an assessment of physical damage for the 10 cities (Aleppo, Dayr az-Zawr, Raqqa, Dar’a, Idlib, Douma, Kobane, Tadmur (Palmyra), Homs, and Hama) as of January 2017. b) Governorate-level assessments. These provided modified extrapolations of city-level assessments to estimate physical damage for the country’s eight most conflict-affected governorates (Aleppo, Homs, Hama, Dayr az-Zawr, Raqqa, Dar’a, Idlib, and Rif Dimash). In both stages, the following sectors were included in the analysis: housing; water and sanitation (public network), energy (public network), health (public facilities), education (private and public facilities), and transportation (primary road network). City-Level Assessments The baseline for all six sectors was created and refined for Aleppo, Dar’a, Homs, Hama, and Idlib cities during previously conducted Damage and Needs Assessments (DNAs). A baseline of facilities was created for the five cities not previously covered in DNA efforts (Dayr az-Zawr, Douma, Kobani, Raqqa, and Tadmur). Table A.1 provides the number of facilities assessed per sector by city. Table A.1: Number of Facilities Assessed per Sector, by City City WASH Power Housing Health Education TOTAL Road FACILITIES length Aleppo 198 333 88,384 121 475 89,511 3,123,681 Dar’a 26 10 9,443 25 137 9,641 224,672 Dayr 23 40 5,712 42 120 5,937 479,199 az-Zawr Douma 11 17 5,578 12 49 5,667 128,462 Hama 99 396 22,977 47 217 23,736 610,216 Homs 26 274 42,017 90 287 42,694 1,113,961 Idlib 30 70 5,896 15 56 6,067 181,346 Kobani 24 3 5,673 3 18 5,721 166,665 Raqqa 13 37 7,810 25 43 7,928 688,997 Tadmur 7 10 1,364 4 24 1,409 271,047 TOTAL 457 1,190 194,854 384 1,426 198,311 6,988,246 THE TOLL OF WAR 105 Baseline identification and remote-based verification of the damage and functionality of facilities was conducted through examination of (1) commercial half-meter resolution imagery, (2) social media data, and (3) publicly available sources of information, including Syrian government reports and nongovernmental organization documents. Employing a multi-pronged approach allowed for cross-validation of data and cycles of refinement. In addition, remote sensing data, such as Visible Infrared Imaging Radiometer Suite day/night band light detection imagery, was consulted as part of the analytic output testing process for the power sector. The physical status of each facility was classified as follows: • Destroyed, if damage exceeds 40 percent of the structure • Partially damaged, if damage is less than 40 percent of the structure • No damage, if damage is at 0 percent. The operational status of each facility is also provided—functioning or not functioning—as source reporting is available. Governorate-Level Assessments As the methodology developed for the city-level analysis relies on detailed observations and labor-intensive analysis, repeating the same methodology for the entire Syrian landscape was prohibited by timing and resource-constraints. Thus, the governorate-level assessment employed extrapolation of all six sectors to provide damage results for the eight governorates in which the 10 surveyed cities are located. Baseline statistics came from the following sources: • Housing and water, sanitation, and hygiene (WASH): access to the public WASH network by residential units, from the Syrian Government 2004 Housing Census). • Power: data on major power plants from Syrian General Institute for Electricity Generation reports. • Education: data on primary and secondary schools from the Syrian Center for Statistics 2010 Survey of the Syrian Education System. • Health: data on private hospitals, public hospitals, medical centers, and specialized medical centers from the Syrian Center for Statistics 2010 Survey on Syrian Medical Facilities. • Transportation: data on main roads, highways, and primary roads from OpenStreetMap. The extrapolation process required statistics to be compiled at the district level (the administrative unit immediately below governorate); however, the 106 Appendix education and schools data available from the Syrian Center for Statistics are only provided at the governorate level. Conflict dynamics analysis at the city and district levels facilitated extrapolation of city-level damage results to the governorate level. Specifically, this entailed creating a conflict-induced damage inference matrix; applying the matrix to cities to calculate composite sector damage estimates for high-, medium-, and low-conflict environments; applying the matrix to districts to first assign expected high, medium, and low conflict-induced damage levels, and next transfer corresponding city-level composite damage percentages to baseline figures for the district; and finally, aggregating these district-level estimates to the governorate level. The conflict-induced damage inference matrix was developed to estimate high, medium, and low damage that take into account three major types of conflict tactics as well as their intensity. In addition, it factors in observed resilience in the country by including reconstruction (undertaken by a variety of local actors) as an indicator type (table A.2). Table A.2: Conflict-Induced Damage Inference Matrix Indicator Type High Prevalence Moderate Prevalence Low Prevalence Low intensity conflict Widespread (3) Targeted (2) Close to none (1) Airstrikes Widespread (3) Targeted (2) Close to none (1) Heavy Artillery Widespread (3) Targeted (2) Close to none (1) (Tanks, Mortars, etc.) Reconstruction Extensive (-3) Moderate (-2) Limited or none (-1) Any city or district with less than 2 points was assigned a low damage estimate; 3 to 5, a medium damage estimate; and 6 or above, a high damage estimate. The damage inference matrix was applied to the 10 cities in the study, yielding the following damage category groupings: • High damage: Aleppo, Dayr az-Zawr, Douma, Idlib, and Tadmur • Medium damage: Dar’a, Kobani, Raqqa, and Homs • Low damage: Hama. Finally, the city composite estimates were used to calculate district- and governorate-level damage. Of 42 districts located in the eight governorates, 29 districts were classified as high damage, eight districts as medium damage, and five districts as low damage. These composite percentages of damage were applied to the baseline figures for each district, at the high, medium, or low level as appropriate. This generated figures for housing, health, and education. The composite percentages for housing were also applied to the transportation sector (using total road length for the major road categories of highway, trunk, and primary roads as the baseline), as well as to the WASH sector (using total number of households accessing the public network as the baseline). THE TOLL OF WAR 107 A P P E N D I X B . SECTORAL SITUATION REPORTS Table B.1: Situation Report: Water and Sanitation Infrastructure City Source Water Availability Damages Coping Mechanisms Aleppo 150 Public and As of 15 February 2017, Pumping stations and piping network Residents are 300 Private Wells no residents in Aleppo heavily damaged, especially in eastern and dependent on wells have accessed water via southern neighborhoods and trucks the public network since 14 January Dara’a Wells and Water Public water network is Damage to two pumping stations and Wells and trucks, in Trucks not fully functional; in connecting pipelines absence of public opposition-held Dar’a networks; trucks face al-Balad, no information difficulty in entering available on water opposition held availability al-Balad Dayr Partially Access to public network Largest water treatment station and most Wells, Euphrates river az-Zawr functioning public is limited; al-Joura and pumping stations are damaged and water tank network al-Qusour neighborhoods receive water once every week for a few hours, on average Douma Wells and hand Public water network is Pipeline likely to be heavily damaged Small scale wells and pumps cut off. because of frequent and heavy bombings, hand-pumps, few some pumping stations damaged. trucks Hama Public Network 90-100% coverage Public network remains undamaged and 0-9% use truck or through public network functioning, but the main line outside the alternative sources. city is damaged Homs N/A N/A Partial damage Availability of 60 wells acts as a back-up Idlib Public Network No consistent water Piping network damages, lack of Trucks and wells supply from public electricity network; Rotating schedule, with each neighborhood receiving water every 12 days Kobani Partially Some neighborhoods are Pumping stations and tanks are not Trucks and wells functioning public disconnected from public accessible, due to ISIS takeover and are network; city relies network; supply limited likely partially damaged on 15 wells to in time and pressure in supply piped connected network neighborhoods Raqqa Public Network Decreased/limited Electricity and fuel shortages, stolen N/A supply equipment; water outages for long periods due to air strikes, likely future damages to pipeline through frequent airstrikes Tadmur N/A No access to public Extensive damage to piping network since network the start of conflict and wells likely N/A remains partially damaged; likely more damages due to latest ISIS takeover 108 Appendix Table B.2: Situation Report: Power Infrastructure City Public Grid (hours of electricity hours per day) Commercial and Private Generators Comments West 0-2 hours (January 2017). The 42%. Commercial and private West Aleppo’s private Aleppo transmission line connecting Aleppo city generators have become an generators are regulated to the grid in Hama has been cut. Most important source of electricity. 108 through the city’s public public infrastructure in Aleppo private generators surveyed electricity administration and governorate has been destroyed or provided on average 8–10 hours of are legally obliged to provide a disconnected. electricity per day. minimum of 10 hours of electricity per day. East 0 hours. The eastern districts of the city No data available. Before the recapture by Aleppo have been largely destroyed and cut off government forces, residents from public services. relied on a combination of public and private generators, as well as improvised solutions, such as solar energy. Damascus 4-12 hours. Reports suggest an average No data available. Data for Damascus vary heavily of 12 hours of electricity a day, with by neighborhoods, especially peripheral neighborhoods faring worse: between central districts and Qudsaya 6 hours, downtown Mezzeh 4 outlying suburbs. hours, Sahnaya and Barza 2 hours and Artouz 0 hours. Douma 0 hours.- Douma has been cut from the 8% on average from private and Private and commercial public grid. commercial generators. generators reportedly rely on the now disused public electricity infrastructure. Hama 1-9 hours. Most neighborhoods in Hama Commercial electricity generators N/A receive between 5 and 9 hours of are available and in use. electricity from the public grid per day. Masyaf 4 hours. No data available. N/A Salamiyeh 2 hours. No data available. N/A Tadmur 0 hours. Following the recapture of the Generators operate to support The city experienced few power (Palmyra) city by ISIS in December 2016, electricity water wells and core infrastructure. outages prior to its initial has been cut. capture by ISIS in May 2015. Latakia 1-4 hours. Reports in January 2017 of No data available. N/A total blackouts for entire areas (Dator and Ramlat Shmali), and 20 to 22 (Dahiet al-Assad, al-Tabiyat, Saliba) hours of rationing for most other neighborhoods. The situation is reportedly worse in the countryside. Syria Direct averages outages at 1 hour of electricity per day. Tartus 6 hours. Officials from Tartus claim 1.5 No data available. N/A hours of electricity alternating with 4.5 hours of rationing (December 2016). Idlib Up to 1.2 hours for most neighborhoods, Roughly 75% of Idlib residents own Idlib’s public electricity system except ad-Dabit (4 hours) and or subscribe to 29 private electricity relies on 11 public generators Market/City Center (2 hours). generators (range 55% in administered by the al-Arbaeen to 95% in As-Souq and opposition’s Public Services Hadiqat al-Basel). Operators are Administration. The city has allowed to provide up to 4.5 hours been cut off from the central per day (July 2016), 18.75%. state electricity grid. Raqqa 8 hours. The city receives on average 8 No data available. N/A hours of electricity primarily from Al-Furat Hydroelectric Dam. Kobani 24 hours. Local sources report that by No data available. N/A January 2017, residents received 24 hours of electricity from the public grid. THE TOLL OF WAR 109 Table B.3: Situation Report: Health Care Facilities City Health Status Alternative Arrangements/Other Efforts Aleppo • 58% of facilities in Aleppo have been destroyed or • Eastern Aleppo is heavily reliant on temporary health damaged. points and mobile clinics • 84% of facilities in Eastern Aleppo were reported to be • IDP camps in outskirts only have access to mobile non-functioning. clinics. Dara’a • 65% of facilities are partially damaged • In Dara’a Al-Mahata, government attempts to re-staff • Both government-controlled Dar’a al-Mahatta and hospitals and provide new equipment. opposition-controlled Dar’a al-Balad experience • In Dara’a Al-Balad, support is received from medical shortages of staff. NGOs. • Hospitals in both areas are under threat. • Some medicines and equipment were also smuggled from government-controlled areas. Dayr • 88% of facilities have been destroyed or damaged • Patients in need from opposition-controlled areas az-Zawr • six facilities are currently functioning were transported to Turkey or regional hospitals. • shortage of medicines and medical supply • Syrian government has reportedly air dropped a ton of medical aid into western Dayr az-Zawr • International NGOs have been providing medical supplies to the city via airdrop since the start of the siege in 2014 Douma • All facilities partially damaged and only two facilities N/A functioning in the city Hama N/A N/A Homs N/A N/A Idlib • Facilities operate with limited capacity; of 15 identified • There are evidences of opening of new facilities and facilities, 9 remain partially damaged, 3 are destroyed, also staff hiring announcements and 3 are intact. Kobani • Two out three hospitals are functioning, one is • The two functioning hospitals were rebuild/repaired experiencing shortages in medical supplies and and re-opened. equipment Raqqa • Before crisis, private clinics provided most of Raqqa’s • International aid organizations began providing health health care services; there was only one large public services, but they were forced to stop following the medical facility in the city takeover of the city by ISIS. Tadmur • Under ISIS, nearly daily attacks damaged/destroyed all N/A medical facilities 110 Appendix Table B.4: Situation Report: Education Facilities City Education Status Alternative Arrangements/Other Efforts Aleppo • 53% of schools are damaged; 10% completely • UNESCO and UNICEF have supported the destroyed construction and rehabilitation of 23 schools between • Nearly all schools in Eastern Aleppo are destroyed December 2016 and January 2017 Dara’a • 15% of facilities are destroyed and 37% are partially N/A damaged. Dayr • 62% of the facilities show signs of damange • Government has re-opened some schools but in ISIS az-Zawr controlled areas, due to security threats, teachers and students stay out of schools. Douma • 90 percent of all education facilities are partially N/A damaged • Schools remain vulnerable: about 10% are used as bomb shelters Hama N/A N/A Homs N/A N/A Idlib • 68% of Idlib’s schools have sustained partial damage • According to a February 2017 statement, 73 percent of Save the Children-supported schools in Idlib and Aleppo have suffered severe airstrike damage. Kobani • 12 schools show no signs of damage, 2 have been • Local rehabilitation efforts have been ongoing in the destroyed, and 4 have been partially damaged. educational sector since April 2015. • 15 schools were rehabilitated, and accommodate 50,000 students as of January 2017. Raqqa • Of the 43 schools identified, 16% have been • In addition to al-Furat University, the Islamic State is destroyed, and 63% registered as partially damaged operating as many as 24 schools. Tadmur • 83% of schools are partially damaged N/A • As of January 2017, no schools or educational facilities are reported operational within the city THE TOLL OF WAR 111 A P P E N D I X C . TECHNICAL SPECIFICATIONS OF THE SIMULATION MODEL A. Key Features of the Model 112 Appendix B. Introduction of Notation THE TOLL OF WAR 113 114 Appendix B. The Model THE TOLL OF WAR 115 116 Appendix THE TOLL OF WAR 117 D. Solution Algorithm E. Scenarios 118 Appendix THE TOLL OF WAR 119 REFERENCES Abadie, A. and Gardeazabal, J., 2003. The economic costs of conflict: A case study of the Basque Country. The American Economic Review, 93(1), pp.113-132. Abadie, Alberto & Diamond, Alexis & Hainmueller, Jens, 2010. “Synthetic Control Methods for Comparative Case Studies: Estimating the Effect of California’s Tobacco Control Program,” Journal of the American Statistical Association, American Statistical Association, vol. 105(490), pages 493-505. Acemoglu, D. and Robinson, J.A., 2013. Why nations fail: The origins of power, prosperity, and poverty. Crown Business. Adhikari, B. and Alm, J., 2016. Evaluating the economic effects of flat tax reforms using synthetic control methods. Southern Economic Journal, 83(2), pp.437-463. Akbar, T., Aoun, J, Arshad, R., and Tamma, W. 2017. Urban Space, Housing and Municipal Services, mimeo, the World Bank. Alix-Garcia, J., E. Artuc, and H. Onder. 2017. The Economics of Hosting Refugees: A Host Community Perspective from Turkana. World Bank: Washington, DC. Almohamad, Hussein, and Andreas Dittmann. 2016. “Oil in Syria between Terrorism and Dictatorship.” Social Sciences 5 (2). Artuc, E. and Onder, H. 2017. Disentangling the Economic Impact of the Conflict, mimeo, The World Bank. Badra, Joude. 2015. “The Syrian Financial Sector.” 2013–2014 Syrian Economic Reconstruction Project, Carnegie Middle East Center. Baker, A., 2014. Why al-Qaeda Kicked Out Its Deadly Syria Franchise. Time Magazine. Barnes, J. 2009. “Managing the Waters of Bath Country: The Politics of Water Scarcity in Syria.” Geopolitics 14 (3). Baroutt, M.J., 2011. The past decade in Syria: the dialectic of stagnation and reform (Part 1 of 5). Research Paper Arab Center for Research and Policy Studies, Doha Institute. Batatu, H., 1999. Syria’s Peasantry. The Descendants of its Lesser Rural Notables, and their Politics, Princeton: Princeton UP. 120 References Cole, J. 2010. Syria Land use Map at Informed Comment, https://www.juancole.com/2010/12/map-syria-land-use.html, accessed at July 4, 2017. Davis, D. R., and D. E. Weinstein. 2002. “Bones, Bombs, and Break Points: The Geography of Economic Activity.” American Economic Review 92 (5): 1269–89. Douwes, D. 2000. The Ottomans in Syria: A History of Justice and Oppression. London: I.B. Tauris. Dun & Bradstreet. 2016. “Syria Country Insight Report.”Dun & Bradstreet, Short Hills, NJ. FAO (Food and Agriculture Organization) and WFP (World Food Programme). 2015. “Crop and Food Security Assessment Mission. WFP, Rome. FAO (Food and Agriculture Organization) and WFP (World Food Programme). 2016. “Crop and Food Security Assessment Mission to the Syrian Arab Republic.” FAO and WFP, Rome. Fearon, James D. 2017. “Notes on the Causes of the Civil War in Syria, Its Continuation, and Implications for International Reconstruction Efforts.” Background paper for the Syria Economic and Social Impact Assessment. World Bank, Washington, DC. Fearon, James D., and David Laitin. 2003. “Ethnicity, Insurgency, and Civil War.” American Political Science Review 97: 75–90. Gallup, J.L., Sachs, J.D. and Mellinger, A.D., 1999. Geography and economic development. International regional science review, 22(2), pp.179-232. Ghanem, H. (ed.). 2016. The Arab Spring Five Years Later. Washington, DC: Brookings Institution Press. Gleick, Peter. 2014. “Water, Drought, Climate Change and Conflict in Syria.” Weather, Climate and Society 6 (3): 331–40. Hamilton, K. and Nguyen, T. 2017. Estimating Syrian Human Capital Loss Over the Civil War Years, mimeo, The World Bank. Hinnebusch, R.A., 1995. The political economy of economic liberalization in Syria. International Journal of Middle East Studies, 27(3), pp.305-320. Hourani, A. 1946. Syria and Lebanon: A Political Essay. London: Oxford University Press. THE TOLL OF WAR 121 HRW (Human Rights Watch). 1996. “Syria: The Silenced Kurds.” Human Rights Watch 8 (4) (E). Ianchovichina, E., L. Mottaghi, and S. Devarajan. 2015. “Inequality, Uprisings, and Conflict in the Arab World.” Middle East and North Africa Economic Monitor, October. World Bank, Washington, DC. IMF (International Monetary Fund). 2016. “Syria’s Conflict Economy.” IMF Working Paper WP/16/123. IMF, Washington, DC. International Crisis Group. 2009. “Reshuffling the Cards? (I): Syria’s Evolving Strategy.” International Crisis Group, Washington, DC. IOM (International Organization for Migration). 2017. National Population Monitoring (NPM): Access to Services: Obstacles and Dynamics. Geneva: IOM. Issawi, C. 1988. The Fertile Crescent; 1800-1914: A Documentary Economic History. New York: Oxford University Press. ———. 1995. The Middle East Economy: Decline and Recovery. Princeton: Markus Wiener Publishers. Kelley, C.P., Mohtadi, S., Cane, M.A., Seager, R. and Kushnir, Y., 2015, December. Climate Change In The Fertile Crescent And Implications Of The Recent Drought In Syria. In AGU Fall Meeting Abstracts. Khalaf, S. 1987. Lebanon’s Predicament. New York: Columbia University Press. Khoury, P. S. 1987. Syria and the French Mandate. London: I. B. Tauris. Mueller, H., L. Piomentese, and A. Tapsoba. 2017. “Recovery from Conflict: Lessons of Success.” Policy Research Working Paper 7970, World Bank, Washington, DC. Nunn, N. and Puga, D., 2012. Ruggedness: The blessing of bad geography in Africa. Review of Economics and Statistics, 94(1), pp.20-36. Onder, H., G. Ponthiere, and P. Pestieau. 2017. “The Domestic Welfare Loss of Syrian Civil War: An Equivalent Income Approach.” Mimeo, World Bank, Washington, DC. Owen, Roger. 1981. The Middle East in the World Economy, 1800-1914. Methuen: London. PHR (Physicians for Human Rights). 2015. “Doctors in the Crosshairs: Four Years of Attacks on Health Care in Syria.” March 2015. https://s3.amazonaws.com/PHR_Reports/doctors-in-the-crosshairs.pdf. 120 References REACH Initiative. 2015. “Migration Trends and Patterns of Syrian Asylum Seekers Travelling to the European Union.” Assessment Report. REACH, Geneva. Sahner, Christian C. 2014. Among the Ruins: Syria Past and Present. London: Hurst & Co. Save the Children. 2017. Invisible Wounds: The Impact of Six Years of War on the Mental Health of Syria’s Children. London: Save the Children. SCPR (Syrian Center for Policy Research). 2013. “Socioeconomic Roots and Impact of the Syrian Crisis.” SCPR, Damascus. SCPR (Syrian Center for Policy Research). 2014. “Alienation and Violence: Impact of Syria Crisis Report.” ———. 2017. “Social Degradation in Syria: The Impact of Conflict on Social Capital.” SCPR, Friedrich Ebert Stiftung, Beirut. Sparrow, A., 2014. Syria’s polio epidemic: the suppressed truth. The New York Review of Books. Syria Public Health Network. 2016. “Syria’s Health: An Overview of Current Knowledge and Policy Priorities.” https://syriahealthnetwork.files.wordpre -ss.com/2016/02/syria-health-policy-brief_london-conf_final-3rdfebruary 2016.pdf. Syrian American Medical Society. 2017. “The Failure of UN Security Council Resolution 2286 in Preventing Attacks on Healthcare in Syria.” Syrian American Medical Society, Washington, DC. UNOCHA (United Nations Office for the Coordination of Humanitarian Affairs). 2016. 2017 Humanitarian Needs Overview: Syrian Arab Republic. New York: UNOCHA. UNDP (United Nations Development Programme). 2005. Poverty in Syria: 1996-2004: Diagnosis and Pro-Poor Policy Considerations. UNDP (United Nations Development Programme). 2011. Poverty and Inequality in Syria (1997–2007) UNDP (United Nations Development Programme). 2014. Human Development Report. New York: UNDP. UNESCWA (United Nations Economic and Social Commission for Western Asia). 2016. Syria at War. UNESCWA and University of St. Andrews Publication. UNESCWA, Beirut. THE TOLL OF WAR 121 United Nations. 2016 “Whole of Syria: Who Does What Where per Governorate.” United Nations Office for the Coordination of Humanitarian Affairs, Syrian Arab Republic. Vanly, I.C. 1992. “The Kurds in Syria and Lebanon.” In The Kurds: A Contemporary Overview, edited by Philip G. Kreyenbroek and Stefan Sperl. Routledge. Wada Y., L. P. H van Beck, and M. F. P. Bierkens. 2012. “Non-Sustainable Groundwater Sustaining Irrigation: A Global Assessment.” Water Resources Research 48 (6). World Vision. 2016. The Cost of Conflict for Children. Amman, Jordan: World Vision. Photography of wall painting in Aleppo by Khalil Hajjar/Middle East Eye (see the original article at: http://www.middleeasteye.net/in-depth/features/pictures-aleppo-artists-paint-pictures-revolution-war-city-walls-2075898635). This report studies the economic and social impact of the armed conflict in Syria. Bringing together conflict-driven loss of lives, forced displacement, physical damages in capital, and overall economic collapse, it takes stock of the consequences brought about by one of the most brutal conflicts of our time. The analysis extends our understanding of the Syrian conflict by focusing on the distinct roles played by physical destruction, casualties, and disruptions in economic organization in driving the economic and social impact of the conflict. Although visual effects, such as physical destruction, are more often noticed, the “invisible” effects, such as disruptions in economic networks, increased rent seeking, and the erosion of social trust, play a greater role in explaining the collapse in economic activity. The persistence of economic effects also differs across various channels. The economic impact of conflict-related casualties is much more persistent than that of other channels. As this report was written, the Syrian conflict was ongoing in its sixth year. The report suggests that the longer the conflict lasts, the slower and more difficult the post-conflict recovery will be.