Lake Chad Regional Economic Memorandum  |  Development for Peace Technical Paper 1. Socioeconomic Trends in the Lake Chad Region Takaaki Masaki (World Bank) and Carlos Rodríguez-Castelán (World Bank) 83 Lake Chad Regional Economic Memorandum  |  Development for Peace 2.1 Introduction The Lake Chad region—which is an economically- main correlates of the long-term transformation of the and socially-integrated area spanning across four region? countries of Chad, Cameroon, Niger, and Nigeria in north-west Africa—has been trapped in a vicious To answer these questions, we draw on the analytical circle of suboptimal territorial development and framework developed in WDR2009 which focuses on fragility. People in the region are confronted with three dimensions of economic geography—namely, multidimensional development challenges ranging from density, distance and division. First, density refers to the limited economic opportunities, poor governance, lack of economic mass per unit of land area, or the geographic basic access to services, among others. These challenges compactness of economic activity. It is shorthand for also have fueled violence and conflicts in the region. the level of output produced—and thus the income Ten years of conflict, mainly driven by the Boko Haram generated—per unit of land area. It can, for example, be insurgency, have left an estimated 12.8 million people measured as the value added or gross domestic product in need of humanitarian assistance in the Lake Chad (GDP) generated per square kilometer of land. Second, region, 2.7 million of whom are people displaced by the distance refers to the ease or difficulty for goods, services, conflict.202 labor, capital, information, and ideas to traverse space. It measures how easily capital flows, labor moves, goods Despite the salience of the Lake Chad region as a hub are transported, and services are delivered between two for regional stability and growth, there is a lack of data locations. Distance, in this sense, is an economic concept, and evidence on the socio-economic landscape of the not just a physical one. Lastly, division, by contrast, refers region. Identifying opportunities for inclusive growth to any restrictions on the mobility of people, goods and in the region entails understanding its current status in services due to border restrictions, territorial disputes, terms of its economic activities/trends, access to services, civil wars, and conflicts between regions and countries, FCV (fragility, conflict, and violence) challenges and among others. human development as well as its status vis-à-vis other areas of the countries.  igure 2.1 graphically shows how this territorial F development perspective fits in the overall conceptual The main objective of this note is to fill in the knowledge framework for the Lake Chad Regional Economic gap on the socio-economic trends of the Lake Memorandum. Chad region by providing a descriptive snapshot of economic activities, poverty, and human development This note shows that the Lake Chad region lags behind in the Lake Chad region. The main questions that this in multiple dimensions of development ranging from note seeks to answer are: (i) what are the current levels poverty, human capital, and access to services. A of economic activity and social inclusion in the region poverty rate in the Lake Chad region is found to be much compared with other regions in the bordering countries?; higher than other parts of the countries surrounding the (ii) how do these socio-economic outcomes change lake. The regional poverty rate in the Extreme North within the region and over time?; and iii) what are the region of Cameroon (59 percent) is three times higher 202 Of the 12.8 million people in need of humanitarian assistance, 10.6 million are from northeastern Nigeria’s three most-affected states: Adamawa, Borno, and Yobe (USAID, 2020). 84 2.1 Introduction Technical Paper 1. Socioeconomic Trends in the Lake Chad Region Figure 2.1:  Conceptual Framework Highlighting Opportunities to Strengthen Territorial Development and Addressing Fragility in the Lake Chad Region Policies to improve territorial development • Supporting higher densities and agglomeration • Ensuring shorter economic and physical distances • Reducing divisions: (social, cultural, ethnic—thick borders) to facilitate regional convergence Suboptimal Territorial Fragility, Violence & Dimensions Development Link between challenges deviates Conflict Challenges of economic lagging regions from their long • Lack of economic term economic potential: • Violent conflict geography: agglomeration Lower: • Ineffective institutions • Density • Limited flow of factors • Climate change/harsh • growth • Distance of production environmental • poverty reduction human • Division • Lack of coverage development outcomes conditions (WDR 2009) (WDR 2010, 2011) Policies to reduce FCV • Service delivery (WDR 2004) • Risk management (WDR 2014) • Governance (Cooperation, Coordination, Commitment)(WDR 2017) Source: World Bank elaboration based on World Bank 2003, 2009, 2010, 2011, 2013b, 2017. that of the rest of the country (19 percent). In Nigeria, the 2014, its Lake Chad region experienced a slight increase Lake Chad region203 has a poverty rate (72 percent) nearly in poverty from 48 percent to 52 percent. The analysis of twice as high as in the rest of the country (38 percent). nighttime light also shows no sign of convergence in the Chad is the only exception, where the poverty rate in the level of economic activities between the Lake Chad and country’s Lake Chad region (31 percent) is lower than rest of the countries. Overall, there is little indication that the rest of the country (40 percent).204 This is explained such spatial gaps in poverty and economic development by the fact that the Chad region around the lake lies near are narrowing. the capital of the country, with a consequently higher urbanization rate and a relatively high population density. One of the key drivers for regional growth lies in improving connectivity to large markets. A large swath There is little sign that the spatial gap between the Lake of rural areas in the Lake Chad region still have limited Chad and other parts of the countries is narrowing access to large markets and, as highlighted earlier, these over time. In Cameroon, for instance, poverty declined areas are typically characterized by slower growth in by 4 percent in the Extreme-North region compared to a economic activities. Our estimate suggests that about 6 percent decrease in the rest of the country between 2007 60 percent of rural population in the Lake Chad region and 2014. While Niger as a whole saw a reduction in live farther than 2km away from all-season roads (proxied poverty from 51 percent to 45 percent between 2011 and by OpenStreetMap), while this rate is much lower in 203 The country’s Borno state is excluded from the analysis given that there is no representative household survey for that state. 204 Mahmood and Ani (2018). 2.1 Introduction 85 Lake Chad Regional Economic Memorandum  |  Development for Peace other parts of the countries, at about 30 percent. Poor The note is organized as follows. Section 2.2 provides road infrastructure undermines intra- and inter-regional key statistics on poverty, sector of work, and human connectivity across the Lake Chad Region, thereby capital indicators in the Lake Chad region vis-à-vis other limiting economic opportunities for people in the region. parts of the country and examine how the Lake Chad lags On the other hand, areas proximate to population centers behind in different dimensions. Section 2.3 provides a tend to have greater market accessibility due to a wider diagnostic of economic geography with a focus on three and dense road network system, which allows people dimensions of density, distance and division. Section living near the city to benefit from its agglomeration 2.4 identifies a set of structural factors, aggregate shocks economies. Our analysis of nighttime light also reveals and selected policies that might be associated with the that areas with greater market access experience a faster dynamics of economic activity and social inclusion across local economic growth, attesting to the important role the region. that road infrastructure and access to markets play in driving regional growth. Larger urban agglomerations continue to expand their economies while leaving smaller ones behind. Our analysis shows that after accounting for various other geographic and socio-economic factors, larger urban agglomerations tend to grow faster economically than smaller ones. In fact, vibrant economic dynamics measured by changes in nighttime light luminosity are spatially concentrated in areas that are already highly populated initially while those sparsely populated areas have continued to see little economic progress over time. This attests to a potential widening gap in major cities vis-à-vis smaller periphery towns or rural areas. Climate and conflict risks also play a role in determining the trajectory of regional growth. Our econometric findings reveal that agrarian areas that experienced favorable weather conditions experienced a faster rate of growth in economic activities compared to less favorable weather conditions. Given that climate change has made weather increasingly more erratic within the Lake Chad and also in the surrounding countries, the economic implications of such climate shocks need to be evaluated carefully depending on local contexts. Another significant economic threat comes from the highly volatile security situation created by the Boko Haram. The Boko Haram activities are also negatively associated with the pace of local economic growth and mitigating security risks should remain among the top priorities for ensuring the sustainable growth of the region. 86 2.1 Introduction Technical Paper 1. Socioeconomic Trends in the Lake Chad Region 2.2 Recent Socio-economic Trends in the Lake Chad The Lake Chad region is an economically- and An estimated 30 million people live in the Lake Chad socially-integrated area in north-west Africa that has region206 and they are primarily involved in agriculture development potential. The Lake Chad region comprises and fishing activities. Agriculture and fishing generate a set of administrative areas across Cameroon, Chad, significant indirect employment in related activities such Niger, and Nigeria that surround the lake (Map 2.1). as food processing, trade, transport, and production of Throughout this note, the boundaries of the Lake Chad crafts. These economic activities contribute to the food region are defined by ten different regions/states: the and job security of people in its hinterlands, including its Chari Baguirmi, Kenam, Lac, and Hadjer Lamis regions two regional metropolises with seven-figure populations: in Chad; the Diffa and Zinder regions in Niger; the Far N’Djaména, the Chadian capital, and Maiduguri, the North region in Cameroon; and Borno, Adamawa, and capital of the Nigerian State of Borno.207 Linkages with Yobe states in Nigeria (Map 2.2).205 neighboring areas suggest that the livelihoods of as many Map 2.1: Map of the Lake Chad  dministrative definition of the Lake Chad Map 2.2: A region N’guigmi C H A D Northern basin N I G E R r r rie Bol Ba at re Southern basin G N I G E R I A Guitté CAMEROON Chari-Logone River N’Djaména Source: Magrin, Lemoalle, Pourtier, 2015. Atlas du lac Tchad. Source: World Bank’s Lake Chad Region Recovery and Development Project (PROLAC). Source: Authors’ elaboration. The shapefiles of the boundaries of the subnational areas are taken from GADM3.6. 205 The proposed administrative definition for the Lake Chad region is consistent with the one used in the World Bank’s Lake Chad Region Recovery and Development Project (PROLAC). 206 Authors’ calculations based on remote-sensing data from WorldPop (2020) (database) and using the proposed administrative definition of the Lake Chad region for this activity. The population estimate was calibrated using the latest census data available from: Thomas Brinkhoff: City Population, http://www. citypopulation.de. 207 See Magrin and Perouse (2018). 2.2 Recent Socio-economic Trends in the Lake Chad 87 Lake Chad Regional Economic Memorandum  |  Development for Peace as 50 million people are linked to economic activities  overty is more severe in the Lake Chad Figure 2.2: P basin vis-à-vis in the rest of the countries based on the Lake Chad region.208 Poverty headcount rate (US$1.90 per day), percent 100– The region’s population is growing at a rapid pace. 90– High fertility rates are the main factor explaining high 80– population growth levels in Cameroon, Chad, Niger, 70– 72.3 and Nigeria. In the four countries, the population could 60– double in the next 25 years. In the Lake Chad region 58.8 50– 51.6 itself fertility rates are very high, which means that the 40– 45.4 43.4 39.8 39.1 population could grow faster than other parts of the 38.1 37.7 30– 30.7 countries it belongs to. The population in the Lake Chad 20– 26.0 18.8 region increased by almost 4 million people between 10– 2000 and 2015. The total fertility rate (TFR) in the 0– Cameroon Chad Niger Nigeria region is higher in the administrative areas of Cameroon J Country J Lake Chad J Rest of country (estimated at 6.8 children per women) and Nigeria Source: Data on poverty based on the latest household surveys conducted in Cameroon (2014), Chad (2011), Niger (2014) and Nigeria (2018). (5.8 children per women) compared to other parts of Notes: Rest of country = outside the Lake Chad region; Lake Chad = within the Lake Chad Basin region. the countries (averaging 4.8 children per women in the rest of Cameroon, and 5.3 children per women in the country’s Lake Chad region (31 percent) is lower than rest of Nigeria). In Niger’s Diffa and Zinder regions, the rest of the country (40 percent). This is explained by fertility rates are also high at 8.2 children per woman the fact that the areas around the lake in Chad lie near (7.5 in other parts of the country). In the Lac and Hadjer the capital of the country, with a consequently higher Lamis regions in Chad, the average fertility rate is high urbanization rate and a relatively high population density. at 6.2 children per woman, albeit slightly lower than the average of 6.5 children per woman in other parts of the There is also a significant spatial gap in poverty within country. the Lake Chad region. Poverty is most prevalent in the parts of the Lake Chad region that lie within the boundary of Nigeria. The poverty rates in Adamawa and 2.2.1 Trends in poverty reduction Yobe States reach as high as 74 percent and 70 percent, which are significantly higher than the national average Given its high poverty rate, low human capital and of 38 percent (Map 2.3, Panel B). These regions also poor access to key services, the Lake Chad region is are home to the largest number of the poor in the Lake characterized as a lagging region. An analysis of the Chad region (Map 2.3, Panel C). On the other hand, most recent household surveys available for each country Kanem Region in Chad has the lowest poverty rate (of 19 shows that households in the Lake Chad region are poorer percent) across the Lake Chad region. compared to households in neighboring regions (Figure 2.2). The regional poverty rate in the Extreme North Not only is the level of poverty relatively high in region of Cameroon (59 percent) is three times higher the Lake Chad region versus that in the rest of the that of the rest of the country (19 percent). In Nigeria, the countries but the pace of poverty reduction in the Lake Chad region has a poverty rate (72 percent) nearly region is also slow. In fact, there is little sign that the twice as high as in the rest of the country (38 percent). spatial gap in poverty between the Lake Chad region Chad is the only exception, where the poverty rate in the and neighboring regions is narrowing over time. In 208 Adelphi (2019). 88 2.2 Recent Socio-economic Trends in the Lake Chad Technical Paper 1. Socioeconomic Trends in the Lake Chad Region Map 2.3: Poverty in the Lake Chad region  overty rates in Chad, a. P  overty rates in the Lake Chad b. P  umber of poor in the Lake c. N Cameroon, Niger, and Nigeria region Chad region Source: Based on the latest household survey in Chad (2011), Cameroon (2014), Niger (2014) and Nigeria (2018). Notes: Poverty rates calculated at USD 1.90 per day (2011 PPP). The Borno state is excluded given that there is no representative household survey for that state due to security reasons. Cameroon, for instance, poverty declined by 4 percent  sset deprivation in the Lake Chad basin Figure 2.3: A versus the rest of the countries in the Extreme-North region compared to a 6 percent decrease in the rest of the country between 2007 and Percent of bottom 40 percent in the wealth index 100– 2014. While Niger as a whole saw a reduction in poverty 90– from 51 percent to 45 percent between 2011 and 2014, its 80– 83.3 82.4 Lake Chad region experienced a slight increase in poverty 70– from 48 percent to 52 percent. This pattern of non- 60– 64.1 convergence in welfare is also corroborated considering 50– 53.3 52.6 household assets across different regions in each country. 40– 46.7 44.2 41.9 The percent of households that are relatively asset-poor— 30– 35.7 38.2 37.9 34.1 31.3 31.9 or in the bottom 40 percentile of asset wealth distribution 20– in a given survey year/country209—shows no clear sign 10– of convergence between the Lake Chad and non-Lake 0– Chad areas of each country (Figure 2.3). For instance, in CMR '11 CMR '18 NER '06 NER '12 NGA '13 NGA '18 TCD '14 J Lake Chad J Rest of country Niger and Nigeria, the percent of asset-poor households Source: Based on the latest two rounds of Demographic and Health Survey (DHS) in each increased in the Lake Chad region over time, while in country. Notes: The x-axis in the figure includes country names (CMR=Cameroon; NER=Niger; NGA=Nigeria; TCD=Chad) as well as year in which DHS was conducted. Cameroon, this share remained almost unchanged between the two latest rounds of DHS. These findings suggest that wealth gaps between the Lake Chad region and the rest of the countries may have worsened over time. 209 We constructed the wealth index for the latest two DHS surveys in each of the LCB countries. Our wealth index is a composite measure of various household assets, including housing materials, access to electricity, cooking fuel, access to improved water, as well as ownership of various items such as televisions and bicycles. We applied a principal component analysis to generate the composite index. To make our wealth index comparable over time within the same country, we applied the same coefficients for use as weights across the two latest surveys. “Asset-poor” households refer to those households whose wealth index score is in the bottom 40 percent of distribution for the given survey country/year. 2.2 Recent Socio-economic Trends in the Lake Chad 89 Lake Chad Regional Economic Memorandum  |  Development for Peace 2.2.2 Local economic dynamics 2.2.3 Human capital outcomes and access to basic services In addition to lagging behind in terms of core poverty and socioeconomic indicators, the Lake Chad region In addition to monetary poverty indicators, the has experienced little economic progress over the past Lake Chad region lags in terms of key human capital three decades. An analysis of local economic growth indicators. The literacy rate (15 years-old and older) and based on nighttime light intensity—which serves as a the completion rate for primary education (ages 14–25) are useful proxy for capturing both the size of local economic significantly lower in the Lake Chad region compared to activities and its change over time210—shows that the the national average (Figure 2.4). Child health conditions intensity of nighttime light is strongly correlated with the in the region are also grim. For instance, child stunting distribution of people and economic activities (Map 2.4, in the Lake Chad region is roughly 10–15 percent higher Panel A). Overall, nighttime light grew faster in areas that in the Lake Chad compared to the rest of the countries appear to be more densely populated or characterized (see Table A2.1). These results, based on original analysis by higher levels of economic activities (as indicated by performed for this report, are consistent with other nighttime light)—particularly in the Northwest and studies. For example, according to the International Southern parts of Nigeria (Map 2.4, Panel B). Crisis Group (ICG), the gross school enrolment rate in the Chadian lake area is below 30 percent, and, Seen from space, the regions near Lake Chad in ‘community teachers’ (largely pupils’ parents) generally Cameroon and Nigeria exhibits a relatively low level stand-in in place of trained teachers.211 In the Chadian of luminosity and lower rates of growth. The gaps in part of the lake, there is only one doctor for every 140,000 the intensity of nighttime light between the Lake Chad inhabitants, that is, a quarter of the national average.212 region and the rest of the countries are particularly stark The low access and quality of education, health and other in Cameroon and Nigeria whereas in Chad and Niger, services in the region has been aggravated by the erosion the average intensity of nighttime light is slightly higher of infrastructure, both public and private, resulting from in the Lake Chad region (Map 2.4, Panel C). The annual the conflict.213 rate of growth in nighttime light are also slower in the Lake Chad region compared to the rest of the countries Access to core public services in the Lake Chad basin in Cameroon and Nigeria. This implies that there has is also lower than in the rest of the countries.. For been no substantial regional growth in the areas around instance, the national average rate of access to electricity the lake. In Nigeria, in particular, increases in intensity in Cameroon, Chad, Niger, and Nigeria is 62 percent, of nighttime light between 1992 and 2013 are much 8 percent, 14 percent and 59 percent, compared to an slower in the Lake Chad region compared to the rest of estimated 20 percent, 2 percent, 10 percent and 38 percent the country (Map 2.4, Panel C and D). in the Lake Chad region, respectively. In terms of access to improved water and sanitation facilities, the Lake Chad areas of Cameroon and Chad suffer from lower levels of access compared to the rest of the country (Figure 2.5). 210 The analysis of nighttime light relies on Defense Meteorological Satellite Program (DMSP) OLS data that are inter-calibrated by Li et al. (2020), allowing for a better comparison over time. The intensity of nighttime light is measured in a digital number (DN) ranging from 0 and 63 that represents an average of lights in all nights after sunlight, moonlight, aurorae, forest fires, and clouds have been removed algorithmically, leaving mostly human settlements. 211 ICG (2017). 212 Ibid. The figure for Chad is much lower than in Cameroon’s Far North region (1/52,000) or Niger’s Diffa region (1/24,500), both close to the lake. 213 It is reported that, during the armed group’s eight-year rebellion, almost 1,400 schools have been destroyed in Borno and more than 57 percent of schools are unable to open because of damage or being in areas that remain unsafe (Al Jazeera 2017). Regarding health facilities: “Insurgents have destroyed about 788 health facilities in the region. In Borno 48 health workers have been killed and over 250 injured. The state has lost up to 40 percent of its facilities and only a third of those left in Borno state remain functional,” (Obi and Eboreime 2017). 90 2.2 Recent Socio-economic Trends in the Lake Chad Technical Paper 1. Socioeconomic Trends in the Lake Chad Region This gap is particularly pronounced in Cameroon, where There are also signs of a widening gap between the the rate of access to improved water and sanitation in the Lake Chad and the rest of the surrounding countries in Far North region is, on average, about 36 percent lower terms of access to improved sanitation and electricity. than the rest of the country. In Cameroon, the share of households with access to improved sanitation in the Lake Chad region declined  idening gap in economic dynamics between the Lake Chad region and other parts of the countries, Map 2.4: W 1992–2013 a. Nighttime light intensity, 2013  nnual rate of change in nighttime lights, 1992– b. A 2013  ighttime light intensity in comparison between c. N  hange in nighttime light in comparison between d. C the Lake Chad region and other parts of the the Lake Chad region and other parts of the countries countries Annual rate of change, 1992–2013, percent 3– 3– 2.87 2– 2– 1.50 1.38 1– 1– 0.99 0.90 0.87 0.80 0.79 0.78 0.70 0.60 0.63 0.60 0.59 0.43 0.37 0– 0– Cameroon Niger Nigeria Chad Cameroon Niger Nigeria Chad J Lake Chad J Rest of country J Lake Chad J Rest of country Source: Panel A shows the mean of nighttime luminosity for 2013 based on stable DMSP inter-calibrated NTL data (Li et al. 2020) at a spatial resolution of 10km while Panel B shows the annual rate of growth in the mean of nighttime luminosity between 1992 and 2013 in percent. Panel C and D show the mean of nighttime light luminosity and annual rate of change in nighttime light luminosity (in percent) during the same time period. The calculation for Panel C and D was performed only on a subset of grid cells that are lit (with a positive value in digital number at some point between 1992 and 2013) thus excluding areas that are largely rural and unpopulated. 2.2 Recent Socio-economic Trends in the Lake Chad 91 Lake Chad Regional Economic Memorandum  |  Development for Peace  he literacy and completion rates for primary education in the Lake Chad are significantly lower Figure 2.4: T than in the rest of the countries a. Literacy rate (+15) b. Completion rate, primary education, 14–25 Percent Percent 100– 100– 90– 90– 80– 83.7 80– 77.8 79.0 78.2 79.0 70– 71.9 72.4 70– 72.3 60– 60– 58.7 60.0 50– 50– 47.1 45.5 40– 41.6 40– 39.4 30– 33.0 34.0 32.8 30– 32.8 29.5 27.8 20– 20– 25.4 21.0 17.4 10– 10– 12.2 0– 0– Cameroon Chad Niger Nigeria Cameroon Chad Niger Nigeria J Country J Lake Chad J Rest of country J Country J Lake Chad J Rest of country Sources: Based on the most recent household surveys conducted in Cameroon (2014), Chad (2018), Niger (2018) and Nigeria (2018). Figure 2.5: Access to core public services in the Lake Chad region across time a. Access to improved water b. Access to improved sanitation c. Access to electricity Percent Percent Percent 100– 100– 100– 90– 90– 90– 80– 80– 80– 84.8 79.2 76.8 76.3 70– 70– 70– 74.8 74.1 74.0 70.4 68.8 68.4 66.9 66.7 65.5 65.1 60– 60– 60– 62.9 61.9 61.3 57.8 56.6 56.4 55.1 50– 50– 50– 48.9 40– 40– 40– 44.5 44.5 41.7 38.4 30– 30– 30– 30.8 30.2 28.5 20– 20– 20– 23.1 19.6 16.2 16.0 10– 10– 10– 14.3 13.7 12.8 10.0 10.4 2.4 9.0 6.0 5.7 0– 0– 0– '11 '18 6 '12 '13 '18 '14 '11 '18 6 '12 '13 '18 '14 '11 '18 6 '12 '13 '18 '14 '0 '0 '0 R R R R R R A A A D D D A A A R R R R R R CM CM CM NG NG NG NE NE NE NG NG NG CM CM CM TC TC TC NE NE NE J Lake Chad J Rest of country J Lake Chad J Rest of country J Lake Chad J Rest of country Source: Based on the latest two rounds of Demographic and Health Survey (DHS) in each country. Notes: The graph shows the percentage of households with access to improved water and sanitation and electricity from left to right. Data on access to these core public services are taken from the two latest DHS in each country. The x-axis in the figure includes country names (CMR=Cameroon; NER=Niger; NGA=Nigeria; TCD=Chad) as well as year in which DHS was conducted. from 42 percent to 31 percent between 2011 and 2018 66 percent to 56 percent). Progress in expanding access whereas the rest of the country experienced a modest to electricity in the Lake Chad region has also stagnated. improvement (from 62 percent to 67 percent) over the In Niger, the regions of Diffa and Zinder saw access to same period. A similar divergence pattern emerges in electricity improve by four percentage points (from Nigeria, where access to improved sanitation in the 6 to 10 percent) between 2006 and 2012, a slightly Lake Chad area decreased from 74 percent to 55 percent lower increase than in the rest of the country (where between 2013 and 2018—a much faster rate of decline access improved by nearly six percentage points, from than in the rest of the country (where access fell from 10.4 percent to 16 percent) (Figure 2.5). 92 2.2 Recent Socio-economic Trends in the Lake Chad Technical Paper 1. Socioeconomic Trends in the Lake Chad Region 2.3 Spatial Dynamics of Economic Development within the Lake Chad Region 2.3.1 Density 2019).216 In Cameroon, most of the economic activities in the Lake Chad region are clustered in the southern The perpetuation of poverty and slow growth in the part of Extrême—Nord, particularly around the city Lake Chad region is closely linked to its economic of Maroua.217 In Niger and Chad, while the volume of geography (WDR 2009). Indeed, the combination of economic activities in the Lake Chad areas is small, they low economic density as well as high distance and division tend to also cluster around the border areas neighboring coalesces to detail the region from its sustainable growth Nigeria and Cameroon. path. Local agricultural systems have been also disrupted The region is characterized by a low economic density following the destruction of farming and irrigation and a lack of agglomeration economies that hinders facilities. Conflict and violence that escalated since it from reaching its potential. While the Lake Chad 2009 have generated losses equivalent to an estimated region accounts for 17 percent of total areas of the four USD 3.7 billion, including profound loss of fisheries, neighboring countries combined, its economy makes up livestock, and agricultural production, ruined irrigation only 5 percent of GDP.214 Most economic activities in the and farming facilities, and the collapse of extension region are highly spatially concentrated in a few large and services.218 Insecurity is exacerbating the existing secondary cities. Beyond N’Djaména and Maiduguri— challenges faced by vulnerable farmers who were already only two metropolitan cities with a population of facing devastating natural hazards including cyclical over 1 million—there are several secondary cities that droughts and floods. The concentration of herds due support the economy of the region, including Damaturu to changes in transhumance flows as a result of these (Nigeria), Jimeta (Nigeria), Mubi (Nigeria), Maroua climatic conditions is putting pastoral resources under (Cameroon), Zinder (Niger) among others. Estimates of extreme pressure and threatening animal health. Gross National Product (GDP) for the Lake Chad region show that economic activity is spatially concentrated in these few urban centers which are clustered around the 2.3.2 Distance national borders of the neighboring countries (Map 2.5). While leading areas tend to have a high economic Most of the trade between Nigeria and Cameroon density, lagging areas tend to present a long distance- flew through the corridor connecting Maiduguri to-density ratio. Distance refers to the ease or difficulty to Kousseri or Maroua although this trade flow has for goods, services, labor, capital, information and ideas been significantly disrupted by the intensification of to traverse space. It measures how easily capital flows, the Boko Haram conflicts since 2009215 (and more labor moves, goods are transported, and services are recently since Nigeria closed its land borders in August delivered between two locations. In this sense, distance 214 Calculated based on Ghosh et al. (2010). 215 In a study conducted by WFP (2016), only one out of 26 transporters surveyed in Cameroon reported Nigeria as its main supply source of cereals suggesting that agricultural trade flows between Northeast Nigeria and North Cameroon have been reduced. 216 The Nigerian government announced on August 22, 2019 the partial closure of three border sites with Benin and Cameroon. The closure was officially extended to all land border crossings in September 2019. 217 See UNHCR and World Bank (2016). 218 World Bank (2020), PROLAC Program Appraisal Document, Report PCBASIC0089548. 2.3 Spatial Dynamics of Economic Development within the Lake Chad Region 93 Lake Chad Regional Economic Memorandum  |  Development for Peace Map 2.5: Economic activity in the Lake Chad region, 2006  stimated subnational real GDP (2006) in a. E  stimated subnational real GDP (2006) in the Lake b. E Cameroon, Chad, Niger, and Nigeria Chad region Source: Authors’ calculations based on data from Ghosh et al. (2010). is an economic concept not just a physical one, related characterized by slower growth in economic activities. to connectivity and access. An area is more likely to be Poor road infrastructure undermines intra- and inter- lagging the farther away it is from leading areas since regional connectivity across the region, limiting economic greater distance-to density implies a lack of integration opportunities for people. This is further compounded into the economy. It also implies poorer access to the by rising transportation costs in water transportation, “thick” markets of capital, labor, goods, services and partly driven by government interventions limiting ideas, and the spillovers of knowledge and information boat circulation in the lake, and the drying of the lake. they provide. Conversely, areas proximate to population centers tend to have greater market accessibility due to a wider and denser road network system, allowing people living near 2.3.2.1 Physical Infrastructure a city to benefit from its agglomeration economies. Access to connective infrastructure across the Lake Rural people in the Lake Chad region are twice as likely Chad region is very limited, particularly among the to be disconnected from all-seasons roads (motorable rural population. A large swath of rural areas in the Lake all year round), compared to areas in the rest of the Chad region still have limited access to large markets. countries. The Market Access Index–a measure of the These areas—with a long distance-to density ratio—are size of population that can be reached within a certain 94 2.3 Spatial Dynamics of Economic Development within the Lake Chad Region Technical Paper 1. Socioeconomic Trends in the Lake Chad Region Map 2.6: Market and rural access in and around the Lake Chad region a. Market Access Index b. Rural Access Index Source: Calculated by the authors on various geospatial data sources (see Appendix 2.C and 2.D for further details). Notes: The map on the left shows areas that score “very high” (top 20 percentile), “high” (20–40 percentiles), “medium” (40–60 percentiles), “low” (60–80 percentile), and “very low” (bottom 20 percentile) in the Market Access Index. See Appendix 2.C for further details on how the Market Access Index is calculated. The map on the right shows the Rural Access Index or share of rural population who live within 2 km away from all-season roads as proxied by OpenStreetMap. See Appendix 2.D for details on the construction of the Rural Access Index. travel time–is relatively high in the Lake Chad region 2.3.2.2 Digital Infrastructure compared with some other parts of the countries (Map 2.6, Panel A). This indicates that, with proper connective Poor road connectivity in the Lake Chad region is infrastructure, people in the Lake Chad region could further complicated by a lack of digital infrastructure. benefit from economic opportunities that large markets— Cell phone ownership as a share of population in the both within and around the region—can offer. However, areas bordering the lake in Niger stood at 13.4 percent the score of the Rural Access Index—i.e. the share of rural compared to 20 percent for the rest of the country. A population living within 2 km away from an all-season similar pattern can be seen in Nigeria, where cell phone road219—is low for the region (Map 2.6, Panel B). Nearly ownership is 4.7 percentage points lower in the regions two thirds (about 60 percent) of the rural population in bordering the lake. Chad is the only exception, where the Lake Chad region live farther than 2 km away from ownership is higher in areas bordering the lake compared an all-season road (proxied by OpenStreetMap), that is, to the rest of the country (20 versus 15.6 percent). about twice the share in the non-lake parts of the basin However, the share of people who are connected is still countries (about 30 percent). low compared to international standards, translating to roughly 80 percent of the population being disconnected from digital technologies. 219 An “all-season road” is defined as a road that is motorable all year round by the prevailing means of rural transport. Trunk, primary, secondary, and tertiary roads in OpenStreetMap are used as a proxy for all-season roads following the methodology by Azavea: https://rai.azavea.com/. 2.3 Spatial Dynamics of Economic Development within the Lake Chad Region 95 Lake Chad Regional Economic Memorandum  |  Development for Peace The lack of digital infrastructure and low levels population, above the regional Sub-Sahara African of cellphone ownership hinder the adoption and average of 27.7 percent. However, the share of unique usage of novel digital technologies. While digital mobile internet subscribers in the Lake Chad countries infrastructure in Sub-Saharan Africa as a whole is remains substantially below regional leaders, such as lagging behind compared to the rest of the world,220 South Africa (52 percent). Chad registered a unique the Lake Chad countries have a particularly low level of mobile internet subscription rate of 17.1 percent of the internet penetration. Approximately 11.8 percent of the population in 2020, compared with 34 percent in Nigeria population in countries bordering Lake Chad reported and 33.8 percent in Cameroon. On the other hand, Niger using the internet, compared to 18.7 percent across Sub- has the lowest mobile internet penetration rate across the Saharan Africa, on average.221 There is heterogeneity Lake Chad region, and among the lowest in Sub-Saharan within the region. Chad lies among the countries with Africa. It is important to identify the main constraints the lowest internet penetration rates in the world, at to adopt internet services faced by individuals to fully 6.5 percent of the population, compared to Cameroon, harness the potential benefits of digital technologies in which at 23.2 percent ranks above the regional average. the region. A similar pattern can be seen regarding mobile internet penetration—measured by unique mobile internet The Lake Chad region is largely disconnected from subscribers222—which is the main mechanism of access to the digital world.223 A large swath of areas in the Lake the internet in Sub-Saharan Africa (as opposed to fixed- Chad region has very little connectivity to fiber optics broadband subscriptions). transmission nodes224 or 3G technology225 except for the Extreme North region of Cameroon where there appears Mobile internet in the LCB countries has undergone to be more comprehensive coverage (Map 2.7).226 Access a rapid expansion, although its pace still lags regional to the internet (either through fixed broadband or mobile leaders like South Africa. Unique mobile internet broadband) can serve as a catalyst for poverty alleviation,227 subscribers across the Lake Chad region increased almost improved labor outcomes228 and the functioning of rural twofold as a share of the population between 2014 and markets,229 specifically regarding price information, 2020. In 2020, this figure stood at 31.1 percent of the access to inputs and consumers230 and access to capital 220 World Bank (2019). 221 Data of 2017, WDI (World Development Indicators) (database) (accessed on 04/07/2021), World Bank, Washington, DC, https://datatopics.worldbank.org/ world-development-indicators/. Internet users are individuals who have used the internet (from any location) in the last 3 months. The internet can be used via a computer, mobile phone, personal digital assistant, games machine, digital TV etc. 222 Given that consumers may use multiple SIM cards to take advantage of discounts or to avoid high charges for off-network calls, market penetration in terms of unique subscribers may provide a better picture of the degree of access to mobile services. GSMA defines mobile internet as the use of internet services by unique users on mobile devices at the end of a given period. Mobile internet services are defined as any activity that uses mobile data (that is, excluding SMS, multimedia messaging services, and cellular voice calls). See GSMA Intelligence (database), Global System for Mobile Communications (GSM Association), London, https://www.gsmaintelligence.com/. Accessed on April 7 2020. 223 Hjort and Poulsen (2019). 224 These nodes correspond to add or drop points (entrance or exit) in the long-haul fiber networks. It is useful to think of long-haul fiber networks as motorways that have junctions (on and off ramps that is, add and drop points) that feed smaller class roads (access fiber, wireline, and wireless networks). In the motorway scenario, even if a household is located close to the motorway, it may be a long drive to the nearest junction. The same applies to fiber-optic networks, in which the speed of fixed broadband Internet is determined by proximity to the transmission nodes rather than the network lines connecting the nodes. 225 While second-generation (2G) technologies enable voice, SMS, and limited Internet access, third-generation (3G) technologies enable more rapid Internet browsing and data downloading. 226 The 2G/3G coverage data should be treated with caution, however, because the Collins Bartholomew coverage maps do not necessarily include all network providers in each country and comparing coverage across these countries may be problematic due to uneven reporting of coverage by country. Overall, this coverage map should be treated as a lower bound of 2G/3G availability. 227 See Bahia et al. (2019) and Masaki et al. (2020). 228 See Hjort and Poulsen (2019); Paunov and Rollo (2014); Fernandes et al. (2019); Chun and Tang (2018); Viollaz and Winkler (2020). 229 See Kaila and Tarp (2019); Goyal (2010); Ritter and Guerrero (2014); Salas-Garcia and Fan (2015). 230 See Aker and Mbiti (2010); Aker (2011); Debo and Van Ryzin (2013). 96 2.3 Spatial Dynamics of Economic Development within the Lake Chad Region Technical Paper 1. Socioeconomic Trends in the Lake Chad Region markets.231 Poor digital connectivity around Lake Chad is trade and the flow of factors of production. Interstate thus another important source of distance that prevents conflict creates the thickest borders. While borders in the region from tapping its full economic potential. the rich world have become increasingly thin, hereby facilitating trade and the movement of people and capital, Map 2.7:  Digital connectivity in and around the Lake borders in many developing countries remain thick, as Chad region (2018–2019) is generally the case in the Lake Chad region. At the same time, borders in the areas around Lake Chad have historically been characterized as relatively porous—with trade and social ties permeating borders. This mobility, however, has subsided over the last decade with the hardening of borders and counterinsurgency measures as a response to the Boko Haram insurgency. 2.3.3.1 Ethnolinguistic and religious divisions The roots of conflict and intercommunal violence in the Lake Chad region are found in multi-faceted factors ranging from competition over resources, local tensions and differences between ethnic groups in the Lake Chad region, as well as climate shocks. As shown by the ethno-linguistic boundaries mapped Source: Data on 2G/3G coverage maps from Collins Bartholomew coverage maps; data on the locations of operational fiber optics nodes from Africa Bandwidth Maps http://www. below (Map 2.8, Panel A) the Lake Chad region presents africabandwidthmaps.com/. Notes: Mobile coverage corresponds to 2018. Fiber optics correspond to 2019. The 2G/3G a wide ethnic heterogeneity.233 The basin has also been coverage data should be treated with caution, however, because the Collins Bartholomew coverage maps do not necessarily include all network providers in each country. Thus, this coverage map should be treated as a lower bound of 2G/3G availability. subject to tensions stemming from wide religious diversity. Different religious groups concentrate in the area, ranging from the predominantly Muslim-majority 2.3.3 Division northern Basin in Niger, to a less concentrated Muslim majority in northeast Nigeria and Kanem in Chad, to Together with density and distance, the third important a Christian majority in the Extreme–North region of geographic dimension for territorial development is Cameroon and the Hadjer Lamis and Chari Baguirmi division. It applies at both national and international regions in Chad (Map 2.8, Panel B). Some of the Boko scales. At the national scale, nations can be internally Haram violent activities may be linked with pre-existing divided across linguistic, ethnic, religious, and/or cultural tensions between specific local communities or ethnic lines. At the international level, divisions mainly arise groups.234 What may be referred to as ‘another attack by from so-called thick borders, i.e., the many restrictions Boko Haram’ in Lake Chad in the international media, countries impose on other countries regarding the flow may instead be a set of reprisals among the Kuri, livestock of goods, capital, people and ideas.232 Thick borders limit farmers, and fishermen.235 Changes in land use and 231 See Hasbi and Dubus (2019); Alibhai et al. (2018). 232 Fratianni and Kang (2006). 233 Vedeld (1999). 234 Cohen (2015). 235 Ibid. 2.3 Spatial Dynamics of Economic Development within the Lake Chad Region 97 Lake Chad Regional Economic Memorandum  |  Development for Peace Map 2.8: Ethnolinguistic and Religious Groups in the Lake Chad region A. Ethno-linguistic boundaries B. Religious boundaries Source: Elaboration based on Ethnicity Felix 2001 and World Religion Map from Harvard Worldmap. water resource management since the 1980s, coupled has been aligned with the Islamic State of Iraq and the with ineffectiveness in the fragile local institutions have Levant since 2015.239 It is not a unified group; in 2016 led to heightened competition among farmers, herders it split into two factions: the Islamic State’s West Africa and fishers along social fault-lines of ethnicity, religion, Province (ISWAP) and Jama’atu Ahl al-Sunnah lil-Dawa gender, and class.236,237 wal-Jihad (JAS).240 At its peak—that is between 2010 and 2015—the group seized a large swath of territories in Northeastern Nigeria, including major cities, pushing the 2.3.3.2 Boko Haram Conflicts Government of Nigeria to declare a state of emergency, (an action that was later followed by other governments The intensification of conflict in the Lake Chad region in the region). While most of the attacks between 2009 since the rise of Boko Haram in 2009 has been a large and 2013 were geographically concentrated in a few states source of division driving the leggedness of the region. in the Northeastern corner of Nigeria, the terrorist group While the group was fist founded in 2002, the insurgency moved some of its activities to the neighboring areas of is considered to have begun in full in 2009 in Nigeria Cameroon, Chad and Niger (Map 2.9). Vigilante groups with bases in neighboring countries. In 2014–2015, it have been created in response to the insurgency, which expanded into northern Cameroon, Niger and Chad. are becoming increasingly violent. Since then, the group has retreated into inaccessible areas, mainly along the borders, but has continued to carry out The ethnolinguistic divisions exacerbated with the more frequent and sophisticated attacks.238 Boko Haram crisis are associated with social exclusion. For example, 236 Onuoha (2010), Béné et al. (2003), Ahmadu (2011). 237 Smith and Walters (2017). 238 Magrin and Perouse de Montclos (2018). 239 Vivekananda et al. (2019). 240 The indiscriminate targeting of civilians appears to have been a major point of disagreement. The extremist group ISWAP avoids harming civilians, focusing mainly on military and government targets (Samuel 2019). 98 2.3 Spatial Dynamics of Economic Development within the Lake Chad Region Technical Paper 1. Socioeconomic Trends in the Lake Chad Region  he evolution of the number of Boko Haram violent events from 2009–2020 Map 2.9: T Sources: Blankespoor (2021). technical paper for this report. The elaboration is based on ACLED (Armed Conflict Location and Event Data Project) (dashboard), Robert S. Strauss Center for International Security and Law, Austin, TX, http://www.acleddata.com/. Kanuris, who are viewed as Boko Haram sympathizers, 2.3.3.3 Cross-border trade have increasingly become the subject of discrimination in Cameroon’s Far North region. Anecdotal evidence Cross-border trade has declined as a result of the suggests that the ethnic group is targeted both by security insurgence. Conflict has affected production and trade forces as well as by local militias groups, or vigilantes, for directly, as well as indirectly, through the counter- their perceived connection to the Boko Haram insurgency. insurgency measures that restrict movement and put In turn, this has been associated with social exclusion and bans on farming and trade.241 Transit flows of livestock marginalization of the group, affecting individuals’ ability seem to be declining. The transit of cattle to Nigeria from to participate in economic activities. On the other hand, Chad and Cameroon decreased by 39 percent between the conflict has also had an impact on social cohesion 2015 and 2016–2017, as shown by customs data from within communities. For example, individuals forcibly the Yagoua livestock crossing point.242 The deteriorating displaced in host communities in Cameroon live together impact that the conflict has had on Chad’s livestock but compete for the same services, which are already exports to Nigeria is even more poignant considering that strained in terms of both access and quality. The conflict these exports are the country’s second source of foreign has also led to distrust among communities in connection revenue after oil.243 In Cameroon, estimates suggest with former fighting. that Boko Haram has stolen USD 6 million dollars’ worth of cattle, sheep and goats since 2013.244 Market infrastructure has also been subject to physical damage, e.g. in Damaturu in Yobe, Nigeria, over 650 shops have been reported as damaged.245 241 UNDP, OCHA (2016). 242 World Bank (2018). 243 World Bank (2015). 244 World Bank (2018). 245 Mercy Corps et al. (2017). 2.3 Spatial Dynamics of Economic Development within the Lake Chad Region 99 Lake Chad Regional Economic Memorandum  |  Development for Peace 2.4 Determinants of regional growth There is a multitude of structural factors and aggregate be statistically significant and positive (or negative), it shocks that could affect a trajectory of territorial shows that areas in the Lake Chad region experienced development in the Lake Chad region. As discussed faster (or slower) growth in nighttime light vis-à-vis earlier, not only does it fall behind in terms of its core other parts of the countries surrounding the lake. Xjrct poverty and socio-economic indicators compared to other contains a wide range of conditional variables that seek parts of the countries, the Lake Chad region also has seen to capture various potential factors that could affect the very little economic progress over the past three decades trajectory of local economic growth. More specifically, it (see Section 2.2). This section seeks to identify these includes population, weather shocks, conflicts, market factors that might be associated with the dynamics of accessibility, as well as various other geographical economic activity across the region using an econometric features, including the density of cropland (as a proxy for approach. agricultural production) and grazing land, access to on- grid electricity, elevation and terrain constraints.246 Lastly, The analysis employs multivariate regression to tease the model controls for region (or ADM1-level) fixed out which of these factors contribute to regional effects Cr to ensure that our results are not confounded by growth. More formally, following Barro and Sala- time-invariant regional characteristics as well as longitude I-Martin’s and Bairro et al.’s (1995), the following and latitude of each grid to account for other geographical convergence model is estimated: confounders. Gjrct1-t0 = α + β1 ln (Lightjrct0) + β2 Lake Chad regionjrc T  o better understand which factors contribute most + β3Xjrct + Cr + εjrct (1) significantly as drivers for local economic growth, we report the results of a Shapley decomposition from those  here Gjrct1-t0 corresponds to annual growth rate in w regressions. nighttime light intensity between year t0 and year t1 in 0.1 degree (≈11 km × 11 km) grid cell j in region r in country c. More formally: 2.4.1 Data Gjrct1-t0 = ln (Lightjrct1/Lightjrct0) / (t1-t0) (2) Given the dearth of data, we turn to various geospatial datasets to capture a multitude of structural factors that may drive local economic growth. Our full sample  here Lightjrct0 denote the initial level of nighttime light w consists of 32,097 0.1×0.1 grid cells defined over the at t0 and β1 essentially captures the rate of convergence four countries of Cameroon, Chad, Niger and Nigeria. after accounting for other confounding factors Xjrct We then look at a subsample of cells that exhibited some There is conditional convergence (divergence) if β1 level of luminosity at some point between 1992 and is statistically significant and negative (positive). Also 2013—the period for which nighttime light data are included in this regression is a dummy variable coded 1 available and comparable over time—and we refer to this if a given grid cell lies within the boundary of the Lake subsample as an “intensive” margin sample. The intensive Chad region, 0 otherwise. In essence, if β2 turns out to margin sample can be considered as a set of cells that 246 See Appendix 2.B for descriptive statistics of variables used in the regression. 100 2.4 Determinants of regional growth Technical Paper 1. Socioeconomic Trends in the Lake Chad Region exclude uninhabited areas or largely rural areas with no by Vicente-Serrano et al. (2010). SPEI takes into account luminosity observed from space. The intensive margin both precipitation and potential evapotranspiration or sample includes 5,211 cells. the ability of the soil to retain water, which depends on temperature, latitude, sunshine exposure, and wind Nighttime light speed. SPEI data are drawn from the Global SPEI O  ur proxy measure for local economic growth comes Database (https://spei.csic.es/). SPEI is expressed in units from satellite data on light emitted into space at night. of standard deviation from the cell’s historical average and Meteorological Satellite Program (DMSP) have been thus has mean 0 by construction in the historical sample. recording data on lights at night using their Operational Linescan System (OLS) sensor since the mid-1960s, with S  ince the impact of rainfall on local economy—and a global digital archive beginning in 1992. DMSP-OLS agriculture for that matter as the entire economy of the nighttime light data are available on an annual basis for region relies heavily on the agricultural sector—is most the period of 1992–2013 and at a resolution of roughly likely much larger during the growing season, we consider 1km (30-arcsecond pixel). The strength of luminosity at SPEI specifically for growing periods in each cell. To night is measured by a digital number (DN), an integer identify a cell-specific growing period, we follow Harari between 0 and 63, which represents an average of lights in and La Ferrara (2018) and rely on the MIRCA 2000 crop all nights after sunlight, moonlight, aurorae, forest fires, calendars data set (Portmann et al. 2010). This dataset and clouds have been removed algorithmically, leaving offers harvest areas by crop and start and end months mostly human settlements. Given that the original DMSP of growing seasons for each crop (available at a spatial NTL time series data are not comparable across years due resolution of 5 arc-minute or roughly 9km at equator). to the lack of on-board calibration, varied atmospheric We first identify the main crop for each cell based on conditions, satellite shift, and sensor degradation, we rely harvested areas according to the MIRCA 2000 crop on inter-calibrated DMSP NTL time series data from Li data and then match with the growing month calendar et al. (2020). for those cell-specific major crops. Finally, we average monthly SPEI values for the growing season months of Population a cell’s main crop in a given year. Higher values of this  ur grid-level population data derives from the Global O variable correspond to more favorable conditions for local Human Settlements (GHS). GHS use satellite data to agriculture. For areas that are considered to be largely obtain for each cell built-up land area over time, more non-agricultural or pastoral, we use the annual average precisely circa 1975, 1990, 2000 and 2013/14, which SPEI. nicely coincides with the end of our period of study. Furthermore, GHS reconstructs grid-level populations Access to gridded electricity circa 1975, 1990, 2000 and 2015, using population D  ata on gridded electricity comes from GridFinder, an levels at a relatively low administrative level circa these open source tool for predicting the location of electricity years and then allocating the population within these network lines.247 Arderne et al. (2020) constructed the administrative areas depending on the distribution of composite map of the global power grid by applying built-up land area. multiple filtering algorithms to night-time light imagery to identify locations most likely to be producing light Climate data from electricity. These light sources (target-locations) are  ur main climate indicator is the Standardized O then connected to known electricity networks through a Precipitation-Evapotranspiration Index (SPEI), developed 247 Data can be downloaded from https://gridfinder.org/ 2.4 Determinants of regional growth 101 Lake Chad Regional Economic Memorandum  |  Development for Peace least-cost routing algorithm following roads and known 2.4.2 Results distribution lines (adopted from OpenStreetMap). Table 2.1 presents the main results and Table 2.2 Conflict data reports the results from a Shapley decomposition C  onflict data is from the Armed Conflict Location and of the regressions. We find clear evidence that overall Event Data Project (ACLED) (Raleigh et al., 2010). We there is a sign of conditional convergence in terms of categorize the conflict data by actor into "Boko Haram" economic activities as captured by nighttime light. More and "non-Boko Haram" observations.248 Using the data, substantively, the level of economic activities grew more we construct the cell-specific number of conflict events quickly in less-developed areas by an annual convergence and number of fatalities. rate of 1 percent between 1992 and 2013 after accounting for other potential confounders including population size Market access and population growth. It is worth highlighting that the  e construct the index of market accessibility, which is W coefficient for the Lake Chad dummy is close to zero and the weighted sum of population in all the major markets far from significant. What this implies is that there is little or urban agglomerations, which are weighted by travel sign of divergence or convergence between the Lake Chad time/distance. See Appendix 2.C for details on data region and rest of the countries overall conditional on sources and methodology used to construct this index. the initial level of development and other demographic, geographic and socio-economic factors. Within the Lake Cropland and grazing land Chad region, we did not see any sign of convergence. D  ata on cropland and grazing land derive from the HYDE 3.2 panel dataset (Goldewijk, 2017), which provides a The growth of nighttime light is strongly driven by spatial estimation of cultivated land (excluding urban population density and its growth over time both areas and pasture land) for each of our grid cells. The across the four countries and within the Lake Chad. HYDE 3.2 data on land classification is generated based This perhaps does not come as a surprise because on satellite imagery of land use and is available at a spatial nighttime lights are both a function of population density resolution of five arc-minute or 10 km by 10 km for the as well as economic activities.249 Indeed, the initial level period between 10 000 before Common Era (BCE) to of nighttime light luminosity and population altogether 2015 Common Era (CE). explains roughly 15–20 percent in the intensive margin and full samples covering the four countries though their Elevation and terrain slope constraints explanatory power is much smaller within the Lake Chad D  ata on elevation and terrain constraints for farming region. This finding accords with other studies250 showing come from SRTM data Version 4 (Jarvis et al. 2008) and that the locations of urban agglomerations remain the Global Agro-Ecological Zoning (GAEZ) database. persistent over time even after controlling for other factors Elevation data are available at a spatial resolution of 90m. that led to their establishments in the first place. These GAEZ data on terrain slope constraints classifies the urban agglomerations continue to grow faster compared level of constraints to a given area or grid cell into seven to more sparsely populated areas and have important different categories from 0 (no constraints) to 7 (severe implications for widening spatial gaps between those core constraints) and distribution of these classes were defined cities and the rest of the countries. for each 30 arc-second grid cell (or about 1km). 248 Boko Haram conflict includes any violent incident where an actor includes: Islamic State (West Africa) and/or Boko Haram - Jamatu Ahli is-Sunnah lid- Dawatai wal-Jihad or Boko Haram - Jamatu Ahli is-Sunnah lid-Dawatai wal-Jihad. 249 Henderson et al. 2018. 250 See Henderson et al. 2018; Jedwab et al. 2017. 102 2.4 Determinants of regional growth Technical Paper 1. Socioeconomic Trends in the Lake Chad Region Market accessibility is positively correlated with the annual rate of economic growth and this relationship is statistically significant. Indeed, besides the initial level of nighttime light and population size, market access has one of the highest contributions to explaining variation in the annual rate of growth in nighttime lights, accounting for 6 percent of variation in the dependent variable. As discussed above, there remains a large number of people in the region who are left disconnected from other potential markets due to poor road networks and infrastructure in and around the Lake Chad region. To better enhance the natural and human potential of the Lake Chad region, better connectivity and mobility within the Lake Chad region and also between the region and other areas of the countries will improve the living conditions of the population and can lead to a catalytic effect where development benefits in other sectors, such as basic services and livelihoods, can be maximized. Climate conditions play an integral role in determining local economic growth particularly in largely agricultural lands. The positive effect of favorable weather conditions during the growing seasons on local economies increases in more agrarian economies, as indicated by a statistically significant positive interaction term between average SPEI for 1992–2013 during the growing seasons and cropland density. These findings suggest that the impact of climate shocks is not spatially uniform and assessing the potential risks that erratic weather conditions may pose to local agricultural economies should be carefully evaluated. Lastly, the incidence of Boko Haram conflicts is negatively correlated with local economic growth. More substantively, a one percentage point increase in the number of BH-related conflicts is associated with the slower annual rate of growth in nighttime lights by 0.4 percent. As discussed above, Boko Haram conflicts have been a major threat to local economies by disrupting regional trade and imposing immeasurable humanitarian costs. Our results seem to corroborate this. On the other hand, it is worth noting that non-BH conflicts are positively correlated with local economic growth. 2.4 Determinants of regional growth 103 Lake Chad Regional Economic Memorandum  |  Development for Peace 2.5 Conclusion The notes has presented an analysis of core development challenges that detail the Lake Chad region from the path of sustainable economic growth. The findings highlight the laggedness of the region compared to the rest of each respective country in multiple dimensions— including poverty, human capital outcomes, access to public services, among others. The Lake Chad region is characterized by a combination of low density as well as high distance/division, which coalesces to present constraints to its regional growth. Due to a lack of vibrant agglomeration economies, the region fails to unleash scale economies that could spur job creation, improve access to core services, and generate new economic opportunities for the poor. Low density is further compounded by high distance that characterizes the economic geography of the region. Due to the poor quality of road infrastructure—in combination with the landlocked geography of the region and ongoing conflict—a large number of people in the region remain disconnected from large markets within their own respective country or in the neighboring countries. That said, simply investing in connective infrastructure does not necessarily yield expected economic benefits unless such investment is combined with complementary policies to mitigate the negative impact of high division rooted in pre-existing ethnolinguistic cleavages and ongoing conflicts. 104 2.5 Conclusion Technical Paper 1. Socioeconomic Trends in the Lake Chad Region Table 2.1: Main correlates with local economic growth: Regression analysis Model (1) (2) (3) (4) Sample Intensive Full Intensive margin Full sample margin sample Within Lake Chad Within Lake Chad -0.010*** -0.006*** -0.007 -0.006 Nighttime light in 1992 (ln) (0.002) (0.002) (0.005) (0.006) 0.000 -0.000 Lake Chad region (0.010) (0.001) 0.003*** 0.002*** 0.002*** 0.002*** Population in 1990 (ln) (0.000) (0.000) (0.001) (0.000) Annual growth rate of population 0.083*** 0.022*** 0.116* 0.025 (1990–2015) (0.018) (0.007) (0.060) (0.016) 0.003*** 0.002*** 0.005 0.002 Access to gridded electricity (0.001) (0.000) (0.004) (0.001) 0.002*** 0.001*** 0.000 0.000** Market accessibility (0.000) (0.000) (0.000) (0.000) 0.006 0.002** -0.008 -0.000 SPEI in 1992 (0.005) (0.001) (0.016) (0.001) 0.000 -0.000 -0.000** -0.000** Cropland density in 1990 (0.000) (0.000) (0.000) (0.000) -0.000 -0.000 0.000 0.000 SPEI in 1992 × Cropland density in 1990 (0.000) (0.000) (0.000) (0.000) -0.019** -0.003** -0.002 -0.003 Avg. SPEI (1992-2013) (0.008) (0.001) (0.033) (0.003) Avg. SPEI (1992-2013) × Cropland density 0.001** 0.000** 0.000 0.000 in 1990 (0.000) (0.000) (0.001) (0.000) -0.004* -0.003** -0.001 -0.002** Number of BH conflicts (ln) (0.002) (0.001) (0.001) (0.001) 0.001 0.002** 0.003 0.005*** Number of non-BH conflicts (ln) (0.001) (0.001) (0.003) (0.001) 0.003* 0.004** -0.005 -0.006 Number of civilian protests/riots (ln) (0.002) (0.002) (0.005) (0.004) 0.000 -0.000 -0.000 -0.000 Grazing land density (0.000) (0.000) (0.000) (0.000) 0.000 -0.000 0.000 0.000 Elevation (0.000) (0.000) (0.000) (0.000) -0.001 -0.000 -0.003** -0.000 Terrain constraints (0.001) (0.000) (0.001) (0.000) Observations 5,211 32,097 663 5,369 R-squared 0.393 0.317 0.254 0.097 Standard errors clustered by ADM1 regions. All the regressions include ADM1 regional dummies and fixed effects for major crops. Note that the intensive margin sample only includes cells that are ever lit at some point between 1992 and 2013 whereas the full sample includes all cells. *** p<0.01, ** p<0.05, * p<0.1. 2.5 Conclusion 105 Lake Chad Regional Economic Memorandum  |  Development for Peace Table 2.2: Shapley decomposition on select indicators Intensive margin Full sample Model Intensive margin Full sample Within Lake Chad Within Lake Chad Shapley Shapley Shapley Shapley Factor Percent Percent Percent Percent value value value value Nighttime light in 1992 (ln) 0.14346 36.53 0.07396 23.33 0.00713 2.81 0.00201 2.07 Population in 1990 (ln) 0.04657 11.86 0.07722 24.36 0.02161 8.52 0.03775 38.88 Annual growth in population 0.00685 1.75 0.0054 1.70 0.01248 4.92 0.00671 6.91 Avg. SPEI (1992–2013) 0.01611 4.10 0.001 0.32 0.01827 7.21 0.00257 2.64 Avg. SPEI (1992–2013) X 0.00636 1.62 0.00273 0.86 0.0192 7.57 0.00338 3.48 Cropland 1990 # of non-BH conflicts 0.00356 0.91 0.00721 2.28 0.00544 2.14 0.00654 6.74 # of BH conflicts 0.00103 0.26 0.00055 0.17 0.00519 2.05 0.00134 1.38 # of civilian protests/riots 0.0061 1.55 0.01012 3.19 0.00273 1.08 0.00207 2.13 Market access index (ln) 0.06337 16.14 0.0276 8.71 0.00901 3.55 0.00561 5.78 TOTAL 0.39271 100.00 0.31696 100.00 0.25351 100.00 0.09709 100.00 106 2.5 Conclusion Technical Paper 1. Socioeconomic Trends in the Lake Chad Region References Ahmadu, H. J. (2011). Farmer-Herder Conflict: Exploring the Causes and Management Approaches in the Lake Chad Region Nigeria [Phd, Universiti Utara Malaysia]. http://etd.uum.edu.my/3399/ Aker, J. C. 2011. “Dial ‘A’ for Agriculture: A Review of Information and Communication Technologies for Agricultural Extension in Developing Countries.” Agricultural Economics 42 (6): 631–47. Aker, J. C. and I. M. Mbiti. 2010. “Mobile Phones and Economic Development in Africa.” Journal of Economic Perspectives, 24 (3): 207–32. Al Jazeera. September 2017. "More than half of Borno schools closed over Boko Haram."Al Jazeera. https://www. aljazeera.com/news/2017/9/29/more-than-half-of-borno-schools-closed-over-boko-haram. Alibhai, A.S., N. Buehren, R. D. Coleman, M. P. Goldstein and F. Strobbe. 2018. Disruptive Finance: Using Psychometrics to Overcome Collateral Constraints in Ethiopia. Washington, DC: World Bank Group. Arderne, C., Zorn, C., Nicolas, C. et al. Predictive mapping of the global power system using open data. Sci Data 7, 19 (2020). https://doi.org/10.1038/s41597-019-0347-4. Bahia, Kalvin, Pau Castells, Genaro Cruz, Takaki Masaki, Xavier Pedrós, Tobias Pfutze, Carlos RodríguezCastelán, and Hernán Jorge Winkler. 2020. “The Welfare Effects of Mobile Broadband Internet: Evidence from Nigeria.” Policy Research Working Paper 9230, World Bank, Washington, DC. Béné, C., Neiland, A., Jolley, T., Ovie, S., Sule, O., Ladu, B., Mindjimba, K., Belal, E., Tiotsop, F., Baba, M., Dara, L., Zakara, A., & Quensiere, J. (2003). Inland Fisheries, Poverty, and Rural Livelihoods in the Lake Chad Basin. Journal of Asian and African Studies, 38(1), 17–51. https://doi.org/10.1177/002190960303800102 Cohen, Corentin. 2015. "Boko Haram, une impossible sociologie politique?" Afrique contemporaine 2015/3 (No 255) p.71-87. Chun, Natalie and Heiwai Tang. 2018. “Do Information and Communication Technologies Empower Female Workers? Firm-Level Evidence from Viet Nam (May).” ADBI Working Paper 545. Debo, L. and G. Van Ryzin. 2013. “Leveraging Quality Information in Stock-Outs.” Research Paper No. 13–58, University of Chicago Booth School of Business, Chicago, IL. ECC Platform. (n.d.). Local Conflicts over Resources around Lake Chad. ECC Factbook. Retrieved December 1, 2020, from https://www.ecc-platform.org/conflicts/lake-chad-africa-local-conflicts-over-survival-resources. Fernandes, A. M., A. Mattoo, H. Nguyen and M. Schiffbauer. 2019. “The internet and Chinese exports in the pre-ali baba era.” Journal of Development Economics 138: 57-76. Fratianni, Michele and Heejoon Kang. 2006. "International Terrorism, International Trade, and Borders." Working Paper No 2006-13, Indiana University, Kelley School of Business, Department of Business Economics and Public Policy Ghosh, Tilottama, Rebecca L. Powell, Christopher D. Elvidge, Kimberly E. Baugh, Paul C. Sutton2, and Sharolyn Anderson. "Shedding Light on the Global Distribution of Economic Activity." The Open Geography Journal, 3, 147-160. Goldewijk, K., A. Beusen, J. Doelman and E. Stehfest. 2017. Anthropogenic land use estimates for the Holocene; HYDE 3.2, Earth System Science Data, 9, 927-953. Goyal, A. 2020. “Information, direct access to farmers, and rural market performance in central India.” American Economic Journal: Applied Economics, 2(3): 22-45. Hjort, Jonas, and Jonas Poulsen. 2019. “The Arrival of Fast Internet and Employment in Africa.” American Economic Review 109 (3): 1032–79. References 107 Lake Chad Regional Economic Memorandum  |  Development for Peace Harari, Mariaflavia and Eliana La Ferrara. 2018. "Conflict, Climate, and Cells: A Disaggregated Analysis." The Review of Economics and Statistics 100 (4): 594–608. Hasbi, M. and A. Dubus, 2019. "Determinants of Mobile Broadband Use in Developing Economies: Evidence from Sub-Saharan Africa," Working Papers hal-02264651, HAL. Henderson, Vernon J., Tim Squires, Adam Storeygard and David Weil. 2018. "The Global Distribution of Economic Activity: Nature, History, and the Role of Trade." The Quarterly Journal of Economics, Volume 133, Issue 1, February 2018, Pages 357–406. https://doi.org/10.1093/qje/qjx030. Jarvis, A., H.I. Reuter, A. Nelson, E. Guevara, 2008, Hole-filled SRTM for the globe Version 4, available from the CGIAR-CSI SRTM 90m Database (http://srtm.csi.cgiar.org). Jedwab, Remi, Edward Kerby, and Alexander Moradi.2017. "History, Path Dependence and Development: Evidence from Colonial Railroads, Settlers and Cities in Kenya." The Economic Journal 127 (August), 1467-1494. Kaila, H. and F. Tarp. 2019. “Can the Internet Improve Agricultural Production? Evidence from Viet Nam”. Agricultural Economics, 675-91. ICG (International Crisis Group). 2017. Fighting Boko Haram in Chad: Beyond Military Measures. Africa Report N°246. ICG, Brussels, Belgium. Li, X., Zhou, Y., Zhao, M. et al. A harmonized global nighttime light dataset 1992–2018. Sci Data 7, 168 (2020). https://doi.org/10.1038/s41597-020-0510-y. Magrin, G. (2016). The disappearance of Lake Chad: History of a myth. Journal of Political Ecology, 23(1), 204–222. https://doi.org/10.2458/v23i1.20191. Magrin, Géraud and Marc-Antoine Perouse de Montclos, M. (Eds.) 2018. Crisis and Development the Lake: The Chad Region and Boko Haram. AFD, Paris. Masaki, Takaaki, Rogelio Granguillhome Ochoa, and Carlos Rodríguez-Castelán. 2020. “Broadband Internet and Household Welfare in Senegal.” Policy Research Working Paper 9386, World Bank, Washington, DC. Mahmood S. Omar and Ndubuisi Christian Ani. 2018. "Responses to Boko Haram in the Lake Chad Region: Policies, Cooperation and Livelihoods." Institute for Security Studies, Pretoria, South Africa. Mercy Corps, USAID, Oxfam, Action against Hunger, CRS, Rescue and COPI. 2017. The North-East Nigeria Joint Livelihood and Market Recovery Assessment. Mercy Crops, Portland, Oregon. Nkiaka, E., Nawaz, N. R., & Lovett, J. C. (2017). Analysis of rainfall variability in the Logone catchment, Lake Chad basin. International Journal of Climatology, 37(9), 3553–3564. https://doi.org/10.1002/joc.4936 Obi, Felix A. and Ejemai Eboreime. May 2017. "How Boko Haram is devastating health services in North-East Nigeria." The Conversation. https://theconversation.com/how-boko-haram-is-devastating-health-services- in-north-east-nigeria-65751. Onuoha, F. C. (2010). Climate change, population surge and resource overuse in the Lake Chad area. In D. A. Mwiturubani & J.-A. Van Wyk (Eds.), Climate change and natural resources conflicts in Africa (pp. 23–43). Institute for Security Studies. http://www.iss.co.za/uploads/Mono170.pdf. Paunov, C. and V. Rollo, 2015. Overcoming Obstacles: The Internet's Contribution to Firm Development, The World Bank Economic Review, Volume 29, Issue suppl_1, Pages S192–S204. Portmann, Felix T., Stefan Siebert, and Petra Döll, “MIRCA2000: Global Monthly Irrigated and Rainfed Crop Areas around the Year 2000: A New High-Resolution Data Set for Agricultural and Hydrologicalv Modeling,” Global Biogeochemical Cycles 24:1 (2010). doi:10.1029/2008GB003435. Raleigh, Clionadh, and Caitriona Dowd. 2015. “Armed Conflict Location and Event Data Project (ACLED) Codebook.”https://www/acleddata.com/wp-contnet/uplodas/2015/01/ACLED_Codebook_2015.pdf. 108 References Technical Paper 1. Socioeconomic Trends in the Lake Chad Region Ritter, P. and M. Guerrero. 2014. “The Effect of the Internet and Cell Phones on Employment and Agricultural Production in Rural Villages in Peru.” Working paper, University of Piura, Piura, Peru. Samuel, Mali. 2019. "Islamic State West Africa Province is proving its economic resilience by generating income from obliging local communities." Economics of terrorism in lake Chad Basin, ISS Today, Institute for Security Studies. https://issafrica.org/iss-today/economics-of-terrorism-in-lake-chad-basin. Salas Garcia, Vania B. and Q. Fan, 2015. "Information Access and Smallholder Farmers’ Selling Decisions in Peru," 2015 AAEA & WAEA Joint Annual Meeting, July 26-28, San Francisco, California 205380, Agricultural and Applied Economics Association. Smith, D., and Wolters, S. (2017). Radio, Social Media and Language as an Agent in Conflict: A Case Study of Dandal Kura Radio International in the Lake Chad Basin. In P. Aall & C. A. Crocker (Eds.), The Fabric of Peace in Africa: Looking beyond the State (pp. 273–288). McGill-Queen’s University Press. http://www.jstor.org/ stable/j.ctt1pq33ws. USAID, “Lake Chad Basin - Complex Emergency,” 2020. Vedeld, T. (2000). Village politics: Heterogeneity, leadership and collective action. The Journal of Development Studies, 36(5), 105–134. https://doi.org/10.1080/00220380008422648. Vicente-Serrano Sergio M., Santiago Beguería, J. Ignacio López-Moreno, Marta Angulo, and Ahmed M. El Kenawy, “A Global 0.5 Gridded Dataset (1901–2006) of a Multiscalar Drought Index Considering the Joint Effects of Precipitation and Temperature,” Journal of Hydrometeorology 11:4 (2010), 1033–1043. Viollaz, M. and H. Winkler. 2020. “Does the Internet Reduce Gender Gaps?: The Case of Jordan." The Case of Jordan (March 13, 2020). World Bank Policy Research Working Paper 9183. Weiss DJ., Nelson A., Gibson HS., Temperley WH., Peedell S., Lieber A., Hancher M., Poyart E., Belchior S., Fullman N., Mappin B., Dalrymple U., Rozier J., Lucas TCD., Howes RE., Tusting LS., Kang SY., Cameron E., Bisanzio D., Battle KE., Bhatt S., Gething PW., A global map of travel time to cities to assess inequalities in accessibility in 2015 Nature. January 2018 553: 333–336. UNHCR (United Nations High Commissioner for Refugees) and World Bank. 2016. Forced Displacement by the Boko Haram Conflict in the Lake Chad Region. 2016. UNHCR and Worldbank. USAID, “Lake Chad Basin - Complex Emergency,” 2020. WFP (World Food Programme). 2016b. Assessing Gender and Markets in the Lake Chad Basin Region. VAM Gender and Markets Study, 2016-2017. World Food Programme, Dakar, Senegal. World Bank. 2009. World Development Report 2009: Reshaping Economic Geography. World Bank, Washington, D.C. World Bank. 2015. Country Partnership Framework for the Republic Of Chad for the Period FY16-20. World Bank, Washington, D.C. World Bank. 2018. Breaking Down the Barriers to Regional Agricultural Trade in Central Africa. World Bank, Washington D.C. World Bank. 2019. Africa’s Pulse: An Analysis of Issues Shaping Africa’s Economic Future. Vol. 19 (April). Washington, DC: World Bank. World Bank. 2020. PROLAC Program Appraisal Document, Report PCBASIC0089548. References 109 Lake Chad Regional Economic Memorandum  |  Development for Peace Appendix 2.A: Supplementary Information Table A2.1: Socioeconomic Outcomes in the Lake Chad Region and Other Regions Cameroon Chad Niger Nigeria National National National National country country country country Rest of Rest of Rest of Rest of LCB LCB LCB LCB Poverty and Human Capital (percent) Poverty 26.0 18.8 58.8 38.1 39.8 30.7 45.4 43.4 51.6 39.1 37.7 72.3 Literacy (15+) 77.8 83.7 47.1 41.6 45.5 21.0 33.0 32.8 34.0 71.9 72.4 58.7 Primary Completion 72.3 79.0 39.4 29.5 32.8 12.2 25.4 27.8 17.4 78.2 79.0 60.0 (14–25) Child stunting 28.9 26.9 37.3 39.9 37.0 51.0 43.9 41.7 52.2 36.8 35.8 48.0 Employment (percent) Agriculture 45.9 39.0 78.0 76.2 77.0 72.2 74.7 75.5 68.5 43.2 42.2 43.2 Industry 15.6 17.1 8.7 7.0 6.9 10.2 7.9 7.2 7.5 9.4 9.5 9.4 Service 31.0 35.3 10.9 13.6 12.8 17.0 15.6 15.2 20.4 37.1 37.7 37.1 Access to Public Services (percent) Piped Water 35.6 39.6 14.8 15.9 16.4 13.8 31.8 34.7 23.5 11.2 10.4 20.7 Improved Sanitation 61.1 66.9 30.8 14.1 16.2 5.7 24.8 28.5 14.3 56.3 56.4 55.1 Electricity 62.2 70.4 19.6 7.7 9.0 2.4 14.4 16.0 10.0 59.4 61.3 38.4 Source: World Bank calculations based on data of the national authorities. Note: Data on education (literacy and primary education), and employment are based on the latest household surveys conducted in Cameroon (2014), Chad (2018), Niger (2018), and Nigeria (2018). Job category by industry includes only working-age individuals (ages 15–65). Data on poverty are based on the latest harmonized household surveys conducted in Cameroon (2014), Chad (2011), Niger (2014), and Nigeria (2018). Poverty rates are based on the US$1.90 international poverty line (2011 purchasing power parity). Data on child health (child stunting) and access to public services are drawn from the latest Demographic and Health Surveys available in each country: Cameroon (2018), Chad (2018), Niger (2012), and Nigeria (2018). Rest of country = outside the Lake Chad region; LCB = within the Lake Chad Basin region. 110 Appendix Technical Paper 1. Socioeconomic Trends in the Lake Chad Region  escriptive Statistics and Sources of Data Appendix 2.B: D Used in the Multivariate Regression Variable N mean sd min max Annual rate of growth in NTL (1992–2013) 32097 0.001 0.01 -0.148 0.15 Nighttime light in 1992 (ln) 32097 0.063 0.372 0 4.156 Lake Chad dummy 32097 0.167 0.373 0 1 Population in 1990 (ln) 32097 1.387 1.386 0.693 9.299 Annual growth rate of population (1990–2015) 32097 0.011 0.022 -0.194 0.244 Access to gridded electricity 32097 0.211 0.408 0 1 Market accessibility 32097 8.062 4.87 0.889 31.963 SPEI in 1992 32097 0.32 0.463 -0.889 1.265 Cropland density in 1990 32097 11.566 16.59 0 68.939 Avg. SPEI 1992–2013 32097 -0.226 0.187 -0.688 0.422 Number of BH conflicts (ln) 32097 0.005 0.095 0 5.568 Number of non-BH conflicts (ln) 32097 0.03 0.223 0 5.333 Number of civilian protests/riots (ln) 32097 0.01 0.136 0 4.883 Grazing land density in 1992 32097 23.55 26.694 0 76.83 Elevation 32097 473.007 263.173 1.339 3015.201 Terrain constraints 32097 2.439 1.029 0 7 Appendix 111 Lake Chad Regional Economic Memorandum  |  Development for Peace Appendix 2.C: Market Access Index Definition  n this report, we define market access as a measure of accessibility from one origin to all destinations based on travel I distance (or travel time). More formally, market access for a given location (or origin) i can be expressed as follows: –θ MAot = ∑ Pdtτodt d  here Pd refers to the population of a location (or destination) d, τodt is travel time from cell o to destination d, and w θ is a trade elasticity or decay parameter measuring how trade volumes fall as travel times increase. We set ϴ at 3.8 following Jedwab and Storeygard (2020) Destinations that are considered in the construction of this index include all urban agglomerations with a population greater than 100,000.251 In other words, market access is the weighted sum of population in all the destinations, which are weighted by travel time/distance. Travel time is computed based on the friction map generated by Weiss et al. (2015). In this study, destinations are defined as all major population centers in those four countries and origins are points on the road network that are closest to the centroid of each village. Data  e calculation of market access requires the following information: 1) census population and geographical coordinates Th (e.g., longitude and latitude) of all population centers; and 2) road network data. Population centers are defined as those half-degree grids with more than 300,000 inhabitants. For road network information, we rely on OpenStreetMap. 251 We used grid-level population data (at a resolution of 1km) from the Global Human Settlements (GHS) to identify urban agglomerations. We first generated a boundary shapefile of urban agglomerations which correspond to all adjacent grid cells with a population size greater than 5,000. We then computed travel time from each grid cell to the centroid of those urban agglomerations. 112 Appendix Technical Paper 1. Socioeconomic Trends in the Lake Chad Region Appendix 2.D: Rural Access Index  e Rural Access Index (RAI) measures the share of rural population living within 2km away from all-season roads. Th To construct this index, we rely on three sources of data: OpenStreetMap, WorldPop 2015 population estimates, and GRUMP Global Rural-Urban Mapping Project, Version 1 (GRUMPv1) (CIESIN, Columbia University, CUNY, CIDR, IFPRI, and CIAT 2017). We apply the following methodology as laid out in https://rai.azavea.com/: • Select commonly used tags from OpenStreetMap (Trunk, Primary, Secondary, Tertiary) that serve as an approximation for all-season roads • Create a mask based on 2 km buffer on these roads • Create a mask based on urban areas as defined by GRUMP urban extents polygons • Summarize the population remaining on the 100 metre raster dataset from WorldPop 2019 Appendix 113