Abstract
Scholars have long debated what is the role of inequalities in organised violence, but the causal mechanisms remain unclear. I argue that mainstream approaches (e.g. deviance, subcultures, grievances, rational choice and Marxism) fall short because they overlook that organised violence is an intensive form of work. These perspectives often exoticise individuals involved in violence for profit or political purposes. Violence is not a deviant feature of social life but an integral part of collective action. By applying occupational lenses, I position those engaged in violence as a specialised class of manual workers recruited in the protection labour market. Scholars can acknowledge the occupational stratification observed before in research about organised violence by understanding violence specialists as workers. This essay explores how the sociology of work can enhance our understanding of inequality reproduction within criminal organisations, guerrillas, armies, police forces, mercenaries and private military companies.
Introduction
Scholars in conflict studies and criminology have an ever-growing theoretical and empirical debate about why humans kill in organisations and how that is related to inequalities. Notably, some are even discussing whether inequalities actually matter. Some argue that these are relevant (Hillesund and Østby, 2023), while others deny it (Vilalta et al., 2023). This debate is rooted in the perception that political or criminal violence increases when income or wealth inequality between classes rises, subsequently producing ‘greedy’ and ‘rent-seeking’ impoverished working-class men (Berdal, 2005; Cramer, 2002). Other scholars argue that violence is the byproduct of horizontal inequalities, that is, some ethnic or religious groups are able to capture wealth and power, and others react against it (Sen, 2008; Stewart, 2008).
Nowadays, this debate has stalled, and the linkages between violence and inequality remain vastly understudied in social sciences, particularly political science and sociology. For conflict and crime scholars, the problem regarding the identification of a causal relation between income inequality and violence depends on how we measure inequality. For example, if we are using correctly the different metrics of income inequality, such as poverty counts or Gini coefficients, or if they must be studied in geographical manner (Cederman et al., 2013; Dixon, 2009). Notably, researchers in this debate are keenly focused on testing whether the increase in youth unemployment in countries with growing populations (youth bulges) can serve as an indicator to predict violence (Blaustein et al., 2018; Urdal, 2006). Still, these explanations assume that individuals living in income-based poverty are more likely to commit violence, join criminal organisations and even instigate large-scale wars.
Nonetheless, as Collins (2008: 35) argued, the overwhelming majority of impoverished working-class individuals do not engage in organised violence, and anyone could potentially commit such acts. However, in modern times, young working-class men make up the majority of those participating in organised violence. This paradox cannot be sorted by only examining the grievances of people experiencing poverty or by measuring income inequality or poverty.
It is crucial to return to the study of the relationship and causal mechanisms between inequalities and violence to understand current conflict dynamics. For example, it is necessary to understand how inequalities shape recruitment intensity. For strategic and political purposes, violence entrepreneurs need not only large armies but also skilled armies. The duration of a conflict also depends on how many individuals are mobilised.
Therefore, it is essential to adopt a historical, sociological approach to explain why some working-class men participate in organised violence. I contend that there is a missing piece, visible to scholars but not fully recognised in these fields: work, not as an independent contextual variable as previously researched, but as a causal mechanism linking violence with various forms of inequality (income, wealth or social mobility). Instead of focusing on a particular inequality measurement, the real challenge is conceptual and theoretical. The sociology of stratification allows us to understand the complex dynamics of inequality as mobility between occupations with different incomes, prestige or social mobility opportunities.
Work is one of the most important spaces where the reproduction of inequalities occurs (Sørensen, 1991). Income inequality between occupations, or stratification, functions in both directions: it keeps some individuals in higher income classes and others in subordinate lower income classes. However, this depends on how societies value the distribution of income or wealth through work. Poverty, lack of social mobility or educational achievement is just an indicator of a broader occupational stratification in which some men decide to work for violent organisations. Nonetheless, as I will argue in this essay, by understanding organised violence as a form of work, we can see that killing is an occupation not necessarily linked to deprivation but to whether violence entrepreneurs recruit certain types of individuals. Nowadays, it is an occupation that is considered socially attributed to the working classes. Conversely, in the past, societies regarded military violence as work for praetorian managerial and ruling classes.
Why does the previous mainstream research have not acknowledged the labour dimension of organised violence? As I will show, classical sociological theories (i.e. deviance, subcultures, grievances or rational choice and Marxism) tend to exoticise individuals performing violence. As summarised by Malešević (2010) and Downes et al. (2016) for conflict and criminology studies, respectively, the mainstream of these fields, even if inspired by neoclassical economics, still depend on the weight of classical sociology to develop models of what they normatively believe is deviant human behaviour.
Conversely, there have been remarkable attempts to formulate consistent theories about organisational violence without considering violence as a deviance in human behaviour, but instead as an integral part of collective action (Blattman, 2022; Bowles and Gintis, 2011; Collins, 2008; Kalyvas, 2006; Malešević, 2017; Mann, 2023; Tilly, 2003). These scholars have gone beyond the older debates that used to argue that impoverished individuals enact ‘irrational’ violence, that they are cultural ‘barbarians’ or that poverty makes them criminal deviants. 1 All of them point out that organisational violence is strategic and serves several purposes depending on where criminal or political collectives operate. More importantly, as Kalyvas (2001) defended, the previous generation of conflict studies conflated the economic conditions and motivations of participants in organised violence with the stated motivations and actual activities of clandestine organisations, thereby promoting the idea that all such groups are essentially looting and non-ideological. Therefore, aggregate measurements of income inequality and youth bulges became a form of evidence for some scholars of rent-seeking non-ideological motivations.
Kestnbaum (2009) argued that, in recent years, sociologists of war have moved beyond the Weberian view of violence as a monopoly of the state and have studied violence as an integral part of social lives. Unfortunately, beyond these efforts, as Walby (2013) argues, the dominance of political science, criminology and economics in this topic demonstrates that violence is still at the margins of sociological discussion. I join the call from these scholars to pivot an examination of the role of sociology in organised violence studies to contribute to this emerging agenda and even contribute from this field to broader debates about solidarity, inequalities, work and crime.
Killing is a challenging endeavour that, in organisational settings, involves labour. Recently, scholars like Hoffman (2011) and Atkinson-Sheppard (2016) have recognised that individuals often consider killing or having the ability to kill to protect or provide services as labour. For instance, they have observed children in guerrillas or criminal organisations in South Asia and West Africa who view their roles as arduous work. Bogliaccini et al. (2024) and Sviatchi (2022) also identified how legal labour markets feed into illicit drug trafficking in Latin America, and Mendoza Rockwell (2024) explicitly argued that participants in drug cartels are part of the working class. In this essay, I take these studies a step further. While they acknowledge the labour component in organised violence, I argue that these individuals must be fully recognised as participants in contemporary capitalism. Therefore, the next step is to conceptualise them as an occupational class, bringing the analytical potential of the sociology of work and stratification into the study of organised violence.
Evidence continues to accumulate in favour of viewing violence as labour. I adopt the ‘labour lens’ metaphor from the study by Atkinson-Sheppard (2023) in the context of organised crime and extend it to develop an occupational labour lens for conflict and criminology studies. Beyond merely identifying violence as labour, I explore how these activities align with research on stratification and occupational markets. From a historical perspective, if we read the findings of previous research carefully, we can glimpse a history of a class and occupation formation (Dingwall, 1983; Thompson, 1963, p. 10): mercenaryism.
Consequently, I propose centring labour into the study of organised violence. First, it is necessary to examine the shadow of classical sociology and economics on organised violence studies to assess why the labour dimension of this phenomenon remains marginal in the scholarly discussion. As I prove, scholars overlooked labour in organised violence because they use this form of labour as an example of acts against social solidarity, posing violence as a disruption. Nonetheless, concurring with the authors cited before, performing violence for organisational purposes is also a form of solidarity.
Second, I introduce some concepts around stratification sociology and class studies to understand violence specialists as an occupational class. Moreover, I will explain why the insights from the field of sociology of work, stratification and class are fundamental to defining the great variety of phenomena regarding organised violence studied so far. Even further, I will shortly defend that studying the stratified nature of violence specialists can shed light on another contentious debate that remains elusive for experts on collective violence: why, in modern times, is organised violence enacted by marginalised young men?
Third, I review the studies that have revealed (explicitly or unintentionally) the labour aspects of organised violence so far by comparing their findings, with particular attention to the divide between those who execute violence in legality (soldiers, police officers, private military agents) and illegality (rebels, criminals, child soldiers). Beyond the legal status of their craft, this is a group of individuals who develop skills that they trade as a commodity in the violence-protection market. They were once named mercenaries before the current forms of labour. However, I will adopt the name given once by Tilly (2003: 35): violence specialists.
For this essay, a violence specialist is a professional who uses violence or the threat of violence as labour power (i.e. commodity). So far, scholars of violence have been focused on the hierarchical superiors of the specialists: violent or criminal entrepreneurs (Tilly, 2003; Volkov, 2002). The violence or criminal entrepreneurs are the ones acquiring that commodified violence for organisational processes. Organisations demand this form of violence for profit (criminal) or to advance other organisational proposes in contentious politics.
Other forms of violence that are not used as protection or waged are not in the scope of this essay. This article does not address everyday incidental tensions that lead to interpersonal domestic violence (Collins, 2008: 4), brawls (Tilly, 2003: 7) or ritualistic violence (hooliganism) (Tilly, 2003: 4). Also, this article does not address symbolic violence (Galtung, 1969). The focus of this article is on intentional and physical harm for organisational purposes, profit or political goals along the same lines argued by Kalyvas (2006: 1), Lessing (2017: 3) and Gambetta (1996: 7). Moreover, property crime (i.e. extractive) or other forms of crime not linked to protecting capital accumulation are not in the scope of this article. However, I hope the insights from this essay are helpful for those fields, as they are useful for this (Letkemann, 1973).
With this conceptualisation, I do not aim to discard other theories of crime and violence, but rather to review them from an occupational class perspective. For the propose of this essay, I am agnostic of the ontological and moral motives of collective violence in which violence professionals participate. In other words, qualifying a violence specialist does not involve denying the ideological or profit motivations of the organisations they participate (Kalyvas, 2001).
My guiding objective is to systematically conceptualise the study of violence as work in class to enable further theorisation to build the necessary causal linkages between violence and inequality beyond the use of aggregate measurements of income inequality or youth demographic changes in the Global South. In particular, I want to emphasise that organised violence has historically been a work alternative for young men. In capitalist modernity, this occupation shifted towards marginalised working classes. In the conclusions, I will suggest some pathways to use occupational analysis and social mobility as ways to re-address the role of inequalities in violence beyond criminalising marginalised young men in the Global South.
Uncovering Work in Crime and Conflict
Even if the studies around organised violence have sprawled widely in political science, economics and criminology, undoubtedly, these have inherited the central tenets from classical sociology and rational choice economics. Malešević (2010) scanned through the war and violence studies and found that although these fields are now conducting more detailed studies of specific organisations (e.g. terrorism), they still replicate ideas that come from the canonical sociological ‘great men’ theories, in this case, Durkheim, Weber and Marx. The same can be said about criminology (Downes et al., 2016). Also, there is an overwhelming influence of rational choice economics based on Gary Becker’s approach to crime (Steinmetz and Pratt, 2024), which is also widely used in rational choice sociology. Indeed, many of the scholars of violence did not explicitly develop their approaches from the biases of classical sociology. Nonetheless, it is essential to identify how their studies resonate in some ways with classical sociological approaches so we can encourage a cross-disciplinary dialogue to expand the study of violence from a stratification approach.
As Connell (1997) discussed, these classical authors’ influence is a historical product of the global expansion of the discipline. Durkheim, Weber, Marx and Becker wrote ‘theories of society’ which attempted to cover all matters possible. The study of violence is a crucial matter because, for these theories, it is a driver of either the rise of the modern state for Weber, the class struggles for Marx, the difference between solidarity and anything else for Durkheim or the lack of deterrence to protect private property for neoclassical economists. Therefore, violence in their theories, although a critical theme, is always a subordinate event to the core ideas these authors proposed. Retrospective pedagogical approaches canonised these theories, and violence studies are one of the casualties of this process.
The anomie underpinnings in sociology have diverted empirical research from the labour aspect of organised violence. Durkheim, Marx and Weber’s theories rest on the presumption that work and violence are antithetical. Some would agree with Connell and adhere to a Southern theory approach to rebuilding the sociological canon for analysing crime and violence. For example, authors like Pereda (2022) have criticised northern criminology because it creates static categories. However, I prefer a historical reinterpretation of classical sociology. Burawoy (2021) showcases how the introduction of W. E. B. Du Bois, which is neither against nor in favour of classical theorists, created a dialogue with classical theory. In this case, black Marxism (concerned by the violent lives that the black diaspora suffered after colonisation) gained its place in the dialogue with the sociology of the north. Nonetheless, as Connell (1997: 1521–1523) explains, these grand theories pushed aside earlier discussions that favoured exploring on eugenics and androcentrism. However, in doing so, they created new blind spots, making it harder to study racism and patriarchy. Although not entirely comparable, work is also part of this list of blind spots in mainstream studies of violence.
In this essay, I do not seek to discard all these approaches but to dialogue with them. I review them under Lakatos’s (1978) epistemological perspective. In my view, in these research programmes, some of their hardcore (their central claims) and protective belts (middle ground derivative theories) were crafted to separate organised violence as a feature of society instead of recognising it as a regular pattern of human behaviour, comparable to other forms of collective action. Furthermore, these programmes also fostered interpretations of how poverty or perceived inequalities push young men into organised violence.
Contrary to these theories, Tilly (2003), Collins (2008) and Mann (2023) have defended that violence is part of the repertoire of social action. Hence, violence is neither human inescapable nature nor something that humans evolved from. Instead, it is contextual to its historical moment. There are four main theoretical premises in main sociological theories about crime and political violence that underscore the labour features of organised violence:
for anomie (deviance) studies, crime and violence are counterforces against social solidarity;
for rational choice and grievances theories, the inherent motivation of violence is purely individualistic economic rationality;
for subcultures and symbolic interactionist perspectives, violence reflects the meaning formed by group interactions and
from Marxism, violence reflects inherent conflicts and class struggles.
Durkheimian anomie deviance-based criminology studies have unfolded into several sub-theories: strains (Agnew, 1992), social disorganisation (Sampson and Groves, 2011; Shaw and McKay, 1942) and institutional anomie (Messner et al., 2008). They all share one theoretical assumption: social norms do not match individuals’ expectations. Anomie as a sociological concept indeed points to the deviation from institutionally prescribed norms (Merton, 1938) and the Durkheimian idea that criminal behaviour is opposed to organic solidarity and the division of labour (Merton, 1934). In other terms, anomie deviance criminology analytically separates work and labour from crime because, in a theoretical sense, they are antithetically used to exemplify solidarity and conflict.
Nevertheless, even if the findings of these theories showcase similar findings to conflict and subcultural approaches in sociology, by doing this analytical separation, my disagreement streams from an overt denial of the work nature of violence. Howard Becker’s (1991) cultural criticism of this perspective is on point. As he argues, these theories enact labelling that later creates theoretical moral prescriptions of human behaviour, a pattern usually found in functionalism. Nonetheless, I agree that clashes between institutionalised mores coded into law and broader social structures play a role in explaining criminal behaviour.
Conversely, rational choice theory recognises rent-seeking as a critical element in crime and/or violence (Becker, 1968). However, rational choice scholars have indeed posed a reductionist view of human behaviour in which all human decisions and actions are based on economic individualistic motives (Green and Shapiro, 1994). Nonetheless, scholars from this research stream recognise the relevancy of monetary gains, wealth accumulation and entrepreneurial behaviour using violence as protection. Remarkably, the mafia’s literature proposed by Gambetta (1996) makes an extraordinary effort to understand extortion as an economic activity without repeating the rational choice overly deductive approach.
In opposition to these two streams of theory, subculture literature aimed to explain working-class participation in violence (Downes et al., 2016). The primary trait of the significant works in this theoretical stream is that criminal organisations create meanings and identities through socialisation rather than economic motivations or just a mismatch with social rules. The seminal literature posits that youth participate in these activities because a collective culture emerges beyond the criminal labelling of adults and legal authorities (Klein, 1997; Matza, 1964). Nevertheless, I partially disagree with these approaches. First, in epistemological terms, ethnographic critical research on gangs tends to pose that each criminal organisation or gang create a system of meanings, and therefore, they are not entirely comparable. In other words, this approach denies some social regularities across cultures. Second, in theoretical terms, although there is comparative gang research (Pyrooz and Mitchell, 2015), the separation of crime and violence as a subculture of alternative meaning to the working class denies the fact that some crime is work itself found in other ethnographies about gangs (Bourgois, 2002; Venkatesh, 1997). Therefore, we can observe several work features within cultural gang research by framing them into an ethnographic analytical point of view by understanding gangs as workplaces in terms of Burawoy (1982).
Subcultural criminology also devolved into conflict criminology. These scholars defended that the subcultures of crime are essentially resistance of the working class to capitalism (Brownfield, 1986; Taylor et al., 1973). Many streams from Marxism critique that the law is a byproduct of the bourgeois class trying to criminalise any act that can threaten the political order and private property. Beyond coinciding or not with this critique, their argument is an alteration from Marx’s actual views about criminal activity. Marx (1852) thought that criminals were essentially part of the extractive and idle lumpenproletariat and a deviation from the working-class solidarity in revolutionary action (Bussard, 1987; Thoburn, 2002). Essentially, another form of criminalising labelling that denies that protection can also be a form of work. Nonetheless, we could expect this from a Marxist scholarship that advocates revolutionary practice. Paradoxically, unlike other theories, the original Marx position recognised criminals as part of a deviant class instead of deviant individuals.
Regarding political violence, Marxism clearly stated two uses of violence: to repress for the state or to promote revolutionary action (Hay, 1999). In this sense, armies, police forces and mercenaries have the role of protecting private property. Interestingly, Moore (1993) recognised the role of class in revolutionary violence beyond the Marxist initial prescriptions. Scholars such as Olson (1993) and Tilly (1985) also claimed that protection from violence is an inherent feature of state-making. In other words, performing violence is work that can either be criminal or political but needed in scale. The initial literature on grievances in political science, posed by Gurr (2015), streamed from this perspective. Hence, even if the Marxist initial idea of crime discards them as lumpenproletariat or conflict criminology as a culture of resistance to capitalism, in Marxism, political violence is performed, either for the state or for private bandits. Violence and protection are acts of work because they seek to protect processes in capitalism, either criminal or by the state.
This review shows how the hardcore and peripheral ideas of main theoretical perspectives in political violence and criminology – inspired by classical sociology – overlooked work as a feature of crime and violence – either by ontological or labelling considerations about those who participate in protection or because of theoretical assumptions about how social organisation arises. Although these theories offer valuable insights into how poverty or inequalities (income, wealth or social mobility) push young men into organised violence, I believe it is vital to adapt them analytically through the occupational lens.
Reconceptualising Organised Violence as Labour
To sort the under-theorisation between work and violence, I propose a conceptualisation of situating violence specialists as an occupational class. This conceptualisation can illuminate empirical challenges in criminal violence and conflict regarding recruitment, mobilisation and dynamics of conflict. In this section, I put forward some notions to study organised violence as labour; for each, some theoretical suggestions come from the sociology of work and class.
Violent Protection as a Labour Market
The demand for violent protection arises from the absence of the state or from avoiding government prosecution (Gambetta, 1996; Schelling, 1971). In addition, as found in literature about rebel and criminal governance, violent protection is widely used to subvert state authority or govern aside the state to protect other activities – rebellions or for-profit activities (Arjona, 2016; Lessing, 2021; Volkov, 2002). Again, being agnostic about the ideological motivation behind the organisations engaging in violent protection for analytical purposes, the critical point is that criminal or political organisations engage in violence that requires workforces to sustain their activities over time. As discussed before, nowadays states are the primary hiring agencies in the labour market for violence protection. States recruit armies and police forces or subcontract them to deploy wars on other countries.
There is a broad debate regarding how and when violence protection occurs and whether it is ‘rational’ or needed. Although some consider violence to be inherently costly and dangerous for any organisation, Snyder and Duran-Martinez (2009) argued that organisations demand violence protection when other forms of nonviolent governance mediation have not functioned in illicit or licit activities. Essentially, violence is contingent on the conflict dynamics. As Gambetta (1996: 41) argues, criminal entrepreneurs need to make their violent acts to be credible racket organisations. In that sense, the skills a recruiter has to look for in the violence market align with each organisation’s necessities to make their threats believable. Even if entrepreneurs do not use violence, violence threats need to be credible by showing the capacities of the specialists.
We can follow four theoretical forms in which violent protection markets are formed by:
sustaining a social conflict in which groups sort tensions through organised violence (Collins, 2008);
governing populations or annihilating them (Kalyvas, 2006);
as a dispute-settling mechanism (Andreas and Wallman, 2009; Gambetta, 1996) and
to protect property and smuggling (Skaperdas, 2001).
Streaming from these perspectives, we can inform patterns of recruitment and mobilisation as a demand for labour. Hence, there is labour power offered in the protection market, and the scale of mobilisation is in part explained by the capacity of the organisations engaging in protection or violence to attain sufficient labour power for their strategic goals. The demand for labour power is inherently related to threats from opposing organisations. Violence entrepreneurs demand this labour as a commodity in the conflict. In other words, income or social mobility inequalities are only relevant to explaining violence if there are vacancies that violence entrepreneurs need to fill to maintain their activities.
For example, guerrillas need intensive mobilisation to remain in combat against the state, or states recur to conscription when needing more labour power in total war. Scholars on violence or crime have been registering processes of labour matching: a process to form voluntary exclusive relations, in this case, for employment (Coleman, 1991). Studying this involves identifying the labour-matching process, which requires us to indicate if we are indeed observing some agency from those recruited by organisations based on violence. As recognised in the literature on youth studies, young people in marginalisation have a bounded agency (i.e. constrained choices) rather than an entirely subordinate position (Evans, 2007; MacDonald and Marsh, 2005). The classical Marxist framework considers this type of work as a result of false consciousness. Beyond a moral prescription from an analytical perspective, humans choose to use their labour power in professional killing.
Violence Specialists as a Professional Occupational Class
I adopted the notion of violence specialists by Tilly (2003: 30) as the ones deploying violence in an occupation because it is a specialised and differentiated activity within the violent protection market. As argued from a neo-Durkheimian interpretation of work and class, occupations are linked to function within a productive process; they require the use of skills that specialists learn, and there is some relational wage-salary recognition of that skill set (Erikson et al., 1983; Grusky and Galescu, 2005: 61–62). The International Labour Organisation (ILO) International Standard Classification of Occupations has recognised the legal violent protection markets (ILO, 2008). ILO defined violence work as ‘protective service workers’ and ‘armed forces occupations’.
Therefore, killing in organisational settings, based on profit or wage relations, has distinctive characteristics as an occupational activity. Performing violence with purpose in long-term organisational terms requires physical strength, learning the use of weapons or the body and situational knowledge about how to deal with combat, brawls or prosecution. States’ legal frameworks have recognised the necessity to train personnel for legal proposes in the use of guns and escalation techniques.
As an occupation, violence specialists have a repertoire of activities that are performed in many types of organisations using violence: violent rituals, destruction, attacks, sometimes brawls and opportunistic ransom (Tilly, 2003: 15). For Collins (2008), this repertoire of violent acts depends on the situations that foster fear or escalation. In criminal violence, violent conflict, rebellion or criminal governance situations, competing factions or state prosecution fosters the need to use it and has specialists perform it. Therefore, it is a highly situational and organisational necessity to confront perceived risk or actual attacks.
Hypothetically any human can hurt, kill, kidnap and threaten others; these skills in armed organisations are fundamentally driven by how they use them for strategic goals and are later taught or learned by doing. They range from the handling of arms; de-escalation techniques; repression of unarmed people; violent interrogations; protection of critical installations, materials or key personalities; sexual violence as a weapon or massive violence that can amount to extermination in cases of genocide.
Most importantly, as Gambetta (1996) and Collins (2008) noticed, sometimes violence can get out of the hands of organisations, leading to retaliatory and costly cycles of violence. Therefore, an essential feature of collective violence is enforcing discipline to control members. In the case of the Sicilian mafia recruitment reviewed by Gambetta, a necessary feature of why someone gets into the mafia is not only the ability to use weapons; it also entails if the organisation sees whether this person is trustable with information and weapons.
Violence specialists, as in other occupations, hold status and prestige in neo-Weberian terms. As noted by Andrzejewski (1998[1954]), for the military, prestige is attached to the capability to use violence as power, which defines how soldiers become influential members of society. Scholars have registered this for other types of violence specialists. As Tilly (1985) argued, militaries and bandits can rule polities by force and overthrow elites. As Lyall (2020) showed, wealthy classes were initially interested in joining the army to become the ruling class until the era of massive popular conscription posed the necessity of civilian control of the armed forces. Working for criminal organisations also transmits prestige. For example, Gambetta (1996) argues that Sicily’s Mafia could efficiently operate in that region because they cultivated social trust and prestige. Prestige is also a crucial motivation for young men who get into gangs (Anderson, 2000).
As a consequence of the previous arguments, if we follow the class schema of Erikson et al. (1983), it is possible to define violence specialists as skilled manual workers. Therefore, we can infer that, within the protection market, there are three ways in which they are part of subordinate classes. First, as part of highly structured organisations (the military, police forces, guerrillas). Second, as contractors for less-structured forms of work (mafias and criminal organisations). And indeed, third, as contract killers for hire on a case-by-case basis (Hart, 2005).
In summary, regular features of subordinate work also occur for violence specialists regardless of the legality of their activities. Even more, like for militaries, guerrilla members require wages, food and housing like any other subordinate worker. Their ideological affiliation does not entail a renunciation of their basic material needs. Indeed, we only see absolute renunciation of life and basic needs in extreme cases of suicide missions (Gambetta, 2005).
Violence specialists are also a byproduct of stratification. Noted by the criminology culturalists and Marxists reviewed before, the prominence of working-class men within criminal organisations, military forces and police forces is not unintended. Violence specialists are a highly stratified occupation, once of the wealthy elites and later displaced by the marginalised (Kestnbaum, 2009). The regular features of occupational mobility theory apply to violence specialists. For example, there are criminal behaviours passed from fathers to sons. In stratification terms, there is an intergenerational transmission of the occupation (Besemer and Farrington, 2012).
Clearly, the violence specialists’ occupation has been predominantly led by young men, making it a masculinised gendered occupation (Bradley, 2015). As Adams (1983) argued, there is no inherent biological sex difference regarding aggression or physical strength that justifies that women cannot participate in war. Instead, organised violence is a gendered stratification byproduct of the broader hierarchical, patriarchal relations because violence is an activity that provides prestige and power. As a consequence, as Cohn (2013) defends, institutions and individuals in war have justified the expected exclusivity of organised violence towards men. As evidence of the stratified dynamics of this gendered occupation, the massive conscription of young men during World War II led to the incorporation of women into the United States (US) workforce (Goldin, 1991).
Analytically, we can understand their recruitment as job searching or networking with the concepts from matching theory in labour markets (Sørensen and Kalleberg, 2001). Other aspects will play into the recruitment into violent organisations: gender, salary premium, personal attitudes, age and peer pressures. These are features scholars cited here have found for those participating in this protection market, but more generally in labour markets and inequalities literature. As Sørensen (1991) defended, the usual indicators from the neoclassical theory of labour matching that identify are also consistent with research on stratification (essentially, labour mobility within occupations) and, more broadly, how markets operate in capitalism.
Comparative Occupational Lenses on Violence Specialists
As Gerring (1999) once argued, the social sciences persistently face semantic obfuscation due to the excessive number of concepts labelling ostensibly the same subject. In particular, as I will review the literature of several violence specialists, there is considerable conceptual fragmentation that leads to unfamiliarity and over-differentiation. There are many specific types of specialists, labelled mainly by the organisations in which they participate. Each group of authors has identified similar behaviours and patterns, yet they present them as unique. Therefore, when one reads different bodies of literature, they may seem incomparable. As previously argued, researchers have conflated the socioeconomic conditions of participants in organised violence with the nature of their organisations, suggesting that specific measurements of inequality explain the surge in violence.
Nonetheless, we can use occupational labour lenses to uncover patterns of recruitment, mobilisation, disputes and work performance typically found in the sociology of work for violence specialists, regardless of their type of organisation or the (il)legality of their craft. The following literature may not recognise work as a feature of violence. Still, many identified those inadvertently, but most do not feature them as an occupational class.
It is not casual that the literature about conflict and crime discusses the supply of relatively young, marginalised, distressed men as a recruitable workforce for violence (MacDonald and Marsh, 2005; Urdal, 2006). That is a consensus. However, the aggregate cross-national comparisons overshadow how violent organisations recruit these young men. As both Appy (1993) and Lyall (2020) documented about the formation of modern armies, modern governments began to push for the recruitment of young working-class men due to the increasing pressure to ramp up the numbers of soldiers in combat. Contemporary militaries have transitioned from their aristocratic roots to massive working-class recruiters. Elder (1986) showed that massive conscription became a work pathway towards adulthood for many young men during the Vietnam War. In another case, the massification of war since the Napoleonic Wars transformed soldiering as a social mobility alternative for young exprobrated French men (Parker, 1976). The parallel between massive industry and armies is unavoidable. Marx and Engels (2010[1848]) once remarked: ‘Modern Industry has converted the little workshop of the patriarchal master into the great factory of the industrial capitalist. Masses of labourers, crowded into the factory, are organised like soldiers’.
The recruitment of violence specialists is indeed highly gendered and skewed towards men. The working culture is another feature of the social construction of aggressive masculinities. In his seminal work on masculinities, Messerschmidt (1999) noted that violent practices in crime, sports, the military and other occupations involving physical strength are the basis of militarised masculinities. Indeed, for adolescents and conscripts, entering militarised masculinities is a pathway for them to adulthood (Messerschmidt and Quest, 2024).
Researchers on gangs and drug trafficking showed the same militarised masculinities as a way of recruitment and socialisation. García Reyes (2022) found that former hitmen for Mexican drug cartels saw their activity as a performance of their masculinity, coolness and enacting true ‘manhood’ that their fathers never taught them when they were children. Equally, Baird (2012) found that Central American gangs provided a workspace for young men looking for masculine roles to enhance their standing in their communities. Similarly, Duncanson (2009) found that peacekeepers in Bosnia Herzegovina saw their masculine military work as duty of ‘warriors’ and as part of their patriarchal perception of women as passive victims in conflict.
In the case of political violence, as Podder (2011; 2015), Hoffman (2011) and Denov (2010) argue, there is evidence that children in West Africa decide to form part of the ranks of guerrillas as a form of work, not only as a product of forced labour. Nonetheless, as they argue, the first wave of scholars and policymakers ignored this evidence in child soldiering studies because the moral assumption that any activity performed by minors is inherently exploitation, influenced by the discourse of children’s rights, shaped the academic and activist conceptions. Also, Marx (2008[1867]: 294) equated child labouring to black population slavery and with women as human labour trade for gangs. However, even in those conditions, Marx noted that children in Bethnal Green, then an industrial neighbourhood in London, were also waged.
However, the analytical sociology literature (Rolfe, 2011) shows that the labour market (regardless of legality) already constrains choice and free will, even in conflict scenarios. Recently, Atkinson-Sheppard (2016) found that gangs in Bangladesh enrolled willing children into their ranks. As she argues, children saw their work in crime and racketeering as alternative forms of employment. Recently, Sviatschi (2022) found that it is more likely for young men in Peru to be recruited by criminal organisations and incarcerated if they were working illegally in coca farming when they were younger. The same trend has been found in Colombia (Hurtado et al., 2023), Mexico (Zepeda Gil, 2024) and Uruguay (Bogliaccini et al., 2024).
Duran-Martinez (2018: 14–19) documented how criminal organisations in Mexico and Colombia acted as contractors of violence specialists in their respective drug wars. In some scenarios, these organisations recruit young men directly into their ranks (insourcing). At the same time, in other cases, they pay independent local gangs to execute violent labour (outsourcing). In the first case, the criminal organisation will adopt the normal wage relations of their violence workers with military hierarchical traits. In the second case, the local gang will pay wages to their members after being compensated by the more prominent criminal organisation. It is essential to highlight that even if gangs conduct violence, there is a significant difference in performing it as the primary skill for their trade or as a way to protect themselves.
Both Venkatesh’s (1997) and Bourgois’s (2002) ethnographic research show that gangs recruit within youth networks ingrained in underground smuggling economies, first for smuggling activities. Then, some of them will rank up in the organisation’s violence-based activities. Suppose we understand violent organisations as workplaces or industries accumulating capital. In that case, these functional changes for members of these organisations are experiencing occupational mobility.
Violence entrepreneurs also recruit based on skills, profiles and social capital. As the literature on private military companies soldiers shows (Berndtsson, 2012; McCoy, 2012), they are professionals who are offering their experience after working in governments (armies, police forces) to these new mercenary companies when states do not want to intervene in foreign wars directly, paradoxically reinstating a job market previously abolished by the formation of conscription armies. Also, the literature about rebels and gangs has coincided that those organisations engaging in violent protection both craft profiles and recruit based on their needs in numbers (Densley, 2015; Gates, 2002). Scholars like Densley (2012) have noticed that urban gangs use different techniques to signal their preferred type of member. Alternatively, guerillas have attracted young children into their ranks by offering peer socialisation (Horgan et al., 2017).
It is vital to notice that not all members of these organisations perform violence or might have to perform it occasionally to protect themselves. For example, police officers promoted to managerial positions might not enact violence on a regular basis. However, organisations train them in case it is necessary due to the nature of their duties. Alternatively, maybe a lorry driver who smuggles drugs through a border holds guns in case police or other gangs try to capture his cargo. Also, any organisation has administrative positions that do not involve violence. The isolation of administrative workers from violence is more likely in licit organisations such as police or military administrative officers, but maybe less for drug cartels or guerillas doing money laundering accounting.
Performing violence is a skilled occupation. Hart’s (2005) research on hitmen echoes this. These individuals saw their work – killing by hiring – as entrepreneurial businesses, conscious of how their ability to kill – physical training, psychological endurance and knowledge on the use of weapons – is tradable. A similar activity, mercenaryism, was socially accepted before European modernity and later displaced as an amoral profession. Avant (2000) argues that this occupational activity was absorbed by governments after princes and monarchs faced military defeats with mercenaries. The control of violence specialists is a crucial topic in military and mercenary studies. Machiavelli (2008[1532]), in other terms, knew that mercenaries are owners of their labour power and capable of betraying polities.
The previous examples show that performing violence is also edgework. As Lyng (2004) argued, some men participate in crime because they desire to perform risky work. Scholars studying sex workers and dangerous sports professionals formulated the notion of edgework as an alternative explanation to rational choice explanations about participation in dangerous labour or crime (Bengtsson, 2013). The neuropsychological literature on crime also shows that risk-taking is an essential marker of humans performing violence (Séguin et al., 2018). For example, as Monaghan (2002) illustrates in his ethnography of private security workers in Britain, their work was based on their bodily capital and their awareness of the limits they faced while facing problematic customers during the nighttime economy.
Other concepts in the sociology of work help us understand the embodied performance of violence (McSorley, 2014). One is dirty work, known as labour, which, for some moral or perceptual reason, is undesirable for society (Simpson and Simpson, 2018). For example, Manrique Rueda and Tanner (2016) found that the Colombian government fostered the formation of paramilitary groups to control the presence of the then-FARC (Fuerzas Armadas Revolucionarias de Colombia) guerillas as dirty work that this government did not want to enact counterinsurgency with the Army. Alternatively, as Huggins et al. (2002) documented in Brazil, police forces perform violence against the gangs in Rio de Janeiro as dirty work to control their expansion beyond the favelas (slums).
On a different note, a relevant difference that brings the state recognition of violence work is the capacity to claim labour rights. In her seminal research, Skocpol (1995) argued that the popular demand to protect injured veterans and families in case of death led to the creation of one of the most critical social protection systems in the US. Nonetheless, soldiers and police officers operate under different legal work regimes that curtail their labour rights, in part because of the threat of violence they pose. Moreover, in recent years, the surge of police unions – even if not recognised legally as such – has brought tensions between policed communities, the states and police officers in the US in the wake of Black Lives Matter and other denunciations of police brutality (Litor, 2023; Marks and Fleming, 2006). A similar tension has surged for United Nations peacekeepers. As they have diplomatic protection, victims and NGOs have denounced them for using that protection to avoid scrutiny of their violent abuses (e.g. sexual violence). Military officers in peacekeeping use their labour hierarchies (e.g. not contravening superiors) to protect themselves from accusations (Podder and Manzillo, 2021).
Hierarchical and military control appears across violence specialists’ workplaces. Recently, Varsori (2021) documented how gangs in Rio de Janeiro began to adopt hierarchical discipline after acquiring heavy weaponry. These gangs needed to control the potential danger of a member defecting with such technical capacities. We have seen the same patterns in the protection workers recognised by states, polities and businesses: police officers, soldiers, private military officers and bouncers, among others. Governments have routinised, taught institutionally and regulated these occupations as work around the world.
As Malešević (2017: 281–285) summarises, the consensus on military studies is that organisations control violence specialists using ideology, peer-to-peer socialisation and vertical bureaucratic command. These techniques allow the social acceptance of violence as an inter-group bureaucratic endeavour, as suggested by Arendt (2006[1963]; for a debate on this, see the study by Brannigan and Perry, 2016). For example, the Nazi regime legitimised their discreet bureaucrats participating in the Holocaust as national patriots (Kershaw, 1993). However, the nature of their acts was hidden until the end, inherently dirty in a cruel performance of bureaucratic obedience (Collins, 1974). Even if it involves particular chauvinistic pride, hiding violence from the public eye is part of this craft.
Overall, researchers from these fields of study have discovered fascinating features regarding organisational hierarchies. For example, in criminology, it is clear that these organisations tend to be networks (franchises) rather than hierarchical structures under one payroll (Calderoni, 2012). Furthermore, we know that occupations and organisational features are essential to understanding wage inequality (Mouw and Kalleberg, 2010). Some licit public armed agencies publish data on wages, and some countries have employment survey data that could reveal wage inequalities between armed occupations and the rest. However, this is clearly a challenge for studying clandestine organisations that, for obvious reasons, do not reveal these data. We have anecdotal evidence though. The journalist Javier Valdez Cárdenas (2007) interviewed several children in Mexican drug cartels, and he found that drug cartels hired them as killers. Some ‘sicarios’ earned salaries per killing or on regular allotments, mostly earning above the average of a whole middle-class family income.
By using occupational lenses, we can clearly see the fascinating concurrences from different perspectives in the ethnographic or quantitative research about organised violence. Notably, these organisations, regardless of their legality in their historical context, recruit relatively young men. They are seeking social mobility, enacting different forms of militarised or violent masculinities, showing that the violence-protection labour markets operate similarly both in clandestinity and legality. For example, military officers move to private military companies because they have a suitable set of skills. As Kennard (2015) found, when the US Army needed to increase its numbers in 2003, officers relaxed the entry requirements to allow the recruitment of former gang members with essential skills for the war. Marginalised young men have engaged in mercenary and organised clandestine violence. In contrast, in premodern times, young male elites populated licit armed organisations. However, as the demand for massive violence increased in modernity, Armies and Police Forces also became recruiters of marginalised boys.
Regarding the performance of work, violence specialists need to gain skills in the use of weapons and particular bodily force. The organisations that recruit them either seek individuals with these skills (drug cartels, private military companies) or train them in these skills (police or military). Clearly, because these organisations are recruiting specialists in violence, controlling their behaviour is a crucial challenge. Most of them adopt highly hierarchical forms of control, socialisation, wages and benefits to maintain compliance with their members. Ideology is a way to maintain organisational cohesion (e.g. nationalism in armies and even genocides), and the formation of national armies was a way to organise mercenaries. Nevertheless, again, violence specialists use their skills as a commodity. They can abandon these organisations and later commit violence as independent contractors.
Implications and Conclusions
Once, Wright (2005) asked, ‘If class is the answer, what is the question?’ He answered that this question and others are generated ‘(. . . ) by curiosity and imagination, organised by theoretical assumptions and animated by normative concerns, encountering the empirical world’ (p. 180). This essay shows that the concurrences in empirical research on conflict, violence and crime are not coincidental. Two recent reviews of violence in conflict (Hillesund and Østby, 2023) and crime (Vilalta et al., 2023) show that there still is a contentious debate about whether inequalities are relevant in conflict. My argument is that we have missed the causal mechanism between violence and inequalities: labour markets. As Sørensen (1991) argues, stratified markets transmit inequalities. Therefore, instead of focusing on aggregate measures of income inequality, we must understand the labour market dynamics and social mobility prospects that constrain the choices of young people living in precarity or poverty, which in some cases leads to their recruitment into organised violence. If no new large-scale violent conflict emerges, increasing income inequality can hardly provoke conflict by itself. Some societies have proven resilient to high thresholds of inequality. Rather than viewing inequality as a precursor to violence, it is preferable to consider it as a pre-condition shaped by labour markets. In other words, inequalities create a reserve army of labour that could potentially be mobilised.
The explicit introduction of labour studies into organisational violence allows a two-way empirical agenda: on one side, we can use the classical questions proposed by Wright (2005:180) about class analysis for organised violence: inequalities, individuals’ perceptions within groups, life chances, standards of living, antagonistic class conflicts, historical variation in these organisations and questions regarding oppression and exploitation of labour. Also, we can conduct research on other participants of the criminal or clandestine world as workers. For example, chemists are producing illicit drugs, lawyers are representing criminals, accountants are laundering money, farmers are producing poppy flowers and drivers are trafficking through borders. Moreover, people are working for criminals or guerrillas whose labour is subordinate and not directly linked to the organisational objectives, such as houseworkers, cleaners and cooks, among others. Finally, other workers in legality provide services for clandestine organisations, such as teachers for the sons of gangsters or guerrilla members.
On the other hand, we can examine the classical question of why humans (usually men) fight wars and commit crimes by using the analytical and conceptual variety of labour studies. By adopting occupational labour lenses, these concurrences become clearer: researchers are seeing features of work in the everyday activities of violence specialists. A few years ago, Kalyvas (2015) warned against comparing civil wars and criminal violence on the basis of the context of where these events happened but suggested new pathways of study of organised violence based on comparing techniques or recruitment. As I argued in this essay, understanding those performing violence as an occupational class is the way forward by asking: why do men work in wars and criminal violence? Paradoxically, men from the ruling classes were involved in praetorian violence in earlier societies, which provided them with the prestige and wealth associated with waging war. In modern capitalist society, however, the massification of organised violence has made it a task for young working-class men. Nonetheless, being a rebel, hitman, soldier or private military contractor can provide impoverished men with new prestige, opportunities and meaning in life and acquire a new, breadwinner, militarised masculine status.
This occupation entails acting against other humans. Looking beyond rational choice or culturalist explanations, we can understand organised forms of violence as edgework. Therefore, depending on the historical and social context, violence can be a risky but desirable masculinised occupation. Like any other employment, the political interaction between states, societies and capitalists is the one that sets the demand for this labour power. In the end, the regularisation of the mercenaries was indeed an organisation of an occupational class. The end of wars (state, civil or criminal) implies the demobilisation and reintegration of violence specialists. Furthermore, no vacuum of power remains like that for long when criminal or political entrepreneurs decide to mobilise them within the margins of the state.
Furthermore, political scientists and criminologists are working to understand several challenging phenomena such as criminal governance, dynamics of violence, militarisation of public security or guerilla demobilisation processes. Let me give some examples from Latin America to highlight that occupational labour lenses will be helpful. For example, maintaining criminal governance is crucial to preserve certain daily work conditions of drug cartel sicarios in Central America. In part, Armies in Latin America are looking to protect their contractual advantages (protection against prosecution) when participating in counternarcotic operations. Moreover, even if demobilisation is political, some guerilla members facing unemployment could join a splinter organisation in Colombia.
This essay makes a case for sociological theory to strengthen the study of violence cross-disciplinary with the findings of the fields mentioned. The overemphasis on political science or economics explanations has over-simplified the violence-related social relations and the role of inequalities in those. Furthermore, this essay challenged several core presumptions of classical sociological theories that pose violence as antithetical to solidarity. Scholars should examine how the findings and suggestions I made in this essay can transform in dialogue the main theoretical perspectives in sociological thinking. Following Burawoy’s (2021) and Skinner’s (1969) historical approaches, we should understand in depth why violence ended up positioned as antithetical to collective action. For example, I think this occupational approach sorts the inner tension between revolutionary and state violence in Marxism by seeing them as work instrumentals to change or govern society. Alternatively, to clarify how the Weberian ‘monopoly of violence’ is formed by controlling the violence specialists, the mercenaries as once Machiavelli (2008[1532]) suggested, or if they become the state as Tilly (1985) famously said.
I hope that by understanding how work influences organised violence, we can develop a more nuanced comprehension of the role of inequalities in this context and, more broadly, address the empirical puzzle of why human-organised violence occurs. In part, as Collins (2008) once underlined, the overwhelming majority of humans do not commit violence. It is an activity highly specialised and, as I argue here, stratified. I hope we can see empirical studies with labour lenses soon, ranging from quantitative class count analysis to thick work ethnographies of organised violence with their perspectives. A research agenda of this nature implies going beyond aggregate measurements of income inequality and reviewing this relation in its complexity. For example, I suggest developing an occupational choice social mobility agenda (Hout, 2015) focused on the social origins of violence specialists as a comprehensive set of pre-conditions.
The prescriptive exoticisation or apologetic absolution of the participants in violence, crime and conflict overshadowed the analytical potentiality of class, work and labour studies in criminology, sociology and political science. Understandably, these activities are considered morally abhorrent by societies in different stages of their histories. Likewise, social sciences are historical byproducts of their time. They are not isolated from these historical mores. Some violence specialists are working in illegality, and the state enshrines others as their guardians. That depends on the legality of the employer and the regulation of this craft in the illicit world.
Footnotes
Acknowledgements
I benefitted from the enriching discussions I had with several colleagues, particularly Sukanya Podder, Mats Berdal, Kieran Mitton, Gareth A. Jones, Brian J. Phillips, Andrea Varsori, Laurie Cohen, Natalia Mendoza Rockwell, Luis Ángel Monroy Gómez Franco, Diego Castañeda, Diego Sánchez Ancochea, Lucia Pradella, John Narayan, Diego Macías and Andreas Moeller. I presented an earlier version at the Contemporary Marxism Research Group in King’s College London. The manuscript improved significantly thanks to the two anonymous reviewers’ kind and encouraging comments.
Funding
The author received no financial support for the research, authorship, and/or publication of this article.
