“WE KNOW WHAT WE DO”: AN AGENCY-STRUCTURE ANALYSIS OF POLITICAL VIGILANTISM IN GHANA

Political vigilantism has come out as an offshoot of the violence that characterises multi-party elections in Ghana. Several studies on political vigilantism give primacy to the role of structural factors in explaining the phenomenon in Ghana. However, such a notion deprives political vigilantes of any sense of agency and further evades how the understanding of the interaction between agency and structure can offer deeper insights into the phenomenon of political vigilantism. The paper argues that political vigilantes should be characterised as agents who can make decisions about the actions that they undertake and not solely determined by the constraints of structure. The actions of political vigilantes are mediated through social structures that shape the options available to them. Using purposive and snowball sampling techniques, 41 key informants were selected for the study in Accra, Kumasi and Tamale. The study employed both primary and secondary data sources. Whereas secondary data was sourced from desk reviews, primary data was collected through interviews and focus group discussions. The narratives were thematically analysed. The study findings show that members of political vigilante groups are conscious agents and for that matter decide and act either autonomously or relationally despite existing pressures from social structures. In addition, the paper shows that the exercise of agency among the vigilantes is inseparable from structure as the structure is reproduced and changed through the actions of the political vigilantes who in turn are shaped through the structures that they occupy. Therefore, policy responses to political violence should consider the two sides of the equation by working simultaneously on the behaviour of political actors and resolving the circumstances that facilitate political violence.


INTRODUCTION
Political vigilantism is an offshoot of the violence that characterises multi-party elections in Ghana.The term refers to the different forms of political violence that characterises the political landscape in Ghana since independence (Gyampo, 2010;Gyampo, Graham, & Asare, 2017).The Fourth Republic of Ghana, established in 1992, has experienced a surge in vigilantism and associated activities.Research indicates that political vigilantes are tolerated and encouraged by political parties in Ghana (Bjarnesen, 2021;Zabyelina, 2019).Available statistics indicates that there are over 27 vigilante groups in Ghana that are aligned to either the New Patriotic Party (NPP) or the National Democratic Congress (NDC).Political vigilante activities are generally violent and have characterized all the eight elections and three transitions in Ghana's fourth republic (Bukari, Ametefe, Osei-Kufuor & Imoro. 2023).
Unfortunately, many of these studies depict political vigilantism as an outcome of structural factors (Attuquayefio & Darkwa, 2017;Bob-Milliar, 2014;Gyampo et al., 2018;Paalo, 2017).This perspective considers behaviour as the outcome of the wider social and political environment in which people are located.In these discussions, the position of the individual as a rational being within the social system is not given much consideration as anyone is subject to the same structural constraints and inducements, and can thus be expected to behave very similarly.
The assumption is that the constraints of structure push people into becoming political vigilantes.Such explanations inadequately consider the different motivations shaping individual participation in political vigilantism in Ghana.In this way, vigilante group members are deprived of a sense of individual agency or the capacity to act independently of the structural constraints (Rapport & Overing, 2000).These limitations constitute an important research gap in studies on political vigilantism.
Consequently, we argue for the need to consider the different motivations of agency and the interactions between agency and structure for a more realistic understanding of political vigilantism.The paper stresses that social structures do not determine the actions of individuals but rather create a reality for them to act in particular ways to either maintain or reproduce these structures (Berger & Luckmann, 1966).Drawing on the ideas of Giddens (1984), Long (1992) and Cleaver (2012;2007), the paper refers to the ability of agents to respond to the challenges that society throws at them, to confront and deal with risk.The works of Giddens, Long, and Cleaver assist the paper by placing individuals with the power of agency engaged in a recursive relationship to structure.The paper uses the narratives of the members of political vigilante groups to illustrate how motivation and actions are critically related to social structure and the exercise of individual agency.In response to these limitations, the following research question were formulated for the study: What structural factors drive political vigilantism?What motivations drive the capacity for action among political vigilante groups?And finally, how do we understand the interaction between agency and structure in relation to the activities of political vigilante groups?
The next section reviews the literature on political vigilantism and further explores theoretical insights on individual agency.Section three presents the methodology for the study while section four focuses on the results and discussion where the factors that shape individual motivation to participate in vigilante groups are discussed and the effects of individual agency in political vigilante groups are espoused.The final part of the paper draws key conclusions emerging out of the research to inform policies on political vigilantism.

LITERATURE REVIEW
This section discusses the dialectic relationship between agency and structure and then follows it with political vigilantism.The review of political vigilantism covers the conceptualisation of the concept, the drivers, and the effects.

Structure and agency: the theoretical debate
The literature is replete with debates on agency and structure and how the two concepts interact to reproduce reality.In much of social theory, the debate centres on the proponents of individual agency who consider the individual actor's ability to write their biographies against those who argue that society determines the actions of individuals.The latter view is emphasised by the position that social actors operate in a field and because a social agent is located in a particular field that is characterised by its own unique principles of vision and division which are used to understand the world, then they are inclined to behave in a particular manner which can be explained through an analysis of structural factors.Bourdieu's (1972) view raises a general debate on how to interpret the behaviour of social actors as driven by circumstances, or as a choice.Response to this question is seen in three variants: those that tend to emphasise structure, those that generally prioritise agency, and those that assume a recursive relationship between agency and structure which are discussed in this section.
According to Giddens (1984), social structure refers to the patterns and relationships between and within groups.Other writers extend social structures to include institutions that are captured as rules of the game (North, 1990).Here the emphasis is on understanding the resources, social institutions and systems that are generalised at a societal level and manifest in recurring patterns of organisation and practices.The theoretical work of Durkheim (1895) underlines the relevance of social facts and rules which structure and organise human behaviour.Durkheim's (1895) account of the social structure suggests that individuals operate within the structures in which they find themselves.This implies that an individual's action may be understood and explained in the context of the social system or structure in which the person is embedded.As expressed by Barnes (2001), society is constituted by "external coercive powers and social structures" that make people act in particular ways.
Discussions of structure tend to lean much towards determinism (Brancati & Snydey, 2013;Fay, 1996;Fortes & Evans-Pritchard, 1987).People's attributes, behaviours and attitudes are seen as the consequence of the wider social and political environment in which they are located (Crothers, 1996;Hughes, Sharrock, & Martin, 2003).Consequently, the structure becomes the determinant for understanding society.For example, increasing unemployment leaves the youth with no other choice but to join political vigilante groups aligned with political parties.Thus, unemployed youth are subjected to the same structural constraints and incentives, and are all expected to behave similarly.This demands that, in order to understand political vigilantism, it is more important to understand the structural constraints that shape the behaviour of political vigilante group members.This position is rejected by the proponents of the agency who argue that individuals and groups are not only shaped by the consequences of the workings of the world but can also act in ways that lead to change (Klotz & Lynch, 2008) Agency places emphasis on human action as purposive and deliberate (Giddens, 1984;2009;Ma, 2010).This position considers human agency as the individual as a person with the capacity act, or possess the power to originate actions.Embodied in this definition of agency is the idea that actions, activities, decisions and behaviours constitute a meaningful choice (Deacon & Mann, 1999).Using the assumptions of the agency perspective, agents act when they are motivated, knowledgeable and capable of rationalising their actions.In addition, agents reflexively monitor their action.In this sense, explanations on human agency place emphasis on the rationalgoal-directed activities of individuals that associate meanings to a particular action.Such considerations of human agency assume that the individual is not a static entity whose actions are determined by social forces; rather, individuals are rational, dynamic and motivated actors.Here, the individual rationalises their choices and will only join a vigilante group when the benefits outweigh the risks.The decision is based on cost-benefit analysis, cognitive processes of alternative choices of maximising their benefits.This can be linked to clientelism, where the politician or political party makes certain incentives available that the vigilante group member hopes to benefits from and therefore decides to join the vigilante group.
From this perspective, the individual is rational with motives that are driven solely by the desire to make maximum benefits.The rational notion of agency has been critiqued for the inadequate consideration of relationality and interdependence (Barnes, 2001;Hemmings & Kabesh (2013).More so, rational explanations of agency tend to overemphasise decision-making and the choices people make rather than on their ability and the resources available for making such choices (Cleaver & Whaley, 2018).
The third strand, which is the position of this paper, adopts a hybrid position in which individuals are characterised as agents who can make decisions that inform their actions.Gidden's (1984) Structuration Theory, constitutes a reference point for most discussions on the interaction between agency and structure.According to the theory, the structural properties within social systems are considered both as a medium and outcome of the practices they recursively organise (Giddens 1984, p.).In this sense, agency and structure interact where individuals act under given social structures and circumstances to reproduce those same very structures through their actions.As stated by Maxwell and Aggleton (2014, p. 801), '... agency is expressed, made visible and, crucially, only has an effect through the particular social practices that it drives'.Therefore, in exercising agency, individuals actively act to solve problems that are considered solutions to the social influence upon their lives.The individual responses, in turn, reproduce, transform or reorient social structures (Archer, 2003).Giddens (1984) stresses that structure and agency discussions should consider the conditions under which an action takes place and the consequences of that action.Of particular interest, is the way social structures shape human actions and cause them to reproduce the very structures that affect them.For Giddens, social structure is produced and reproduced through everyday human activities, governed by rules and regulations, organisation practices and use of resources.According to Giddens (1984) agency connotes knowledgeability, intentionality, and capacity which are exercised within the confinements of the broader structural limits that shape agency.For this paper, authors explore the dialectic relationship between agency and structure as it provides meaning on how certain norms, practices, and behaviours over time become ingrained in society which eventually turn into structures that then influence the interests and actions of the actors involved in political vigilantism.The decision to join a vigilante group and participate in the activities of the group are borne out of deliberate cognitive processes which the individual or group is responsible for and this is in turn produce ways of perceiving the vigilante activities in politics in Ghana.As agents, political vigilantes are conscious of their actions which may be either practical or discursive.Drawing on these perspectives, this paper explains the motives of political vigilantes by putting them in a lifeworld, to know the different motives that shaped the decision to participate in this group and how their actions reproduce political vigilantism within the larger frame of political activism.

Political vigilantism
This section of the article looks at the issue of political vigilantism in Ghana by first conceptualising it, exploring the drivers, and finally the effects of political vigilantism.

CONCEPTUALISATION OF POLITICAL VIGILANTISM
The word vigilante is of Spanish origin and means watchman or guard, but its Latin root denotes vigil, meaning, awake or observant (Rosenbaum & Sederberg, 1974 p 542).Scholarly definitions of vigilantism refer to it as a private activity with no direct state involvement (Martin, 2010;Abrahams, 1998;Johnston, 1996).Vigilantism is also defined by Abrahams (1998) as taking 'the law' into one's hands.Similarly, Johnston (1996, p 220) considers vigilantism as either the informal regulation of acts of criminal deviance or the informal regulation of society's moral standards by non-state actors.From this conceptualisation, Johnston (1996, p 232) defines vigilantism as: a social movement giving rise to premeditated acts of force by autonomous citizen.It arises as a reaction to the transgression of institutionalised norms by individuals or groups … Such acts focused upon crime control and aim to offer assurances of security both to participants and to other members of a given established order.For Johnston (1998), the initial thought of taking charge of the law means challenging the state's capacity to implement law and order which may be considered illegal.However, Johnston considers this conception of vigilantism to be problematic regarding that vigilante activity can occur under self-protection during conditions of disorder and state failure.In line with this, Kyed (2018) uses the concept of 'street authority' in Mozambique and Swaziland to show how in the absence of effective state capacity, vigilantism offers citizens protection and order through swift, direct actions using violent means.In instances where the state's capacity to respond to crime is weak, vigilante groups emerge to provide all sorts of security and regulatory functions (Singh, 2008).
While a precise definition of what vigilantism exactly may be missing in the literature, generally it is understood as an organised attempt by a group of ordinary citizens to enforce norms and maintain law and order on behalf of their communities, often by resorting to violence, in the lack of trust in state institutions and the perceived absence of effective official state action through the police and courts.The inexplicable nature of vigilantism as a concept requires focusing on the key features that are common in the definitions and how these features explain the phenomenon.More so, it is essential to further understand why vigilantism emerges.Most of the definitions of vigilantism capture elements of violence, group mobilisation, protection, crime, security vacuum, and poor criminal justice system (Abrahams, 1987;Baker, 2008;Buur, 2006;Kantor & Persson, 2010).
Common to all these definitions is the use of violence to offer protection due to the perception of increased criminality or social deviance which threatens social order.With vigilantism, protection becomes a matter of survival and self-responsibility (Rosenbaum & Sederberg, 1974;Sircar & Jain, 2017) and such groups are assumed to flourish not only in places where states lack the capacity to protect citizens from crime but also where the state itself is believed to be impartial in the provision of security (Boege et al. 2008).In all these, violence is predominant in the narratives around vigilantism (Pratten, 2008;Smith, 2004).Therefore, vigilantism emanates in response to the state's perceived inability to respond to problems related to insecurity.This notion of vigilantism has been applied to explain the phenomenon of political or electoral violence in Ghana since the inception of constitutional rule in 1992 (Bob-Milliar, 2014;Bjarnesen, 2021;Gyampo et al. 2017;Kyei, 2020;Tankebe, 2019).The term emanated from certain media houses that sensationally used the word political vigilantism without sufficient clarity to mean political violence (Gyampo et al., 2017).With this, political vigilantism became the socially accepted phrase for any form of political violence in Ghana.Vigilantism in Ghana is mainly about violence associated with the electoral process.According to the Vigilante and Related Offences Act, 2019 (Act 999), political vigilantism is "the act of enforcing law and order without authority".The Act considers actions beyond political violence to include behaviours that are unruly, violent, disruptive, unlawful, and in some cases, criminal.
With these considerations in mind, this paper conceives political vigilantism as the activities of non-state actors, such as a person or an organisation, that collectively utilises the structures of political parties for organised mob violence to pursue the interest of their political parties with the intention of influencing the electoral process or shaping the outcome of an election.Inherent in this definition is the presence of a group that is associated with a political party, uses mobilisation and collective action to perpetuate violence, and performs protection or security roles.

Overview of political vigilantism in Ghana: drivers and its effects
Even though political vigilantism is considered a recent phenomenon, the political history of Ghana suggests it gained prominence in the period just before independence.The political history of Ghana is dotted with many incidences of violent attacks committed by groups aligned political parties.For example, the Convention's People Party (CPP) and the National Liberation Council (NLC) used the Young Pioneer Groups and the Action Troopers respectively to perpetrate violent attacks on their political opponents (Aryee, 2019; Gyampo et al., 2017).Another incidence of political violence was recorded under the Provisional National Defence Council (PNDC) in the early period of the revolution with the mobilisation of people to form groups such as the "Mobisquad", Committee for Defense of the Revolution (CDR) and the Workers' Defence Committee (Gyampo, 2013).
The shift to democratic rule in 1992 witnessed the re-emergence of electoral violence in Ghana's Fourth Republic with political parties relying on youth groups for campaigning and mobilisation (Bjarnesen, 2021;Bob-Millar, 2014).These activisms of political youth groups often resulted in violence at the behest of the political parties due to the intense competition for political power.It is widely known that the main actors of vigilantism in Ghana are the two main political parties, the New Patriotic Party (NPP) and the National Democratic Congress (NDC) (Armah-Attoh, 2017; Attuquayefio & Darkwa, 2017;Dumenu & Adzraku, 2020).The NDC and NPP are associated with and patronise political vigilante groups mainly due to the zero-sum nature of the political system in Ghana.As a result, the two main parties, the NDC and NPP patronise political vigilante groups for support with election-related activities.
Political vigilantes are known for their canvassing potential and their willingness to use violence to improve the chances of their political party winning an election.Consequently, both political parties have formed vigilante groups and consolidated them within the party's organisation with the aim of using them to win political power.According to Aning and Albrecht (2020), the political vigilante groups take on names that are associated with power and fear such as Pentagon, Al-Qaeda, Al-Jazeera, and NATO Forces.Others groups are named after prominent political party big shots who are who are capable of financing the group.Example of such groups are the Azorka Boys and Maxwell Boys that are named after NDC Chairman in Tamale and an NPP founder in Kumasi respectively (Bukari, Ametefe, Osei-Kufuor & Imoro, forthcoming).
According to Gyampo et. al. (2017) and Bob-Milliar (2014), political vigilantes are dominated by party youths who are mostly found at the bottom of the party hierarchy.They sacrifice their time and energy to canvass support for their political parties (Moyo, 2020).Their activities are seen as contentious, violent, aggressive, and informal with their brand of political activism characterised by lawlessness and hooliganism (Bob-Milliar, 2014;Jackson, 2012;Kwarkye, 2018).
Many activities of the political vigilantes are focused on elections and other major political party-related activities.All the major elections since 1992 have been associated with violent acts of political vigilantism (CDD, 2017;CODEO, 2017).Activities of political vigilantes often include the snatching of ballot boxes, intimidating of political opponents, disruption of the voting process, and disturbances during the counting of ballots (Birikorang & Aning, 2016;Gyampo et al. 2017).The issue of violence is even more intense during by-elections.A clear case is the Ayawaso West Wugon by-election in 2019 which attracted national attention due to the enormity of violence associated with that specific by-election (Bjarnesen, 2021;Graham & Hagan, 2021) Beyond inter-party clashes, political vigilante groups are being used internally within political parties (Aning & Albrecht, 2020).Bjarnesen (2021) highlights the internal nature of the violence of vigilante groups, often related to claims of neglect of foot soldiers by their political patrons.For example, when the NPP won political power in 2016, some toll booths being managed by members of the opposition NDC were seized by agents of the NPP.Similar incidents of seizure of public facilities were reported at various places in Ghana.The members of the Delta Force, a political vigilante group affiliated with the NPP government, attacked the Ashanti Regional Security Coordinator in Kumasi and brutally assaulted him and some of his staff members (Graphic Online, 24 March 2017).The same group also attacked a Kumasi Circuit Court and freed about 13 members of their vigilante group who were facing court charges for causing disturbances at the Regional Coordinating Council.All these violent acts by members of the NPP, that had won the 2016 election, were related to the benefits to be derived from the political spoils.Political party foot-soldiers turned vigilante groups, notably the Delta and Invincible Forces, resorted to the use of force and in other instances violence, to seize and manage control state assets and public facilities (CDD 2017;Gyampo et al. 2017;Owusu-Kyei & Berckmoes, 2020) The above review drives home the argument that political vigilante groups have been a nuisance to Ghana's democracy with the two main political parties involved.In any democratic regime, elections are supposed to be competitive, free, and fair.Unfortunately, political violence dominates the electoral process affecting the conduct of elections in Ghana.Understanding the structural conditions driving this violence is central to studies on political vigilantism but using an agency-structure analysis will offer a better understanding of the phenomenon in practice.

METHODOLOGY
The study is qualitative, using both primary and secondary data sources.The primary data was collected from members of political vigilante groups in Accra, Kumasi and Tamale.These study sites were purposively selected due to the presence of vigilante groups as captured in the Conflict Map of Ghana and CODEO Reports between 2000and 2017(CODEO, 2017;Kendie, Osei-Kufuor & Boakye, 2014).Using snowball sampling approach, the research team used interlocutors to get access to a key member of the vigilante groups in each of the three sites.These individuals then supported the research team to gain access to other members of the vigilante groups.
As indicated in Table 1, a total of 22 political vigilantes from Greater Accra, Ashanti and Northern Regions were sampled for the study.The regions were selected because they had recorded incidence of violence associated with political vigilante groups.Purposive sampling procedure was further used to select three constituency representatives of both the NDC and NPP from each of the selected study regions.In each constituency, the constituency organiser was purposively selected because they are responsible to all party organisations at that level.The activities of the vigilante groups are at the constituency and polling station levels, hence the decision to interview political actors in that space.In addition, the Executive Secretary of the National Peace Council (NPC) in each of the study regions was interviewed because they had implemented sensitisation and education programmes for political vigilantes.The Regional Police Crime officers were purposively selected for the study.In each of the selected three regions, the key contacts arranged for a group discussion with the members of the vigilante group.The participants for the group discussions ranged between six and nine and were predominantly males with ages ranging between 21 and 44 years.
Key informant interviews and focus group discussions were the main methods of data collection used.The key informant interviews were held with selected members of the political vigilante groups, constituency representatives of both NDC and NPP and the Executive Secretary of the NPC in each of the study region.Three focus group discussions were conducted with youth groups affiliated with a political party.During the period of the study, the Vigilantism and Related Offences Act, ACT 999, 2019, had been promulgated which made it difficult for the research team to get access to the Political Vigilantism groups.Existing political Vigilante groups were banned with the coming into effect of ACT 999.It must be emphasised that each informant belonged to a political party.However, for the sake of anonymity, the paper did not mention the political party, or the name of the vigilante group.An informed consent was obtained from all the study participants before being interviewed.Information on the informed consent form was read to the FGD participants and the key informants with those consenting to participate in the study made to sign or thumbprint the form.
The data collected were transcribed and further analysed using thematic analysis.In addition to this, the transcribed data was summarised and produced in the form of Meeting Notes which were numbered in the order that they were filed, followed by a date.The discussion quotes the Meeting Notes rather than attribute quotes to individual informants in order to guarantee their anonymity and protect sources.The secondary data sources for the study were collected from relevant materials, including publications in local newspapers, online sources, journal articles, and reports of election observer missions.

Structural Drivers of Political Vigilantism
In this paper, social structure is applied to explain the drivers of political vigilantism.The study reveals a connection between socio-economic conditions such as unemployment and political vigilantism.During the interviews, high youth unemployment and illiteracy came out strongly as contributing factors to political vigilantism.According to a key informant, politicians take advantage of idle youth who have no other option than to follow the politicians to get some money (Interview A5, 8 May 2021).Another informant revealed that the support political vigilantes received from the political elites is through the provision of 'food and money every day because we are loyal followers who work for them' (Interview B6, 24 June 2021).Likewise, some informants suggested following political elites is lucrative because people make a lot of money through that.
The interviews reveal that unemployment and illiteracy affect youth vulnerability which results in the political vigilantes associating with political elites (Aning, 2020).Thus, political vigilante activities constitute a source of employment.Sharing a similar position, a key informant was of the view that: political elites offered the members of the vigilante groups money in exchange for their services throughout the electioneering period.The situation is further compounded by the fact that the majority of the young people are inactive and are easy to be exploited as many are out of school and with limited opportunities for socialisation.For the party youth, going to the residence of either the party executives or the aspiring member of parliament every morning to take instructions, is a sure way of getting money and food (Interview A2, 6 May 2021).The narrative above suggests that political vigilantism provides livelihood options for many of their members.This study therefore supports the view that political vigilantism is associated with youth unemployment (Asamoah, 2019;Attuquayefio & Darkwa 2017;Bjarnesen, 2021;Dumenu & Adzraku, 2020;Gyampo et al. 2017).
The use of political vigilante groups by the political elites for election-related activities came up during the interviews.As indicated by participants in an FGD, the opportunities associated with winning elections have driven many youths to join the vigilante groups of the two main political parties.The participants noted that the politicians prefer the youth because they can quickly mobilise for violence (FGD 03, 7 June 2021).Another informant mentioned that members of the political vigilante groups are mobilised to do the party bidding due to the false feeling of threat (misconceived) that if they do not do something their party will lose power or will not win an election.This can be attributed to the absence of values in our political parties (Interview B7, 26 June 2021).Thus, these narratives point to unemployment and livelihood concerns as a key structural driver of political vigilantism in Ghana.
The study further notes that the use of political vigilantes to offer security during elections was as a result of the perception that the security agencies are unfair and biased.This was confirmed in all the FGDs where the participants agreed that the provision of security by the political vigilantes is due to the perception among the members of the opposition political parties that the security agencies are aligned towards the ruling party (FGD 01, 15 May 2021).The lack of trust in the state's ability to provide security and justice to all political players, especially to opposition political parties, explains the rise in political vigilantism over the years.Therefore, the desire of parties to win power by all means gives impetus to the creation of vigilante groups to intimidate their opponents (Interview A4, 9 May 2021).Many of the political vigilantes interviewed mentioned the lack of trust between the two main political parties as a key driver of political vigilantism.A key informant indicated that: the two political parties, especially those in the opposition think that when something happens during the elections and they report it to the police they won't get justice.They consider the police not to do enough to support their cause.The alternative is to use the vigilante groups to provide security for them even though this is a function of the police and the military.In the provision of such security services, the boys clash with the police and this leads to violence (Interview C4, 28 May 2021).
The interview above shows that the mistrust of the security agencies makes the political parties to use their own security operatives to deal with election-related security issues.Several of such informal security roles are offered by the political vigilante groups who with time become armed to provide safe and secure environments for the political elites during political campaigns.In addition, political vigilantes are used by the political elites for personal protection services.As a result, senior members of both NPP and NDC are on record to have admitted to the use of political vigilantes to offer informal security services during political campaigning and elections (Mubarik, 2018).This development has resulted in violent clashes between the two political parties and also between the vigilante groups associated with these parties and state security as witnessed in the Ayawaso West Wugon by-elections in 2019 (Myjoyonline, 2019).These findings are consistent with the studies by Bjarnesen (2021), Gyampo et. al. (2017) and Mawusi & Adzraku (2020).
Besides the informal security services offered by vigilante groups, political patronage and the culture of rent-seeking were mentioned as key drivers of political vigilantism.As one informant put it, rent-seeking has rendered politics to become a business and these politicians need power to stay in business.Therefore, the politicians will go to every length to win elections since the consequences of losing an election are dire and dangerous (Interview A1, 4 May 2021).This interview shows that the benefits that the political elites derive from politics will make them adopt all manner of strategies to stay in power.Individuals and groups invest heavily in political parties with the aim of reaping bountiful dividends through state contracts, appointment into offices and boards of state organizations when their parties win power.
Closely linked to this assertion is the winner-takes-all politics that allows the winning party to appoint its members to occupy positions in government.According to Gyampo (2015), winnertakes-all politics leads to state capture that results in the partisan monopolisation of state resources, facilities and opportunities as well as the exclusion of political opponents from national governance.Bob-Milliar (2014) links the winner-takes-all party politics to low-intensity violence that characterises the activities of the political vigilante groups.The exclusion of all other Ghanaians who are not members of the victorious political party from benefiting from the political spoils leads to the proliferation of political vigilante groups whose mission is to work for their political party to either stay in power or win power.This confirms both Bjarnesen (2021) and Gyimah-Boadi's (2007) argument that political vigilantism is a creation of political elites, patronage and insecurity within their Ghanaian political space.( In other responses, the complicity of the Electoral Commission is mentioned as a key driver of political vigilantism.According to an informant, the EC is considered to be biased since its decisions tend to push the interest of the ruling party (Interview A2, 6 May 2021).The activities of the Electoral Commission were perceived by the FGD participants to be an impetus for political violence.In addition, the challenges with elections are a key structural driver for political vigilantism.

Unpacking agency in political vigilantism
In this section, the paper argues that political vigilantes are initiators of action whose motives are driven by pecuniary incentives.The notion of intentionality is used to first point out that individuals have capacities to act in purposeful and meaningful ways and secondly, to show the diverse motives underlying the decision to join political vigilante groups.Two recurring themes were dominant in the motives of the political vigilantes.The first was that they joined the group as a conduit to get formal employment when the political party wins' power and secondly as a source of living.The following narrative captures these motivations: Most of our boys are people without jobs so we are perpetually prepared to join political vigilante groups and discharge these roles for the politicians with the motivation that we can get money.At my age, I rely on my parents for food.I joined the group because in the end, I look forward to getting a job, especially in the national security, military, police or in immigration service.The economic system here in Ghana is unthinkably hard, we face all the sufferings as a people, without joining such a group it is not easy to survive the untold hardships, sometimes we are deeply troubled since our parents even see us as failures in life as a result of unemployment, we, therefore, have no option than to join a group like this with the aim of getting money to feed ourselves and our family (Interview A5, 8 May 2021).
The prospect of jobs drives the zeal to partake in political vigilantism.For that reason, joining a political vigilante group is for economic motives which is self-interest driven but beyond this are other motives.According to Dumenu and Adzraku, (2020), the promise of job opportunities by politicians and personal aspirations for success constitutes a powerful source of agency for political vigilantes.As posited by Daddieh and Bob-Milliar (2012), the activities of political vigilante groups are based on the notion of reciprocity and the provision of personalised goods, such that people join the political vigilante groups as a result of some motivation factors.They further explain that individuals are pushed by lots of motives to join political vigilante groups.
Evidence from the study reflects the diverse motivations and varying strategies underlying the decision to become a member of the group.Rational motives dominate the desires of these political vigilantes.Such motivations were associated with the benefits to be derived from being a member of a vigilante group.As illustrated by a statement from one of the informants, he had his driving license when his political party came to power (Interview C3, 22 May 2021) while another indicated that he got a passport and the opportunity to travel abroad (Interview B1, 17 June 2021).Another informant stated that he got employment as some IT personnel at the School Placement Centre at the Ghana Education Service in Accra (Interview C5, 29 May 2021).Accordingly, political vigilantes reflect upon and evaluate their present situation to make certain choices.One more aspect worth mentioning is the way the political vigilantes integrate aspirations and optimism in realising their life goals.Such forms of human agency are considered meaningful and strategic (Cleaver, 2001;Giddens, 1984;Long, 1992).
Expanding the discussion brings to the fore the need to critically consider the variations that result from the exercise of agency.A critical issue is the unequal ability of the vigilantes to meet their desired goals.Some made it very clear that though they have been members of these groups their desired goals have not been met.Control and power came out clearly as shaping the outcome of agency for the political vigilantes.Some members of the groups had power which resulted in a greater scope of action and choice than others.Some low-positioned members had the desire to use the political vigilante groups as a vehicle for getting a job but they seem not to have fulfilled those goals.One member indicated that he nearly resigned from the group because he was not benefitting from the political spoils (Interview B3, 10 June 2021).This implies that political vigilantes are differentially positioned within the social system which affects their ability to act in ways that lead to desired outcomes or meet their life goals (Cleaver, 2012) Even as they waited for their turn to benefit from political power, vigilantes within the governing NPP acted through violent means to drum home their disenchantment with the political party.Aggrieved members of the Delta Force, a political vigilante group associated with the NPP, on Sunday, 7 th October 2018 attacked and chased out a government minister for "failing to fulfil the promise made to them during the 2016 elections (Ghanaweb, 10 October 2018).Political vigilantes possess agency in the sense that they use violence to get the party to respond to their concerns and be the agents of their own development.In this example, the political vigilantes acted because they felt the government minister did not share rents which incentivised them to mobilise in response to party activities.
The narratives of the political vigilantes further highlighted the diverse motivations that shaped their actions.For example, an informant indicated that gaining recognition within the party was another main reason for joining a political vigilante group.The idea was that joining a vigilante group would be instrumental in making him known within the party and further enhance the prospects of meeting his expectation to travel outside the country.For another political vigilante, the desire to gain recognition in the group and within the constituency, made him to use his own money to fuel his car for party functions.The informant explained that this was through his own effort and party people can attest to it (Interview B1, 17 June 2021).Such narrative reflects Dumenu and Adzraku (2020) view that factors such as "recognition within the party; protection from the law; and for the purposes of revenge and protection" motivate people to join these political vigilante groups.Furthermore, this reinforces the idea that incentives that shape agency are not only rationally motivated but are mixed and embedded in socially dominated principles.Though the dominant narrative for joining a political vigilante group was economic, the motives varied with some being fluid and socially-embedded while others were taken in the interest of the group (Cleaver, 2017;Dumenu & Adzraku, 2020).
Another dominant feature in the narratives of the political vigilantes was the use of social relations to enter the group.Among the political vigilantes interviewed, the aspiration to become a member of the group was channelled through friends who were influential either within a particular political party or the vigilante groups.This shows that the ability to have access or to be successful in joining a vigilante group is facilitated through existing networks and not necessarily through agency capacity.Even within the vigilante group, social relations are critical in upward progression or being influential in the ranks of the vigilantes.Energy, social capital and knowledge of party mobilisation were mentioned as key ingredients for progression in the group.This shows that people who possess these qualities can create their own trajectories within the political vigilante group and by extension, in the political party as well.Thus, future projects and aspirations are shaped not only by the individual's capacity but also by their social relations.
Another dimension of agency noticeable in the narratives of the political vigilantes is their individual characteristics.The objective dimension of the individual agency considers individual skills and resources that can be mobilised for the achievement of personal goals as critical in agency achievement.Since political vigilantes indulge themselves in activities that are often physical in nature, they require people who are well-built and muscular with the physique to perform these acts.Dumenu and Adzraku (2020) in their study of political vigilantism in Ghana showed how "party youth are organised to use their physical strength to help and protect the party".The same study described political vigilantes as well-built and muscular men who provide security and protection for politicians.This reinforces the idea of an interactive association between embodiment and agency (Giddens, 1984;Long, 1992) that often manifests through participation in group activities.
Many people interviewed noted that political vigilantes organise and execute their actions through violence.The use of violence to pursue their agenda suggests some form of agency, though negative (Bob-Milliar, 2014;Dumenu & Adzraku, 2020).According to a vigilante interviewed, violence does not occur in a vacuum but it is always in response to the happenings of either the state security apparatus or based on the strategies of their political opponents (Interview B1, 17 June 2021).The interviews show that political vigilantes in using violence often act within a certain context to exert power.As indicated by an informant, we assess the situation before violence is used but most often violence is the main weapon.When we notice that the security agencies will use force which will disadvantage us, we use violent behaviour to cause confusion and then move (Interview B1, 17 June 2021).Extant studies on political vigilantism in Ghana document the numerous ways in which political vigilante groups used unlawful forms of violent strategies during pre-electoral periods, during the elections, and in post-electoral periods across the country.The strategies employed by these political vigilante groups include burning party offices, cars, paraphernalia, looting properties, seizure ballot boxes, forcefully taking over the management of public facilities, disrupting party rallies, physically attacking appointees or removing officials from offices, gunshots, and in some instances inflicting wounds on political opponents (Armah-Attoh, 2017; Asekere, 2020;Dumenu & Adzraku, 2020;Paalo, 2017).However, there have been instances where violence shapes intra-party activities.Political vigilantes deliberately use violence to change happenings within their political party.The dilemma is whether the use of violence is an active exercise of agency or a crippling structural constraint?The choice to use violence is a conscious act that is aimed at making their displeasure about some decisions known to party leaders and therefore constitutes an active exercise of agency.These use of acts of violence are ways that political vigilantes endeavour to ensure the contextual continuity of political vigilantism.However, the use of violence involves greater reflexive engagement.The findings suggest that using violence is a common strategy to ensure the relevance of political vigilantism in the political arena.The political vigilantes consciously act in ways that reproduce the narrative that political vigilantism is violent and through that draw attention to themselves.These acts require reflexive management

Structure-agency analysis of political vigilantism
In this section, the paper combines aspects of the structure-agency perspective to explain political vigilantism and how the actions of political vigilantes reproduce the phenomenon over time.The paper argues that individuals are capable of exercising agency and at the same time shaped by social structure (Giddens, 1984;Long, 1992).According to Giddens (2009), social structures predispose individuals towards certain particular acts which in this case are considered as the actions associated with political vigilantism.Political vigilantes draw on existing social structures to act in ways that produce and reproduce political vigilantism in Ghana's fourth republic.This duality is illustrated with examples in this section to show that political vigilantism is produced and reproduced by agents through the structures that have been discussed in the earlier sections of the paper.
It is common knowledge that violence is a characteristic of political vigilante groups in Ghana (Armah-Attoh, 2017;Asamoah, 2020).The decision to use violence is informed by certain contextual factors such as the actions of the security agencies, the posturing of the Electoral Commission and the tactics of the other political parties.For instance, the mistrust of the security agencies makes political vigilantes act in ways that seek to curtail the power of the security agencies.According to an informant: When we were deployed to support the Party during the by-election, the police arrested two of our members and placed them in police cells.Our leader through the parliamentary candidate requested bail and the release of the two colleagues but the police refused.There were heated exchanges between the police and our leaders.We started pelting the police with stones which eventually led to the firing of warning shots by the police.Some of the voters at the polling station who were in the queue ran to safety (Interview B3, 10 June 2021).Such stories showcasing the use of violence are not isolated but run through a lot of interviews.The reproduction of the system of violence associated with the actions of the political vigilantes occurs through the structures creating political vigilantism.Weak enforcement of electoral laws by the EC and the failure to prosecute criminal offenses related to elections by the security agencies shape the behaviour of political vigilantes in ways in which they act to reproduce the violence associated with political vigilantism.The reproduction of the violent related actions that are associated with political vigilantism reflects Giddens's (1984, p. 19) argument that "structure is both medium and outcome of reproduction of practices that recursively organise".In this way, agency and structure become an iterative process where political vigilantes act under given structural conditions.Undoubtedly, the actions of the political vigilantes led to the social reproduction of familiar systems that are associated with political vigilantism, and then this structure manifests.It further shows that political vigilantism and its associated systems are recreated through the actions of the vigilantes.
The duality of agency and structure as used by Giddens (1984) offers pathways to understanding the recurring political vigilantism in the political history of the fourth republic.The winner-takes-all mentality, the grabbing of political spoils after elections by political parties that won the elections.Each of these drivers provides the medium through which the political vigilantes exhibit conscious goal-driven activity to reproduce the system.According to Giddens (2009), structure evolves with social interactions to create change.Political vigilantes as shown earlier are "purposive agents" who are aware of the intended outcomes of their actions.This awareness becomes a reflector that monitors their everyday routines thus creating a context for mutual relationship between agency and structure, ordered across space and time.This mutuality further creates a setting for reproducing and maintaining political vigilantism.

CONCLUSION
The aim of this article is to use the interaction of agency and structure to offer a better understanding of political vigilantism in Ghana.The study demonstrates that political vigilantes have their own way of shaping their life course that is not reducible to social context.In particular, the paper has shown that agency (personal motivations) seems ever-present in the actions of political vigilantes.Through their actions, these vigilantes express their capacity to assess issues and respond consciously beyond the restrictive orders, norms and structures of the political environment.Thus, political vigilantes are actors who do not only make choices but shape their circumstances.Intentionality and capacity to display goal-driven behaviours shape the agency of political vigilantes.The paper shows that political vigilantes are not determined by the structural constraints that drive political vigilantism but they rather respond to the environment in which they are located.Through actions that are purposive and relational, political vigilantes see an opportunity and therefore draw on this structure as the medium to reproduce political vigilantism.Therefore, the paper is of the view that when the structural constraints that drive the phenomenon are tackled, the incentive for becoming a political vigilante to achieve life goals will diminish.
The outcome of political vigilantism is experienced in the way through which the drivers of political vigilantism offer the opportunity for violent actions that affect the electoral process at different stages.Therefore, structure and agency are intertwined and as such, all policy responses to political vigilantism should place emphasis on this duality and work on both sides of the equation.Thus, policymakers need to pay attention to the drivers of political vigilantism and introduce measures that can resolve them whilst at the same time working to direct the behaviour of party youth to the use of non-violent means of getting grievances resolved.Political parties should introduce reward systems that do not privilege the use of violence.Again, there is the need to re-examine the accrued benefits in terms of the associated power and dominance when a party wins political power.That the centralisation of power in the hands of a sole political party which is sometimes termed as winner takes all further makes electoral activities extremely antagonistic.Further discourse ought to project the path of coalition governments to ensure diffusion and division of power as well as increased consensus building.