Understanding Vulnerability to Violent Extremism: Evidence from Borno State, Northeastern Nigeria

ABSTRACT This paper analyses the links between socioeconomic concerns and one of the most significant conflicts in the world, the Boko Haram-led insurgency in Northeastern Nigeria. In doing so it centers group dynamics for analysis of how women and youth constituencies intersect with vulnerability to violent extremism. It offers sophisticated quantitative analysis of new and original gender- and age-disaggregated survey data, with over 80% female respondents. The paper finds that while poverty can influence vulnerability to violent extremism, women and youth constituencies interact in particular ways with structural factors and certain youth constituencies exhibit lower propensities to violence.


Introduction
Understanding the factors that drive violent conflict is a longstanding concern across academic, policy and practitioner communities globally. Attention to this subject is underscored by the need for complex responses to manage the implications and outcomes of violence as well as address its root causes. Within violent extremism studies, the expectation is that understanding and addressing root causes is important for resolution and longer-term peace and reconstruction. The UNDP and the World Bank emphasize conflict prevention through comprehending and managing economic, political and social dynamics. 1 This paper is concerned with understanding the socioeconomic dynamics that underpin one of the most significant conflicts in the world, the Boko Haram-led insurgency in Northeastern Nigeria. Meagher 2 has been clear that "addressing the Boko Haram insurgency . . . requires policy makers to look beyond western security templates . . . to grasp the underlying causes of what is primarily a Nigerian conflict." Since 2009, the militant group, Jamaatul Ahlul Sunna li Da'wati wal Jihad, known widely as Boko Haram, has been responsible for violent attacks on civilians, state and non-state security actors and destruction of physical and social infrastructure across the public, private and charity sectors. According to the Armed Conflict Location and Event Data Project, 3 over the five-year period of 2014-2019, Boko Haram has been responsible for 2,800 incidences and over 31,000 fatalities. The United Nations Refugee Agency (UNHCR) reports that the Boko Haram insurgency has displaced nearly 2.4 million people in the Lake Chad Basin with over 2 million people internally displaced in Nigeria and about 300,000 Nigerian refugees. 4 The paper asks the central question, to what extent does violent extremism interact with poverty as a motivator, within youth and women constituencies? Against this background, it has two objectives. The first is to critically interrogate arguments on drivers of violent extremism through a quantitative datadriven approach. Schuurman 5 has noted that academic scholarship tends to focus on the reexamination of secondary literature and eschew quantitative analysis as well as the examination of new data (post-2007). This limits empirically grounded insights. The second is to center group dynamics for analysis of how women and youth categories intersect with vulnerability to violent extremism. It challenges generalized assumptions through attention to the perspectives of these constituencies on poverty, the notion of the collective and the distinctness and particularities of social groups.
The paper makes a unique and cogent contribution in its rigorous analysis of new quantitative data to understand and explain the extent to which socioeconomic factors, in relation to social groups, interact with vulnerability to violent extremism. In doing so, it offers advances across the fields of economic development, conflict, and peace studies. It analyses gender-and age-disaggregated survey data (2019) from the NEEM Foundation in Nigeria on vulnerability to violent extremism in Borno State, Nigeria. This dataset was produced from a 2019 survey of respondents (N = 4932, Mean =age 36.79, 83% female and 17% male). The majority were female respondents thus elevating the analytical focus on women. The dataset focuses on both internal vulnerabilities (including attitudes, belief, and intent) and external vulnerabilities (including social and economic concerns); the paper's analysis focuses on the latter.
The paper finds that poverty can be significant for vulnerability to engagement with violent extremism for women and youth constituencies. It highlights the influence of structural factors that impinge on social categories of women and youth. The paper also offers a methodological intervention in the use of "difficult to access" valuable quantitative data from NGOs working in conflict-affected contexts. This paper is a timely contribution to understand the underlying dynamics of violent conflict in the Northeast as it analyses data at the 10-year mark, since the start of the conflict.
The rest of the paper is divided into four sections. The second section presents a review of the empirical and theoretical debates on linkages between socioeconomic factors, notably poverty, and violent conflict, especially violent extremism. The third section presents the methodological approach of quantitative analysis of survey data from the NEEM Foundation. These include assessment of the level, degree and direction of association and correlation/covariance between variables in the data set that measure the interaction between poverty and vulnerability to violent extremism with attention to age and gender categories. The fourth section critically integrates the data analysis with the key conceptual and theoretical arguments on the interactions between poverty and violent conflict, and therein violent extremism. It focuses on key constituencies of youth and women. The fifth section concludes and shows how some constituencies exhibit a lower propensity to violent extremism, despite the influence of poverty. This is a critical finding for exploring nonviolent forms of exercising agency to support peacebuilding and reconstruction.

Violent extremism and root causes: theoretical, conceptual, and empirical debates
There are a range of cross-disciplinary factors across economic, political, social, and environmental concerns that have been put forward as influencing violent extremism. Debates on the root causes of violent conflict and insurgencies in general and in the case of Boko Haram have spanned themes across disciplines, including high poverty rates, 6 7 8 high unemployment rates, 9 10 socioeconomic inequality, 11 12 13 high rates of illiteracy, 14 15 extremist ethnoreligious ideology and fundamentalism, 16 17 18 political segregation, minority exclusion and lack of opportunity 19 20 21 22 and the confluence of development challenges and climate change. 2324 This paper is concerned mainly with the factors that can be linked to a socioeconomic logic with focus on the theme of poverty and how this may interact with vulnerability to violent conflict and violent extremism. To be sure there are concerns about privileging economic factors that can reinforce economic determinism in understanding the factors that drive conflict. It is thus important to consider these in concert with social, environmental, and political factors that attend the occurrence of violence. Although this paper focuses on economic concerns, it does so cognizant of the ways that philosophical, psychological, environmental, social, political, and economic considerations are defined, refined and sustained by one another. It analyses socioeconomic factors as external vulnerabilities, as only part of the picture. The wider data set shows the complexities of comprehending vulnerability to violent extremism in the interdependencies between external and internal vulnerabilities that comprise belief, ideology, intent, grievance and cognitive style among other factors. 25 The focus on the pivotal issue of poverty is for two reasons. First is the attention to this theme as critical to violent conflict including violent extremism, especially in analyses of the Boko Haram-led insurgency. The second is that addressing poverty is pertinent to the success of peacebuilding interventions, including Nigeria's North East Development Commission. Bjorgo and Silke 26 argue that "understanding what has caused the violence in the first place will help identify effective solutions to current conflicts."

Empirical concerns: analyses of socioeconomic factors across constituencies
The most persistent arguments around the underlying drivers of the Boko Haram insurgency have centered on socioeconomic factors including poverty and (un/under) employment. From Table 1, for Borno state, the 2010 dollar per day poverty rate is lower than the national average while the 2018 multidimensional poverty index (MPI) measure is higher than the national average as is the unemployment rate in 2011. 27 Some estimates place poverty at 76% for Borno state. 28 Poverty and unemployment in the region tend to be higher than the national average and remain important factors in this context. In fact, the MPI is almost entirely higher than the national average in the North of Nigeria and is entirely lower in the South of Nigeria. 29 The high level of poverty, income inequality and unemployment, in the Northern region especially among youths, amidst the excesses of elite groups, has been identified as the perfect pool for attracting foot soldiers for this conflict. 30 31 32 The key message across these papers is the extent to which poverty and unemployment make Northern Nigerian youth increasingly susceptible to recruitment to violent extremism as part of the Boko Haram insurgency. Analyzing these dynamics in Borno state provides a framework for pursuing similar analyses in the region.
Studies including Botha and Abdile 38 and Iyekekpolo 39 argue that other factors play a role in violent extremism and poverty cannot be treated as an isolated root cause. Some note the need for focus beyond individual conditions. Mustapha and Meagher 40 challenge the contention that poverty is not a central factor in the rise of Boko Haram. They argue that poverty be considered in more complex ways beyond individual deprivation with attention to group marginalization that is as much about macroeconomic disparities as individual experiences of want. This article engages this call for attention to the collective.
In considering the impact of youth (un)employment conditions (and connections to poverty) on the conflict, Botha and Abdile 41 suggest that young people are vulnerable and can be recruited because Boko Haram presents opportunities for employment. Explaining the participation of the youth constituency in the insurgency, Adelaja et al 42 surveyed public opinion from communities affected by the conflict on the root causes of terrorism and their objectives. They find a widely held perception that the Boko Haram members are typically unemployed and are motivated to join due to poverty and economic related challenges.
The experience of violent extremism in the Northeast has exhibited continuities and discontinuities to gendered violence. There are gendered constructions in debates on women and violent extremism. These depict men as violent and women as victims of violence through abduction, rape, and killings with limited contemplation of their participation in more active roles. Agbaje 43 focuses her analysis on the objectification of the female body within the Boko Haram insurgency. These important narratives are being contended given increased female association with, and participation in, violent extremist groups, including Boko Haram. Zena 44 highlights how women have come to sympathize with violent extremist groups to advance their rights in an environment engrossed with poverty, illiteracy, and early marriage. In a 2017 BBC interview, the NEEM Foundation Director, Dr Fatima Akilu, notes the social and economic power that some former wives of Boko Haram combatants have experienced vis-à-vis conditions within their home communities. In 2019, International Crisis Group was vocal about this complex conflation of women as militant, sympathizer and forced accomplice and how this can undermine peacebuilding efforts. 45 Zena 46 notes that "women should not be represented and analyzed only as bystanders and victims . . . . a significant number of women play a crucial role within the group as messengers, recruiters, smugglers, as well as suicide bombers." A range of arguments are put forward to explain women's participation and association with Boko Haram including structural inequality and their marginalization as well as possibilities for improved access to social and economic resources. 47 48 There is reference to pervasive economic challenges due to poverty and social concerns including early marriage that see women seek to advance their economic rights by participating in the Group. 49 50 Some authors challenge the economic deterministic approach to explaining violent extremism in the Northeast. They include ideational, social, and political dynamics, that underpin the Boko Haram insurgency. 51 52 53 While this paper addresses poverty, it is attentive to the complex social and political realities that are also at play. These are reflected in its attention to the constituencies of youth and women.

Theoretical and conceptual concerns: relative deprivation and societal dynamics
This paper centers social constituencies, of youth and women, and the logic of these collectives as a basis for interaction with violent extremism. It does so with reference to how certain socioeconomic factors might underpin vulnerability to engagement with violent extremism. The most prominent theoretical lens that has been used to explain the causal factors for the Boko Haram insurgency is that of relative deprivation that describes individual or societal experience of discontent when being deprived of an entitlement. 54 This has been noted as a driver of violent extremism when linked to higher levels of poverty and inequality in the North especially vis-à-vis the South. In fact, former Boko Haram leader, Mohammed Yusuf noted the relevance of "conspicuous consumption and opulence of the Western educated elite" amidst abject poverty. 55 Bjorgo 56 has been attentive to nuanced discussions about the root causes of terrorism and violent extremism, including poverty. This work articulates structural causes and factors including relative deprivation and the impact of political and economic dynamics at macro levels on segments of society, facilitator causes in the means of participating in violence, motivational causes rooted in how structural factors interact with responses from groups (and individuals) and triggering causes with attention to events and occurrences that precipitate violent responses.
Discussing the theory of frustration-aggression, Breuer, and Elson 57 highlight cognitivist perspectives that relay frustration with delayed, reduced or removed rewards as well as non-attainment of goals that can then trigger aggression. The contexts within which frustrations can be articulated at societal levels (in addition to individual levels) include "severe economic recessions, a lack of or restricted access to resources, or systematic and/or institutional discrimination against certain groups." 58 There are clear synergies with the logic of relative deprivation, particularly with reference to societal dynamics.
Relative deprivation focuses on measures of economic, political, or social deprivation that are relative, rather than absolute. 59 Significantly the idea of relative deprivation is based on perceptions. Gurr 60 builds on the logic of frustration-aggression and considers relative deprivation as a potential driver for collective violence based on its depth or intensity with some emphasis also on societal dynamics. The logic of deprivation here is connected to the idea of attainable goals, already noted. Specifically, Gurr 61 considers relative deprivation as a gap between what is deemed achievable, that is value capabilities, and what is expected which is linked to what people feel entitled to, that is value expectations. There is some synergy between the idea of goals from the theory of frustration-aggression and values from relative deprivation. 62 These theories have been used to explain the causes of conflict in many contexts, including Nigeria. To support their findings about the root causes of the Boko Haram conflict, Ogunrotifa, 63 Adelaja, Labo and Penar 64 and Tayimlong 65 adopt elements and versions of the theories of relative deprivation and frustration-aggression for establishing their claims. In addition, economic deprivation is seen as increasing vulnerability to violence as a means to express grievances and reinforces the conclusions drawn by other scholars like Ewi and Salifu. 66 These contributions have deepened understanding around the motivations for participation in the Boko Haram insurgency across various constituencies including youth and women. They are in line with Cramer's 67 arguments on the significance of poverty as an analytical lens for explaining and understanding violent conflict across a range of theoretical, conceptual, and empirical approaches. Piazza 68 and Abadie, 69 also show the prominence of analyses that hold poverty and relative economic deprivation as significant to understanding root causes of violent extremism.
However, much of the literature on the Boko Haram conflict does not draw on extensive and rigorous data analysis in putting forward these arguments. This can be due to the difficulty in accessing such data. In some cases, it relies on old data from the pre-2010 period that does not consider interactions between socioeconomic factors such as poverty and unemployment and violent extremism. 70 Cramer 71 notes that understanding pre-and post-conflict poverty are major challenges associated with the paucity of available data. This has resulted in blanket (negative) conclusions drawn about entire social categories as has been the case with the youth constituency and limited understandings of the place of women. These outcomes undermine evidencebased policy interventions for addressing the conflict. This is a gap the article addresses in its analysis of new survey data from Borno State. It offers more nuanced findings, with insights on the lower potential participation in violence by specific youth constituencies. This opens opportunities for engaging such constituencies in interventions, as is discussed later.

Analyzing survey data on poverty and violent extremism from Borno state: data and methodological approach
This section presents processes for using original quantitative data to interrogate interactions between poverty and vulnerability to violent extremism. The paper draws on survey data on Countering Violent Extremism (CVE) that was collated by the Abuja-based NEEM Foundation, an NGO launched in 2017. In addition to data-driven peacebuilding research, NEEM delivers mental health and peacebuilding programmes. The Foundation established Nigeria's first CVE Programme.
This dataset was produced from a 2019 survey of respondents (N = 4932, Mean =age 36.79, Standard Deviation = 11.359, 83% female and 17% male) through the Foundation's Counselling on Wheels, its principal peacebuilding and mental health intervention. The survey instrument was deployed to participants of the programme and assessed using the Vulnerability to Violent Extremism Assessment (VVES) Scale. The VVES is a ranked structured tool that builds on the Violent Extremism Risk Assessment 2 Revised and the Extremism Risk Assessment Scale that is used together with clinical and professional judgments to assess vulnerability to violent extremism. 72 It is part of a more complex process of understanding how a range of factors impinge on vulnerability and is thus not conclusive; it is a "gauge of vulnerability in affected communities" with focus on social and economic aspects as opposed to a conclusive measure. 73 The respondents are from communities experiencing the impacts of violent extremism and from which recruitment is believed to occur in one of the core states affected by the insurgency, Borno State, Northeastern Nigeria. They include Dubai, Moramti, Mairi, Kaleri, Old Maiduguri, Dala, Gwenge I, Gwenge IV, Umarari and Wulari.
The VVES features seven groups and sixteen subgroups of variables, ranging from internal to external vulnerabilities. 74 Each subgroup has five items measured on a Likert scale. Internal vulnerabilities are key to understanding motivations through attitudes, fixedness of belief and intent, for instance, and external vulnerabilities to placing protective factors, social and economic and other contextual factors as well as histories. This study focuses on the latter set of factors, with attention to history, social (conformity and group think) and the economic factor subgroups of the VVES dataset. These include the variable items that address poverty: I can join bad groups if there is no work anywhere; If I get too hungry, I can join bad groups; If I have nowhere to sleep, I can join bad groups; Being poor can lead me to do things I don't want to do; Poverty can make me join bad groups; Money can make me join bad groups. The variable items were measured using a five-point Likert scale, ranging from 'Strongly Agree to "Strongly Disagree." The VVES is noted as "not intended to proscribe or criminalize individuals and communities but to assess vulnerabilities in a manner that does not pathologize." 75 A subset of variables was extracted from the VVES dataset to form a new dataset intended for the analysis. These variables had 4932 respondents. Overall, 3556 responses representing 72.1% of the respondents had at least 1 missing value. Only 1376 respondents had no missing data value. According to Madley-Dowd et. al., 76 statistical guidance articles state that data missingness is likely to lead to bias and that more than "40% data missingness should only be considered as hypothesis-generating." 77 Since the dataset missingness is 23.1%, which is above the 10% threshold, it is necessary to address this to avoid biased inferences in our analysis.
It is not uncommon to see large missing data values in social science data. King et. al. 78 estimated from papers and posters presented at the Society for Political Methodology, that incomplete data values averaged between 50% and 90%. 79 Most researchers hardly report missing data, but this can lead to biased results.
From Table 2, missingness for variable items, I can join bad groups if there is no work anywhere; If I get too hungry, I can join bad groups and If I have nowhere to sleep, I can join bad groups, stands at 64%. As a result, the paper's analysis will exclude these variables. It will rely instead on the key variable items which give far lower and acceptable levels of missingness in the dataset. This enables us to carry out our analysis with confidence.
Responses rely on self-reporting by participants. While self-reporting is a dominant approach for collating data on vulnerability to violent extremism it is also subject to certain challenges. Biases that have been reported include "underreporting socially undesirable items, overreporting socially desirable items, interviewer effects, bystander effects, and more." 80 However, NEEM Foundation's broad and long-term community engagement across its intervention programmes in psychosocial support, peacebuilding and education offer a context of knowledge within which to cross-check the veracity of the data. Another critical point is the support that is offered to respondents by NEEM within a very challenging context of violence and vulnerability. This study centers the respondents' reflections and understandings of poverty and how this can interact with the dynamics of vulnerability to violent extremism. Attention to the sense of privation of those affected is highlighted as an important counterpoint to so-called objective measures. This strengthens and deepens gendered understandings of poverty and its impact as well as elevating key voices and experiences in how this is understood. 81 Ultimately, this approach offers a snapshot of how those within these contexts consider the interactions between violent extremism and poverty. The focus on socioeconomic factors and therein external vulnerabilities has variable items that refer to violent extremist groups as "bad groups." Prior to administration, the survey was tested in the community, in affected areas, to ensure a shared understanding of "bad groups" as violent extremist groups, such as Boko Haram.
The data were collated for a different purpose namely for vulnerability assessments. However, some of the survey questions fit into the remits and objectives of this paper, namely on the interactions between poverty and vulnerability to violent extremism. To this end, this paper, therefore, assessed the level, degree and direction of association and correlation/covariance between the variables. The Spearman's rank correlation coefficient (ρ) analysis was employed because the scales of the variables are ranked. This nonparametrical statistical test is privileged for study samples that are not normally distributed over Pearson's correlation and other related metrics for normal or Gaussian distributed samples. The Kruskal-Wallis H test and Mann-Whitney U tests were conducted to determine the nature of the differences between the responses of the participants to the variables by age and gender categories.
The Kruskal-Wallis H Test was conducted to examine if there were differences (significant) in participants' rankings across their responses to the variables, Poverty can make me join bad groups, Money can make me join bad groups and Being poor can lead me to do things I don't want to do, across five age groups (16-25, 26-35, 36-45, 46-55 and 56-65). The attention to the range of age categories is to enable analysis of the particularities of certain social group dynamics, including youth. Notably, the 2009 Nigerian Youth Policy denotes youth as 18-35-year-olds, which was updated in 2019 to 15-29year-olds. 82 Given this recent change, the paper retains the understanding of youth as 18-35-year-olds. The focus on the three variables is to understand respondents' perceptions of poverty and as underscoring vulnerability to violent extremism. Where statistically significant differences exist, a pairwise comparison with significance value by the Bonferroni correction for multiple tests was conducted to control for Type I errors across tests and to also investigate the granular differences in responses to the variables between the age groups.
Comparison between gender categories was confined to two groups-female and male. The Mann-Whitney U test was conducted for examining the differences between the two groups in relation to their responses to the variables. The focus on these groups is to analyze the particularities of women and men as social categories. Where these differences are established to be statistically significant (p > .05), that means the distribution of the tested variables is different across the categories of gender; in cases that are not statistically significant, the distribution of the variables tends to be the same across the categories of gender -that is, both female and male respondents probably highly rated the variables similarly.
We build on previous work such as that of Nelson and Scott 83 who use correlation matrix to assess media response to terrorist incidents. Several scholars have also adopted this method to establish the significance or otherwise of an association between different variables. Morakabati and Kapuściński 84 use a correlation matrix with a dataset with similar characteristics to the VVES dataset to consider how travel is altered by terrorist attacks with attention to personality types. Goodwin et al 85 use this method to examine the predictive power of demographic factors on threat perception and the consequences therein on behavior and relationships. Bloomberg et al 86 use Spearman's correlation to evaluate the prevalence of terrorism on a per capita basis and the relationship between income and terrorism. Weinberg et al 87 also use Spearman's correlation to study the relationship between terrorism and governance. Elbakidze and Jin 88 use the Kruskal-Wallis test to evaluate differences between mean annual terrorism counts of two groups of terrorists.

Discussing interactions between poverty, violent extremism, youth, and women: findings and discussion
To the best of our knowledge, this study is the first to analyze gender-and agedisaggregated survey data on vulnerability to violent extremism in what is the epicenter of the Boko Haram insurgency affected conflict, Borno State. This section interrogates these interactions through analysis of gendered and youth narratives on poverty and violent conflict, with findings from the Spearman's correlation analysis of the variables, the result of the Kruskal-Wallis test, the two-tailed Mann-Whitney test as well as the descriptive statistical analysis of the variables by age and gender groups.

Poverty and violent conflict
Poverty has been put forward as a causal factor for the conflict, often based on scant evidence. Erhardt and Sani Umar 89 note "the ubiquitous problem factor of poverty" as a push factor for Boko Haram members. Gendered analysis of poverty in Nigeria, including in the Northeast, shows a shift over time from 1980 when female headed households presented higher rates than maleheaded households to a reverse pattern by 1996. 90 91 However, this pattern of lower rates for female headed households is reversed as the household sizes increase. 92 Despite the pattern on female-headed households, Nigerian women are noted as "experiencing poverty more ceaselessly than their male equals." 93 Furthermore, the poverty status of women is influenced by their positioning in male-headed households; for instance, as of 2013, only 15% of total households in Nigeria were female-headed. 94 This is a critical factor in Northeastern Nigeria where women's access to social and economic resources is heavily mediated by their relationships with men across various spheres. 95 These factors are important for a more complex gendered reading, especially around debates on the feminization of poverty that move beyond static headcounts of male vs. female to understanding the dynamics of how poverty changes over time and the place of women therein. 96 The findings from the data analysis show interactions between poverty and vulnerability to violent extremism in Borno State. Most respondents indicate that poverty can increase their vulnerability to participating in bad groups as seen in Figure 1. In this regard, most respondents also note that poverty can  see them partaking in activities that they would rather not engage in, from Figure 1. Notably, more than a quarter of respondents strongly disagree. There is also a high correlation between these variables that is statistically significant, from Table 3. This is reinforced by research showing that beyond poverty, there is an absence of articulated "serious interests" across men and women in joining the Group. 97 Cramer 99 contends that across the ideological spectrum, interactions between violent conflict and development factors, from structuralist to economic individualist logics, poverty features prominently as a causal factor. But in recognizing the ways in which some very poor societies do not erupt in violent conflict he highlights the significance of mobilization toward conflict.
Here we can reflect potentially on the group dynamics that can underpin such mobilization, as with youth and women.
There is a positive correlation between variables 1-3. This implies that respondents have similar responses for variable items, Being poor can lead me to do things I don't want to do, and Poverty can make me join bad groups and Money can make me join bad groups to quite high degrees. The place of "groups" also signals Cramer's argument about mobilization toward violent conflict.

Women and participation in violent conflict and violent extremism
The discussion on group dynamics in this paper engages the place of women as actors in the Boko Haram insurgency. From this unique data set that constitutes mostly female respondents, there is an opportunity to consider the interactions between women, poverty and vulnerability to violent extremism.
Variables 1, 2 and 3, from Table 4, all feature insignificant Mann-Whitney U test results. This means there is no significant difference in the participants' responses to these variables between male and female respondents.
This data analysis shows that women's responses on agreeing and strongly agreeing that Being poor can lead me to do things I don't want to do as well as make me join bad groups (Variables 1 and 2) have higher mean ranks than men with no statistically significant difference in the distribution of responses from Table 4. This outcome might be linked to the large proportion of female respondents. Zena and 100 Matfess 101 suggest that women's participation in violent extremist groups is linked to navigating poverty and structural inequalities and accessing economic and social resources. On average, women show some disagreement in similar patterns across the variables. There are higher levels of agreement on the variables, i.e. Being poor can lead me to do things I don't want to do, and Poverty can make me join bad groups and Money can make me join bad groups in Figure 2. Over a quarter of respondents also strongly disagree across the variables. These patterns are in sync with their male counterparts.
There is a synergy between these findings and analyses of female participation in violent conflict and violent extremism. This is with consideration of how structural economic, social, and political dynamics locate women, as a segment of society, in a disadvantaged position. Usman et al are 102 clear, for instance, on the extent to which social norms of women's dependence on men can "shape their limited access" to economic opportunities. This speaks to Bjorgo's 103 points around the links between structural factors to vulnerability to violent extremism.
Participation in Boko Haram is presented as potentially providing economic (and social) opportunities in a context that is defined by structural economic and social challenges. This can be read as attempts to manage gaps between value capabilities and value expectations within relative deprivation. In particular, the data analysis on the dissonance between willingness and participation in violent extremist groups on account of poverty can in some ways be connected to ICG's 104 assertion that women have lived with the insurgents in fear but have yet gained access to economic and social resources. There are of course reflections on degrees of willingness as a complex notion in this regard. Usman et al 105 highlight women being coerced, compelled, pressured, encouraged, and convinced and highlight the challenge with distinguishing between these categories across women as victims and in more active roles.
Critically women's participation in the Boko Haram insurgency is documented as spanning various areas such as domestic work, spiritual guidance, communications, intelligence, recruitment and training, explosive experts, foot soldiers, suicide bombers and leadership (Botha and Abdile 106 ; Ehrhardt and Sani Umar 107 ; Usman et al. 108 ). These categories are nuanced in terms of the extent to which they might also be considered victims of the insurgency due to the conditions they work in, the harm they face, the degrees of willingness at play and the fatalities that may result. Yet the diversity in participation challenges the extent to which there is a singular narrative, including the degree to which poverty constitutes a driving factor. Usman et al 109 note that not all female participants in Boko Haram are necessarily radicalized. The data analysis does not offer more granular details on the motivations of the survey respondents. Getting a sense of this may enable deeper reflection into identifiable group dynamics that underscore vulnerabilities to participation in violent extremist groups. How these group dynamics may interact with individual motivations is an element of what Bjorgo 110 describes as motivational causes of violent extremism. It is vital to note that while there are structural factors that locate women in particular positions, this is not a monolithic category and attention to the interplays between individual and group motivations is an important consideration.
Despite the welcome attention to critical examination of women's participation in violent extremist groups there is a risk of binary readings of this constituency as either victims or perpetrators. It is important to challenge these dichotomies and allow for greater complexity in analysis. Ehrhardt and Sani Umar 111 report the transition of forced female participants that emerge as strong advocates. Usman et al 112 draw attention to considering women in counter-insurgency roles beyond their characterization as victims and active participants. It is also possible to consider fluidities that can challenge a strict trichotomy across these categories.

Youth as a constituency and participation in violent conflict
Onuoha's 113 attention to youth as a constituency offers the space to reflect on the critical factor of group dynamics as an intrinsic element of violent extremism. Ismail 114 notes that youth constitute the largest proportion of participants in violent extremism although these are a marginal proportion of the overall youth population in these contexts. Indeed, it is now well acknowledged by the policy community that most of the youth are not involved in violent conflict or extremism. 115 Umar Sani 116 argues for the need to move beyond methodological individualist approaches to understanding the significance of poverty as an underlying factor in violent extremism. This requires greater complexity with the engagement of notions of group marginalization. Focus on the dynamics of groups as an element of conflict can be understood in relation to Stewart's 117 and UNDP 118 arguments on horizontal inequality and the challenges wrought by intergroup (across age-groups as well as ethnic groups, among others) inequalities. It also links to Bjorgo's 119 attention to structural and motivational causes of violent extremism. Against this background, Cramer's 120 point on the importance of mobilization as critical to violent conflict is pertinent in helping to understand the path to participation. Ismail 121 has reflected on the ways in which youth participation in violent extremist groups can be understood to "reverse their (shared) alienation from mainstream society and formal processes" including employment. Essentially these can be seen as attempts to narrow the gap between value capabilities and value expectations.
For female youth specifically, Usman et al 122 discuss how participation within Boko Haram may offer the opportunity for marriage to male fighters. This can be a key route to adulthood given how women's access to social and economic resources may be mediated through their relationships with men and especially husbands. These narratives speak to youth as a particular constituency with specific characteristics. Honwana 123 highlights youthhood as a social and cultural category with a segment of the population waiting and seemingly stuck in a state of transition that results from complex interactions across relationships, and practices as well as wider structural factors including policies, politics and law. The prominence of young people in the age groups of 16-25 as well as 26-35-year-olds linking poverty to vulnerability to violent extremism reinforces some of the ways in which youth have been positioned in the conflict from Figures 3-8. This does extend also to the 36-40 age category who are placed outside the youth category. This calls for caution in the generalizations that tend to be made about the interactions between youth and interactions with violent extremism.
However, the pairwise comparisons in Figure 3 show that there was very strong evidence 125   groups from Figure 6. This suggests the need to reflect on the underlying factors that underpin such patterns. Figures 7 and 8 show the differences between the responses of participants in the 26-35 years age category and those of the other participants are statistically significant on whether Money can make me join bad groups.  Respondents in the other age categories show similar patterns of agreement that Money can make me join bad groups from Figure 8. There is need for specificity and a degree of granularity in discussions as we see those in the 16-25 age category expressing similar patterns of (strong) agreement with the 36-  45, 46-55, and 56-65 age categories that money can make them join bad groups.
It is necessary to consider the agency of the youth constituency in these activities. Reflecting also on the logic of social dynamics through group activity, from Figures 4, 6 and 8, within the youth constituency, data analysis shows respondents in age categories 16-25 agree (slightly) more that Poverty can make me join bad groups and Money can make join bad groups vis-vis Being poor can lead me to do things I don't want to do. 128 Yet, the 26-35 age group stands out with lower levels of agreement on how poverty and money might motivate participation in violent extremism. Ismail and Alao 129 have critiqued the tendency to focus on the economic and security factors with limited attention to youth agency in their participation in violence and especially the capacities for mobilizing based on shared interests. The consistency with which the 26-35 age group showed higher levels of disagreement as well as lower levels of agreement with the variables raises interesting questions about alternative ways in which this agency is exercised beyond participation and mobilization or forced recruitment. Further research is needed to establish what separates the motivations of this specific group from others and how the potential propensity of the group away from violent extremism might be engaged for violence prevention and peacebuilding.

Conclusions
This paper examines contemporary interactions between conflict, security and development factors thus contributing to an interdisciplinary understanding of violent extremism. Using a range of conceptual lenses, it moves beyond individualistic analyses to privilege group dynamics and structural factors across age and gender categories. The paper provides an evidence-based and data-driven quantitative study with the analysis of original and recent data from Borno State in Nigeria that has not been utilized previously.
The paper showcases the value of co-creation of knowledge between academia and non-academic constituencies. This is pertinent especially where it is logistically difficult to obtain data and more significantly where it is critical to recognize the complexity and delicacy of contexts by working with those that are committed to addressing these complexities through long-term engagement and supporting progress in these spaces.
This collaborative effort has enabled the use of methodological approaches that have not been used extensively in this context. As such, the paper offers an exemplar for examining the underlying dynamics that may influence participation in violent extremism through quantitative analysis.
The paper has highlighted three key themes that are pertinent in this discourse. First, is interrogating narratives on the interaction between structural socioeconomic factors, notably poverty, and violent extremism. It shows how quantitative analysis of original data can be used to critically examine established ideas on the root causes of violent conflict. Here we see that poverty is an important factor through gendered analysis. Second the study engages the influence of group dynamics with attention to the youth through analysis of age-disaggregated data. Third and on a similar note, the study is unique in its analysis of survey data that comprises over 80% of female respondents and therein moves beyond a typical focus on men. The datadriven attention to gender analysis shows women and men do not present (statistically significant) different responses on a range of issues despite the structural realities. This indicates the importance of producing data sets that can enable more complicated readings of the place of women.
This evidence-based study offers key lessons that should inform efforts to address violent extremism in three ways. The first is in the importance of poverty, as presented by these constituencies, as a pivotal factor and how this should be centered in interventions. The second related point is the need to be attentive to the group dynamics and therein interventions that engage structural factors that impinge on certain groups. For many women, navigating male-dominated economic and social structures is a case in point. The third is the value of understanding young people as a diverse constituency and being attentive to variations. From this analysis, there is need for further investigation into certain groups that suggest a higher propensity to nonviolence, that is, the 26-35-year age group within a dataset that is dominated by female respondents. It is necessary to explore what potential opportunities may exist for nonviolent forms of exercising agency to prevent conflict and support peacebuilding and reconstruction.