Oyewole Simon Oginni | Researcher, Armed Conflicts in Sahel, Bonn International Centre for Conflict Studies

Abstract

The Lake Chad Basin (LCB) region is battling the scene of continuous violent conflicts and climate-induced displacement. While the state counterinsurgency measures have helped regain some of the areas controlled by non-state armed groups, a large portion of the agriculturally-rich areas in the region remain a battlefield between the state and insurgent groups. The inaccessibility of the civilian population to the contested areas has disrupted food production systems, causing a sharp rise in food prices in state-controlled areas and influencing civilians’ preference for armed-group-controlled areas for food security. In this paper, we examine the effects of state and non-state intervention programs on civilian risk preferences in the frontline areas. Drawing on empirical data from two frontline cities in Nigeria and Cameroon, we find that the intervention programs targeting housing, healthcare and business/skill training reduce the civilian propensity to migrate to the insurgent-controlled areas to secure agricultural access resources. Specifically, young and middle-aged business/skill training participants are significantly likelier to stay in state-controlled areas over the insurgents-controlled areas than non-participants. While women profit more from healthcare and housing support than men, we do not find any gender-specific effect of business/skill training on civilian risk preferences. Our study offers insights into short-term measures that can be explored to reduce civilian risk preferences in conflict-torn spaces where non-state armed groups use to access to agricultural resources to control civilians and civilian behaviours.

Keywords: Armed conflicts, Boko Haram, Cameroon, Food security, Nigeria

1. Introduction

Prolonged conflicts exacerbate food insecurity, affecting civilians’ coping strategies, risk preference and trust in the state’s ability to honour its social contract on providing welfare and security (Boege et al., 2009; Orjuela-Grimm et al., 2022). Recent studies have shown that civilians’ inaccessibility to agricultural resources during armed conflicts can sway their loyalty in favour of non-state armed groups who appear to have greater control over the territory (Bahiss et al., 2022; Kemmerling et al., 2022). Such risk preferences are well documented during armed conflicts in Mali, Afghanistan, the Philippines and Somalia, where civilians had to pay taxes and levies to the insurgents to access agricultural resources (Bahiss et al., 2022; Revkin & Ahram, 2020; Weigand, 2017). In the Lake Chad Basin region, which traverses Nigeria, Chad, Nigeria and Cameroon, civilians’ coping strategies include bargaining and co-optation with armed groups or a deliberate outmigration to the areas controlled by armed groups in order to access lakes, rivers and other agricultural resources.

Responsiveness of the state to the critical needs of civilians living in crises can improve state legitimacy, enhance how civilians make better decisions on the frontlines and reduce the possibility of altered loyalty in favour of insurgent groups (Balch, 1974; Bøås & Strazzari, 2020). As a result, the state often implements or supports intervention programs to improve its presence in the areas where the non-state armed groups are perceived to be gaining ground and to minimize the effects of armed conflicts on the civilian population in frontline areas (Obamamoye, 2019; Raeymaekers, 2011). However, while chunks of research on the impact of food insecurity in conflict situations have mainly focused on rural-urban migration (Orjuela-Grimm et al., 2022; Riebe & Dressel, 2021), few studies have attempted to explore civilian outmigration from frontline cities to rural areas, especially the areas under the control of non-state armed groups.

This research aims to examine the effects of state and non-state intervention programs on civilian risk preferences in the frontline areas. To achieve our objective, we analyse how intervention programs affect civilian migration choice on access to agricultural resources in state-controlled and armed-group-controlled areas in the Lake Chad Basin region. The LCB region faces violent conflicts and climate-induced displacement (Ruppel & Funteh, 2019). While the state’s counterinsurgency operations have helped regain some of the areas controlled by non-state armed groups (Oriola, 2022), a large portion of the agricultural-rich areas remain a battlefield between the state and insurgent groups (Oginni, 2021). The inaccessibility of civilians to the contested areas has caused food shortage and a sharp increase in food prices in state-controlled frontline areas (Fudjumdjum et al., 2019; The Economist Group, 2022), driving civilians’ preferences for armed-group-controlled areas for food security (Oginni, 2023, forthcoming).

Drawing on empirical data from two frontline cities in Nigeria and Cameroon, we find that the intervention programs targeting housing, healthcare and business/skill training reduce civilian risk preferences. Specifically, young and middle-aged business/skill training participants are significantly likelier to stay in state-controlled areas over the insurgents-controlled areas than non-participants. Furthermore, while women profit more from healthcare and housing support than men, we do not find any significant gender-specific effects of business/skill training on civilian risk preferences. Our study offers new insights into short-term measures that can be explored to reduce civilian risk preferences in conflict-torn spaces where non-state armed groups employ accessibility to farmlands to control civilians and civilian behaviours.

2. Background

Increasingly, there have been spatial shifts in the nature and scale of urban displacement in the last few years In the Middle East, for example, Syria’s conflicts have forced many people to flee to cities in Jordan and Lebanon – about 80 per cent of the Syria refugees live in cities across the world (Lintelo, Lakshman et al., 2018; Sangapala, Sultana et al., 2017). Similarly, the protracted conflicts caused by the Al-Shabaab insurgent groups forced many Somalis to relocate to cities, towns and suburbs in Kenya, Uganda and Ethiopia for protection and livelihoods (Nikuze et al., 2019). The situation is not different in the Lake Chad Basin region, in which frontline cities have been central in the ongoing Boko Haram crisis.

LCB region is connected in many ways, especially regarding resources and rich cultural ties, but has been devastated by the violent extremism of Boko Haram terrorist groups. Boko Haram terrorists were initially presumed to have been fighting the war against the corruption and secularity of the Nigerian government before they turned violent in 2009 (Azgaku, 2015; Marc-Antoine Pérouse de Montclos, 2015; Adesoji, 2019). Since the emergence of violent extremist groups, over 4.5 million people have been forcibly displaced from their homes, while 20,000 lives were lost to the crisis (IDMC, 2018; Okoli & Iortyer, 2014; Yusuf, 2019). Although the Multinational Joint Task Force (MJTF) established in 2015 by the governments of the four countries has succeeded in “technically” containing the expansion of terrorist activities (Nagarajan et al., 2018; Zenn, 2017 Obamamoye, 2019), the continuous onslaughts in villages and communities in the remote areas remain a significant setback on food security in the region.

Frontline cities in the Lake Chad Basin have attracted a large number of civilian population (Magrin et al., 2018). However, state presence is very weak in these cities, characterised by multiple challenges that range from governance deficit to localized land-tension, climate fragility to widespread poverty, youth unemployment and low level of illiteracy (IDMC, 2018; Nagarajan et al., 2018). About 80 percent of the displaced resettling in these cities live outside the registered camps, usually in the informal settlements where there are poor health facilities and housing, transport system and inaccessibility to clean water (Fudjumdjum et al., 2019).

Nigeria, Chad, Niger and Cameroon, facing a decade of armed conflicts from non-state armed groups, have remained within the range of the top fifth to seventeenth rank on Failed State Indicators (FSI) over a decade, suggesting that elements of state fragility are very much present within the LCB region. Compared to Gabon (in Central Africa region) and Ghana (in West Africa), which have witnessed a declining trend on FSI, the four LCB states have a high record on security threats to the state, fractionalized leadership, fragmentation of state institutions or power struggles, poverty and uneven development, amongst others (Bøås & Strazzari, 2020).

Since 2017, several intervention programmes have targeted civilians in frontline areas of the LCB, following the success achieved by the Multinational Joint Task Force (MJTF) on counterinsurgency in many frontline cities that were occupied previously by the armed groups (Albert, 2017; UNDP, 2021). State and non-state intervention programs, though often limited to the areas densely populated by internally displaced persons (IDPs) and returnees, have been implemented across the frontline cities in line with the Lake Chad Basin Regional Stabilisation Strategy. For example, the multisectoral recovery project (MRP) in the LCB covers livelihood recovery, service delivery and rehabilitation. The commonly identified interventions within the LCB RSS are housing support, healthcare and business/skill training. Our interest is to examine the effectiveness of these interventions on civilian risk preferences in terms of migration choice between staying within state-controlled areas and leaving for insurgent-controlled areas. To what extent do intervention programs influence civilian risk preference in food-insecure and conflict-torn spaces?

3. Methodology

This study adopts mixed methods. We combine cross-sectional data with 5 focus group discussion and 54 in-depth interviews. The data was collected in Mubi (Nigeria) and Maroua (Cameroon) between between August and December 2019. The study is part of a larger research project on mapping stabilisation vectors in the LCB region (Alupo et al., 2018; Oginni, 2021; Oginni et al., 2018, 2020; Opoku et al., 2020, Oginni, 2023a, 2023b, forthcoming). Our cross-sectional data consists of 2024 households: 761 households in Nigeria (Mubi) and 1263 households in Cameroon (Maroua). This represents 81% response rate of 2500 household survey in the frontline cities after adjusting for incomplete response. A city which has been previously occupied by armed group takes on 1, otherwise 0. In this case, Mubi takes on the value of 1 and Maroua takes on the value of 0. Intention to migrate to access farmlands takes on 1, otherwise 0 and we coded this as stay and exit option respectively.

We are interested in the effects of intervention programs on civilian risk preferences on access to agricultural resources in the frontline cities. Specifically, we intend to estimate on average how different versions of intervention programs (housing, healthcare and business/skill training) affect civilian migration choices in the frontline. Thus, we implemented a causal Double Machine Learning (causalDML) for program evaluation to estimate the effect heterogeneity of intervention programs on migration choice (Heiler & Knaus, 2022; M. C. Knaus et al., 2022). CausalDML is developed based on DML methods and it allows the flexible use of standard statistical tool such as OLS, t-test, supervised learning for estimating the causal parameter of interests. CausalDML is applied under unconfoundedness which assumes access to a vector of pre-treatment variables (M. C. Knaus et al., 2022).

The advantage of DML compared to ML is that it overcomes overfitting problems associated with regularisation through the use of Neyman orthogonal scores and cross-fitting that help reduce bias and produces N ^½ consistent to estimates the parameter of interests (Bach et al., 2022). For our study, we evaluate three intervention programs under the multisectoral recovery project (MRP) of the LCB Regional Stability Strategy, namely, housing, health and business/skill training. We are interested in the average potential outcomes (APO), average treatment effect (ATE), average treatment effect on treated (ATET) and to compute group average treatment effect (GATEs) to determine who benefit or suffer more or less from participating intervention programs within the area based intervention areas: Maroua I, II, and III as well as Mubi North and Mubi South.

APO is the average potential outcome of the intervention if the whole civilian population affected by armed conflicts were assigned to the intervention programs. ATE indicates the average effects in the population while ATET covers the subpopulation that is observed in the intervention programs. GATE is a conditional average treatment that is useful for subgroup analysis (categorical variables) or heterogeneity along continuous variable (Heiler & Knaus, 2022; Knaus et al., 2022).

First, we code the version of the interventions to be multivalued treatment using R-package following the guideline for estimating causal parameter of interests (Heiler & Knaus, 2022). Then, we implement causalDML via R-package. The result produced APOs, ATEs and ATETs. So, we obtain APOs, ATEs and ATETs from the means of the estimated doubly scores estimated through causalDM R-package. We estimate GATEs through OLS of the pseudo-outcome of pre-treatment variable (that is, gender, age group, income, education and city-specific characteristics (Heiler & Knaus, 2022). We also estimate GATEs for income via kernel regression because they are continuous variable. The procedures we followed to arrive at our results, including the detailed steps on how to implement causalDM on R-package, are outlined in Knaus (2022).

4. Findings

Civilian risk preferences in the frontline areas of the LCB region

Prolonged conflicts can increase civilian risk preference, especially when the state is seen as incapable of providing food security and other livelihood opportunities to cushion the effects of conflicts on the civilian population (Bahiss et al., 2022; Jackson et al., 2022). We asked study participants what could influence their migration choice and whether they consider leaving their former farmlands in insurgent-controlled areas. Table 1 shows factors influencing civilian risk preferences in frontline cities in Nigeria and Cameroon. Some participants shared that a sharp rise in food prices and their inability to engage in other works to earn income informed the decision to return to their former farmlands even if Boko Haram controlled the area. Due to land scarcity in state-controlled areas, civilians fleeing the scene of continuous are unable to secure farmland in relatively secured areas due to intense competition and population growth.

 

Table 1: Mapping civilian risk preference on access to agricultural resources in the frontline

Key drivers	Participants’ view
Rise in food prices in frontline cities	‘I spent my whole life over there. There is absolutely nothing to do here in the city to feed my family except I go back to my farm. I am limited to what is available in the markets compared to planting varieties in my farm. It is expensive to feed a family of seven here.’ (47 years, Maroua, Cameroon)
Land scarcity in state-controlled areas	“You know that it is very difficult to secure farmland here. You have to think of moving far from the city and these areas are not safe too. So, the locals have to feed themselves first by allocating farmlands to their people (32yrs, Mubi, Nigeria)
Selective restrictions on access to rivers, lake and farmlands in contested territories.	‘I heard some people still farm in the areas but they are paying some levies to the commanders there (that is, Boko Haram). Some military officers also worked with local security to fish there. Why should I not return to the areas? I am willing to pay any levy to feed my family. (28yrs, Mubi, Nigeria)
Social networks and skill mismatched	‘Our friends and families still live in the area. The only problem is that we have to find someone to speak to the commanders first so that they allow us to farm during this season and pay levy later after harvest. I was only trained as a farmer by my father before his death. People here drive repair handsets, drive tricycle and barbing salon. I did not learn that. So I am not connected to the city in anyway’ (44yrs, Mokolo, Cameroon)
Unequal access to humanitarian supports.	‘Some people were allocated land because the NGO registered them as a refugee. NGO distributed fertilizers to them before farming season. You know I do not have any connection to anyone. I will go back to my farm at any cost.’ (Mokolo, Cameroon)
Future uncertainty about conflict resolution	‘You know this crisis appear endless. We often hear that government has cleared some villages occupied by Boko Haram but you know our people there always come back here to tell us that the same Boko Haram attack their farms. But they are still better there than us here in the city because they can feed themselves and pay no rent like us here (52yrs, Mubi, Nigeria)

 

Future uncertainty about resolving armed conflicts in the LCB affects civilians’ risk aversion or tolerance of access to agricultural resources. Future uncertainty emerges from low confidence or trust in the state’s capacity to restore peace and enhance livelihood recovery in conflict-torn spaces. Some participants shared that the state’s declaration of victory in some villages did not run in parallel with the scene of continuous onslaught in most remote areas where Boko Haram, ISWAP and other militias collect levies and punish whoever refuses to do so. Figure 2b-g shows the change in the spatial pattern of attacks from NSAGs between 2011 and 2022. While the Boko Haram insurgency was limited to Nigeria territory between 2011 and 2012 (Figure 2a), the insurgents expanded to the Far North region of Cameroon after two years (Figure 2c).

Later, Boko Haram launched its first attack in Chad after the Chadian government indicated an interest in participating in joint counterinsurgency operations in the LCB region. While the success of the operation of MNJTF limits the activities of Boko Haram to remote areas between 2016 and 2018, the emergence of ISWAP and other NSAGs beyond conventional frontlines has increased civilian causalities in recent years (Figure 2f-g). The new security threats, coupled with the challenge of early warning and response, have been exploited by Boko Haram and ISWAP for their recruitment drive in the LCB region. Between 2021 and 2022, the armed conflict incidents spread to the northcentral of Nigeria and spiralled into farmer-herders conflicts due to the southward migration of conflict-affected populations around the LCB region (see Figure 2g). However, a recent study has shown that very few people migrate outside conflict zone in the LCB region (Oginni, 2023b, forthcoming).

Effects of intervention programs on civilian risk preferences

Table 1 shows the intervention by program types and the demographic characteristics of participants and non-participants. Participants are civilians who participated in any of the three intervention programs. In contrast, non-participants who live within the targeted intervention area are, therefore, eligible to participate but did not eventually participate. About 52% (1053 households) participated in the programs, while 48% (961) did not participate in any of the three programs. More men participated in the three programs than women, while young and middle-aged adults participated in the three programs than older adults. Overall, participants who participated in one of the three intervention programs have a higher income than those who did not.

Table 1 Descriptive statistics of participation in interventions by program type

Participants (N = 961; 48%)	Non-participants (N = 1063; 52% )
	No program (1)	Housing (2)	Healthcare (3)	Business/Skill training (4)
No of observations	1063	601	131	229
City/Country:
Mubi (Nigeria)	0.56	0.19	0.03	0.22
Maroua (Cameron)	0.51	0.36	0.08	0.05
Gender (men)	0.58	0.52	0.55	0.55
Marital status (Single)	0.26	0.21	0.19	0.26
Age
Middle aged adults	0.21	0.37	0.24	0.30
Older Adults	0.27	0.28	0.24	0.31
Young Adults	0.52	0.35	0.52	0.39
Education
High	0.13	0.09	0.06	0.09
Low	0.54	0.63	0.60	0.45
Medium	0.33	0.28	0.34	0.46
Income (XAF/NGN)	47309	52213	32139	50099
Migration choice (stay choice)	0.341	0.664	0.773	0.715

 

Our interest is how intervention programmes affect civilian risk preferences, especially civilian migration choices on access to agricultural resources in state-controlled areas and Boko Haram-controlled areas. Here, we compare three programs (housing, healthcare and business/skill training) to non-participation. Specifically, our interest is to compare ATE and ATET (see Table 2). A higher ATET suggests the effectiveness of the program assignment (Knaus et al., 2022). ATE and ATET estimates indicate significant differences in the effectiveness of intervention programs. The three programs (housing, healthcare and business/skill training) show a positive effect (about 0.05 or 5% on average) on the civilian choice of staying in the state-controlled frontline cities. However, a comparison of ATE and ATET reveals no significant difference between the three programs. As Heiler and Knaus (2022) and Knaus et al., (2022) suggest, this might indicate that the program assignment fails to take advantage of the effect heterogeneity because we expect ATETs to be higher than ATEs if the program assignment is well-targeted.

 

Table 2: Average effects of pro-state intervention programs

	Estimate (1)	Standard Error (2)
Panel A: Average potential outcome (APO)
No participation (Control)	0.357	0.014
Housing support (shelter)	0.743	0.015
Healthcare and food support	0.742	0.034
Skill Acquisition and Business	0.668	0.058
Panel B: ATE
Housing support – No participation (Control)	0.386***	0.020
Health and Food Supports – No participation (Control)	0.385***	0.036
Skill Acquisition and Business - No participation (Control)	0.311***	0.059
Panel C: ATET
Housing support – No participation (Control)	0.377***	0.025
Health and food Supports – No participation (Control)	0.343***	0.050
Skill Acquisition and Business - No participation (Control)	0.298***	0.040

 

Notes: The Table reports DML based point estimates and standard errors of average effects

(p < 0.001 **; p < 0.01 **; p< 0.05*)

 

Heterogeneous effects of targeted interventions on civilian migration choice in the frontline.

Here, we study the effect heterogeneity at different scales by estimating group average treatment effect (GATEs) for gender, income, age groups and city of the subgroups. We performed the GATES using a standard OLS regression with the pseudo-outcomes of pre-treatment variable (Knaus et al., 2022). After estimating the DML for average effects, we use the dummy variables for gender, age groups and city and their respective reference groups as covariates (for procedure see Heiler and Knaus, 2022; Knaus et al., 2022). Then, we used kernel regression GATEs for variable income based on the R-package casual DML package (Knaus et al., 2022).

Table 3 shows the results of the standard OLS regression for GATEs. The results show significant gender differences in the effect of housing and healthcare programs. Women, as a reference group, profit more on average significantly from participating in the intervention programs. However, the effect disappears for women on business/skill programs. Panel C shows the OLS regression result of the age group but with two dummies (Middle-aged and Older adults). Again, young adult is the reference group. The F-statistic is not significant at 5% level for health, housing and business/skill training programs. However, it shows that young adults benefit substantially more or suffer less from participating in the three programs. Older adults only benefit more or suffer less from housing programs, while young and middle-aged adults benefit from business/skill training.

Table 4: Group Average Treatment Effects (GATEs) of intervention programs

	Housing	Healthcare	Business/Skill Training
Panel A
Constant	0.436** (0.031)	0.461** (0.057)	0.405** (0.080)
Gender (male)	-0.086* (0.042)	-0.159** (0.077)	-0.136 (0.107)
Panel B:
Constant	0.453** (0.032)	0.323** (0.058)	0.335** (0.080)
Non-returnee	-0.117* (0.042)	0.087 (0.077)	-0.011 (0.107)
Panel C:
Constant	0.365** (0.025)	0.346** (0.045)	0.267** (0.062)
Age (Middle aged Adults)	0.056 (0.0.052)	0.085 (0.095)	0.254* (0.132)
Age (Older Adults)	0.185* (0.086)	0.138 (0.158)	0.161 (0.22)
F-statistic	2.614	0.688	1.944
Panel E:
Constant	0.457** (0.034)	0.447** (0.062)	0.315** (0.087)
City (Maroua – Cameroon)	-0.111* (0.043)	-0.119 (0.079)	0.0233 (0.110)

 

Note: This tables highlight OLS coefficients and their robust standard errors.

In addition, we performed nonparametric GATEs for income the subgroups. While we find a notable effect of business/skill training on income, we do not find any effect of housing and healthcare on income. Figure 3a-c show the result of non-parametric GATEs for income). Our survey did not cover which business/skill trainings the participants benefit from. However, some participants shared that some soft trainings improve their level of understanding and satisfaction with the state efforts on counterinsurgency.

I have participated in a series of peace dialogue sessions with government and NGOs. I understand that government is fighting to protect us from the armed group [Boko Haram]. So, I would rather stay where I feel more protected even though I can barely feed my family (31, Mubi, Nigeria)

5. Discussion and policy implications

Access to agricultural resources during armed conflicts remains a critical tool that non-state armed groups often use to control civilians and civilian behaviours (Kemmerling et al., 2022). This is particularly evident when prolonged conflicts prevent the state from guaranteeing food security in the frontline areas and increase civilian risk preferences, including co-optation and bargaining with non-state armed groups for alternative livelihood (Jackson et al., 2022; Weigand, 2017).

As demonstrated in this study, the inaccessibility to agricultural-rich areas of the LCB region has influenced civilian risk preference for the insurgent-controlled areas to secure alternative livelihoods. Boko Haram and ISWAP insurgents capitalised on the rise in food prices, land scarcity in state-controlled areas, restrictions on access to insurgent-controlled areas, and perception of unequal access to humanitarian assistance to control civilian behaviours. Targeted interventions on healthcare, housing and business/skill training have some potential to reduce civilian risk tolerance in the frontline and outmigration to insurgent-controlled areas. Interventions that target business/skill training may be a short-term option to pursue in a situation where the state cannot guarantee accessibility to farmlands due to evolving security scene. Yet, the more significant proportion of the conflict-affected population resettling in urban areas is farmers without the required skill sets to navigate the urban economy.

In addition, the study findings suggest that the conventional narrative of rural-urban migration during armed conflicts only holds for some situations. In the LCB region, like other regions battling violent conflicts worldwide, we observe outmigration to remote areas for food security among civilians whose livelihood depends on access to farmlands, rivers, lakes and other agricultural resources. Here, there is a potential area for further research on the dynamic of migration within crisis zones.

References

Adesoji, A. (2019). Boko Haram and the Global War on Terror. Oxford Research Encyclopedia of Politics, April, 1-20. https://doi.org/10.1093/acrefore/9780190228637.013.851

Albert, I.O. (2017). Rethinking The Functionality of the Multinational Joint Task Force in Managing the Boko Haram Crisis in the Lake Chad Basin. Africa Development, 42(3), 119-135.

Alupo, B.A., Oginni, S.O., Opoku, M.P., & Torgbenu, E.L. (2018). Psychological experiences of refugees and the response of the community in the Lake Chad region. Behavioral Sciences of Terrorism and Political Aggression. https://doi.org/10.1080/19434472.2018.1463552

Bach, P., Chernozhukov, V., Kurz, M.S., & Spindler, M. (2022). DoubleML-An Object-Oriented Implementation of Double Machine Learning in R *. 

Bahiss, I., Jackson, A., Mayhew, L., & Weigand, F. (2022). Rethinking armed group control Towards a new conceptual framework. 

Balch, G. I. (1974). Multiple Indicators in Survey Research: The Concept “Sense of Political Efficacy”. Political Methodology, 1(2), 1-43. http://www.jstor.org/stable/25791375

Bøås, M., & Strazzari, F. (2020). Governance, Fragility and Insurgency in the Sahel: A Hybrid Political Order in the Making. The International Spectator, 55(4), 1-17. https://doi.org/10.1080/03932729.2020.1835324

Boege, V., Brown, M.A., & Clements, K.P. (2009). Hybrid Political Orders, Not Fragile States. Peace Review, 21(1), 13-21. https://doi.org/10.1080/10402650802689997

Fudjumdjum, H., Leal Filho, W., & Ayal, D.Y. (2019). Assessment of Barriers to Food Security in North-Eastern Nigeria. Handbook of Climate Change Resilience, 1-15. https://doi.org/10.1007/978-3-319-71025-9_99-1

Heiler, P., & Knaus, M.C. (2022). Effect or Treatment Heterogeneity? Policy Evaluation with Aggregated and Disaggregated Treatments.

IDMC. (2018). On the GRID: The global displacement landscape. In Comparative Education Review. https://doi.org/10.1086/447117

Jackson, A., Weigand, F., & Tindall, T. (2022). Understanding agency in civilian-armed group interactions. 

Kemmerling, B., Schetter, C., & Wirkus, L. (2022). The logics of war and food (in)security. Global Food Security, 33(April), 100634. https://doi.org/10.1016/j.gfs.2022.100634

Knaus, M. (2022). Replication Notebook “Double Machine Learning based Program Evaluation under Unconfoundedness”. 

Knaus, M.C., Bonev, P., Huber, M., Kennedy, E., Lechner, M., Semenova, V., Strittmatter, A., Wager, S., & Zimmert, M. (2022). Double Machine Learning based Program Evaluation under Unconfoundedness.

Lintelo, D., Lakshman, R., Mansour, W., & Soye, E. (2018). Wellbeing and Protracted Urban Displacement: Refugees and Hosts in Jordan and Lebanon (Issue April). 

Magrin, G., Montclos, M.P. De, Chauvin, E., Lemoalle, J., Raimond, C., Rangé, C., Doua, S.A., Dangbet, Z., Don-donné, G.P., Hessana, A., Higazi, A., Fougou, H.K., Mahamadou, A., Mohamed, K., Tafida, A.A., Tukur, A.L., & Zieba, F.W. (2018). Crisis and development The Lake Chad region and Boko Haram. 

Nagarajan, C., Pohl, B., Rüttinger, L., Sylvestre, F., Vivekananda, J., Wall, M., & Wolfmaier, S. (2018). Climate-Fragility Profile: Lake Chad Basin. 

Nikuze, A., Sliuzas, R., Flacke, J., & van Maarseveen, M. (2019). Livelihood impacts of displacement and resettlement on informal households – A case study from Kigali, Rwanda. Habitat International, 86(February), 38-47. https://doi.org/10.1016/j.habitatint.2019.02.006

Obamamoye, B.F. (2019). State Weakness and Regional Security Instability: Evidence from Africa’s Lake Chad Region State Weakness and Regional Security Instability: Evidence from Africa’s Lake Chad Region. International Journal of Conflict and Violence, 13, 1-13. https://doi.org/10.4119/UNIBI/ijcv.639

Oginni, O.S. (2021). Arrival or Return? Temporality and Materiality of Transit Sites in Overlapping Displacement Context in Border Cities of the Lake Chad Basin Region. Urban Forum, 0123456789. https://doi.org/10.1007/s12132-021-09450-8

Oginni, S.O., Opoku, M.P., & Alupo, B.A. (2018). Terrorism in the Lake Chad Region: Integration of Refugees and Internally Displaced Persons. Journal of Borderlands Studies, 33, 1-17. https://doi.org/10.1080/08865655.2018.1457975

Oginni, S.O., Opoku, M.P., & Alupo, B.A. (2020). Terrorism in the Lake Chad Region: Integration of Refugees and Internally Displaced Persons. Journal of Borderlands Studies, 35(5), 725-741. https://doi.org/10.1080/08865655.2018.1457975

Okoli, A.C., & Iortyer, P. (2014). Terrorism and Humanitarian Crisis in Nigeria: Insights from Boko Haram insurgency. Global Journal of Human Social Science, 14(1), 39-49.

Opoku, M.P., Nketsia, W., Oginni, O.S., Atim Alupo, B., & Saka, B.N. (2020). Societal perceptions of counterterrorism in Cameroon: The voices of those far from the battlefield. Journal of Human Rights, 19(2). https://doi.org/10.1080/14754835.2020.1716700

Oriola, T.B. (2022). Nigerian Troops in the War Against Boko Haram: The Civilian-Military Leadership Interest Convergence Thesis. Armed Forces and Society, 0(0), 1-35. https://doi.org/10.1177/0095327X211072894

Orjuela-Grimm, M., Deschak, C., Aragon Gama, C.A., Bhatt Carreño, S., Hoyos, L., Mundo, V., Bojorquez, I., Carpio, K., Quero, Y., Xicotencatl, A., & Infante, C. (2022). Migrants on the Move and Food (In)security: A Call for Research. Journal of Immigrant and Minority Health, 24(5), 1318-1327. https://doi.org/10.1007/s10903-021-01276-7

Raeymaekers, T. (2011). Forced Displacement and Youth Employment in the Aftermath of the Congo War: From Making a Living to Making a Life. Ssrn, January. https://doi.org/10.2139/ssrn.1762751

Revkin, M.R., & Ahram, A.I. (2020). Perspectives on the rebel social contract: Exit, voice, and loyalty in the Islamic State in Iraq and Syria. World Development, 132, 104981. https://doi.org/https://doi.org/10.1016/j.worlddev.2020.104981

Riebe, K., & Dressel, A. (2021). The impact on food security of a shrinking Lake Chad. Journal of Arid Environments, 189, 104486. https://doi.org/https://doi.org/10.1016/j.jaridenv.2021.104486

Ruppel, O.C., & Funteh, M.B. (2019). Chapter 5: Climate change, human security and the humanitarian crisis in the Lake Chad Basin region: selected legal and developmental aspects with a special focus on water governance. In Law | Environment | Africa (Issue 2016). https://doi.org/10.5771/9783845294605-105

Sangapala, P., Sultana, N., Sayegh, A., & Al Ber, S. (2017). SETTLEMENT EXPERIENCE OF SYRIAN REFUGEES IN ALBERTA Sandeep Agrawal and Seraphine Zeitouny With assistance from. 

The Economist Group. (2022). Global Food Security Index (GFSI). In The Economist Intelligence Unit. 

UNDP. (2021). Conflict Analysis in the Lake Chad Basin. 

UNHCR. (2018). Forced displacement in 2017: Global Trends (Issue 19 June 2018). https://doi.org/10.1080/09584939308719709

UNSC Res/2349. (2017). Security Council Documents (Issue UNSCOR, 7911th Sess, 31 March, pp. 1-8). https://doi.org/10.5040/9781472565105.ch-004

Weigand, F. (2017). Afghanistan’s Taliban – Legitimate Jihadists or Coercive Extremists? Journal of Intervention and Statebuilding, 11(3), 359-381. https://doi.org/10.1080/17502977.2017.1353755

Yusuf, I.U. (2019). Countering Boko Haram Insurgency and Terrorism : Assessment of Media Campaigns. Global Media Journal, 17(32), 1-11.

Zenn, J. (2017). Demystifying al-Qaida in Nigeria: Cases from Boko Haram’s Founding, Launch of Jihad and Suicide Bombings. Perspectives on Terrorism, 11(6).