What democratic policing is … and is not

ABSTRACT

Democratic policing is a multidimensional, multilevel, and contested concept rooted in political ideology. It is not singular or politically neutral. I argue there are four typologies of democratic policing: right, centre-right, centre-left, and left. In Latin America, in the 1980s and 1990s, countries went through the dual processes of democratisation and the implementation of neoliberal economic policies. The latter increased inequality in wealth and led to deeply divisive debates regarding the place of equality and violence in the definition of democracy. Putting aside these debates on the meaning of democracy, police reform projects in Latin America have embraced community-oriented policing as synonymous with democratic policing. Yet, democratic policing is not a singular concept and political debates matter to its various meanings. The article uses Goertz’s (2006. Social science concepts: a user’s guide. Princeton University Press) three-level concept analysis to assess the theoretical similarities and differences between the four types of democratic policing. It then tests the theory with empirical data from the cases studies of Argentina (Menem and Kirchners) and Chile (Bachelet and Piñera). The case studies are informed by field research in both countries (2006–2015), and draw on media and human rights reports as well as secondary data. The study finds a gap between theory and practice that calls for more research on policy convergence. More importantly, it reveals the need to situate ideal definitions of democratic policing within political debates on democracy, paying close attention to the role of political ideology.

KEYWORDS:

• Democratic policing
• community policing
• political ideology
• Latin America

Latin American police forces have a recent history of involvement in authoritarian regimes and civil wars. With the resumption of electoral democracy, they assumed a central role in crime control, a pressing policy issue throughout the region. This context of democratic transition has elicited a great deal of scholarly and practitioner interest in how to democratise the police. Yet, as Sartori reminds us, it is important to have clear concepts before we move to operationalise them. To this end, this article asks: what is democratic policing? The answer, I argue, is not singular. Political ideology matters.

Indeed, a study of Latin America highlights the importance of political ideology to the definition of democratic policing. In the 1980s and 1990s most countries in Latin America experienced the dual processes of democratisation and neoliberalisation. Democratisation was accepted as referring to the return of electoral democracy but anything beyond that definition was the result of compromises within the traditional elite (military, political leaders, and the industrial and agrarian elite) (Munck 2015, p. 368, also see Nun 2003). Neoliberal economic reforms were supported by important factions of the elite but often imposed on countries by the IMF through structural adjustment loans offered in response to the region's debt crisis. The economic reforms included significant cuts to regulations, taxation, and social services. Neoliberal economic reforms, as now recognised even by the IMF, widen inequality of wealth (Ostry et al. 2016). This consequence is the focus of political debates regarding the definition of democracy throughout the region. The key question is whether this inequality is acceptable and natural (a mere outcome of democracy) or unacceptable and structural (problematic for democracy). In many cases political leaders’ support for democracy – even electoral democracy – has come to depend on the definition of democracy (Munck 2015, p. 371).

Putting aside these debates on the meaning of democracy, police reform projects in Latin America have embraced Community-Oriented Policing (COP) as the international best practice (Bayley 1994, Sklansky 2008, p. 82, Müller 2010, p. 22). Indeed, much of the literature on COP in Latin America juxtaposes it to mano dura (iron fist) policing, offering COP as the democratic alternative (Arias and Unger 2009, Lum 2009, Müller 2010, Dias Felix and Hilgers, this issue). Yet there is no one definition of COP. Very broadly it can be understood as police working with communities, in partnership, rather than against them. What this means in practice varies between COP programmes (Dias Felix, this issue). Certainly some academics and practitioners advocate that COP must contain particular features (e.g. Bayley 1994, Arias and Ungar 2009). However, no programme contains all the features on any such checklist and some definitions of COP are so vague as to make it indistinguishable from traditional policing practices (Manning 2010, Dias Felix and Hilgers, this issue). Thus it is not clear that all COP programmes are equally democratic.

To verify, the literature on democratic police reform appears to agree that democratic policing can be understood as when elected political leaders are able to effectively use police to uphold the rule of law (implied to refer to both crime control and protest policing) and that the police, as public servants, respond to citizen complaints, are accountable, use a minimal level of coercion, and respect human rights and notions of justice and equality (Jones et al. 1996, Linz and Stepan 1996, Holston and Caldeira 1998, pp. 282–283, Bayley 2006, p. 19, Manning 2010). Jones, Newburn and Smith argue that of these characteristics of democratic policing, equity is the most important (1996, p.190). This consensus is very useful for practitioners as it then allows for the establishment of checklists of specific technical police practices that need to be in place in COP or other police reform projects. Both the definition, and associated checklists, generally correspond with various ideal forms of liberal democracy and are consistent with the assumed experience in Anglo-American countries as well as early definitions of democracy found in the literature on democratisation in Latin America, such as the Schmitter and Karl (1991) article from which I have taken and revised the title for this article.

For example, in his well-cited book on promoting democratic policing abroad, Bayley defines democracy as ‘a government that is constitutional in the sense of being based on law, with authority exercised on behalf of representatives elected at frequent intervals by universal suffrage through processes that are free and fair’ (2006, p. 18). He explains that he has chosen this definition because ‘countries most interested in assisting in its development’ (democracy promotion) use it (2006, p. 18). From this definition he then identifies four key police reforms (2006, p. 18–19).

While neither Bayley’s (2006) definition of democracy nor his corresponding technical police reforms are wrong, they do problematically reinforce the idea that both are politically neutral. Certainly there are good arguments to be made for liberal democracy and liberal democratic policing. However, it is not the only definition of democracy or democratic policing. To assume there is one definition of ‘democratic policing’, that it corresponds to a single understanding of liberal democracy, and that it can be exported as a technical and politically neutral democratic fix, is simply inaccurate. It also problematically obfuscates the cause of COP failures and persistent police violence as a lack of ‘political commitment’ rather than recognise the politically contested nature of COP and democratic policing itself. While I do not dismiss the establishment of an ideal, such an ideal must be situated within the political debates on the concept.

In Latin America, democratic police reform projects are nested within deeply divisive debates regarding the definition of democracy (Nun 2003, Munck 2015) and the place of policing within it. Such debates are not unique to Latin America, but rather reflect the unresolved tensions between democracy, political ideology, and policing in all countries. Democracy is the people's sovereign yet in every democracy the police function as the coercive arm of the state. How best to balance this tension depends on political ideology.

I argue that there are four typologies of democratic policing: right, centre-right, centre-left, and left. Thus, rather than start with policing practices as a technical matter, and assume a consensus on the concept of democracy, I flip the conversation around and frame this article around political ideologies and their corresponding definitions of democracy and policing. This reflects that democracy (and thus democratic policing) is a ‘multidimensional and multilevel concept’ (Goertz 2006, p. 6).

Methodology

To capture this complexity I apply Goertz's framework of three-level concept analysis. The first level of analysis is what Goertz calls the basic level. This is the core meaning of the concept that remains constant regardless of what adjective is attached. Putting aside normative issues, the basic core element of democratic policing is that all four typologies contain the minimalist democratic feature of regular, free, fair, and competitive elections, which enable political parties and leaders with opposing ideas on policing to compete, debate, and pursue different policing practices and policies.

The secondary level of analysis is where we identify the characteristics we attach to the concept when connected to particular adjectives (in our case, the four political ideologies). The characteristics of democratic policing, such as equality and respect for civil and social rights, will vary depending on political ideology. Like some definitions of democracy, some definitions of democratic policing encompass a broader range of democratic practices than do others. COP is compatible, in particular forms, with some concepts of democratic policing more than others. Thus I begin this article with a brief theoretical discussion of the contested meanings of democracy based on political ideology and their associated visions for policing (including COP).

The article then moves to Goertz’s (2006) third level of concept analysis, the indicator/data level. The cases studies of Argentina and Chile provide empirical evidence to assess how the theory translates into practice. This section is informed by field research in both countries from 2006 to 2015 and draws on media and human rights reports as well as secondary data. Argentina offers examples of the concept of democratic policing on the right (Menem) and left (Kirchners) within the context of an institutionally weak delegative democracy (O’Donnell 1994). Chile offers the same for the centre-right (Piñera) and centre-left (Bachelet) within the context of a state with stronger liberal democratic institutions. The case studies are not meant to be an exhaustive overview of these governments’ democratic policing reforms (for more extensive analyses see Bonner 2014, 2016, 2019). Instead, the brief case studies aim to illustrate how the use of a concept of democratic policing that is based in political ideology can open up new research questions.

In this third level of concept analysis I find a mismatch between theory and practice. In practice, there appears to be a convergence across the political spectrum that favours the use of the right/centre right concept of democratic policing. Rather than undermine the value of theoretical typologies of democratic policing based in political ideology, this finding highlights their importance for better understanding political debates on police reform.

Political ideology and definitions of democratic policing: theoretical considerations

That there is not one definition of democratic policing and corresponding reforms, such as COP, is not surprising. As Sklansky (2008) notes: ‘no such account is possible, and no such agenda can make much sense. The meaning of democratic policing will remain in flux – because our ideas about democracy and our ideas about the police will remain in flux’ (2008, p. 191). Certainly, the US (as well as other countries), NGOs, and international organisations have promoted community policing as conceived by both the left and right (as well as other variations). One interviewee from the Washington Office on Latin America explained that there are a tremendous number of bilateral and multi-lateral donors interested in police reform in Central America and as a result in the

Best-case scenario you had a lot of duplication of efforts and in other cases just a contradiction in terms of what the donors were doing. [For example] in Honduras at one point you had three or four different community policing programs. All based on different models.¹

So what then are the issues of contention in the concept of democratic policing and how does political ideology shape different definitions? At its core, democracy is about public participation in state decision-making. However, as Bobbio (1996) argues, the degree of participation one sees as intrinsic to democracy is rooted in key values that separate the political right and left, as well as the extreme and moderate versions of both ideologies. These values pertain to perspectives on inequality/equality (a key distinction between the right and left) and the justifiable use of violence (a key distinction between moderates and extremists).

To summarise what follows, based on theory, the four typologies of democratic policing are outlined in Table 1.

Table 1. Typologies of democratic policing.

	Right (neopopulist)	Centre-right (conservativism)	Centre-left (social democratic/liberal socialism)	Left (populist)
Democracy	Electoral democracy	Liberal democracy	Social democracy	Electoral democracy with social rights
Inequality	Natural	Natural	Structural	Structural
(Police) Violence	Acceptable	Minimum necessary	Minimum necessary	Acceptable
Participation/Political Rights (affects community participation in COP and repression of protests)	Elections, limited free speech, COP unlikely	Elections, free speech, top-down COP	Elections, free speech, Co-equal COP	Elections, more limited free speech, COP unlikely or top-down
Civil Rights (affects levels of acceptable police violence)	Selective	Important	Important	Selective
Social Rights (affects degree to which COP will include social programmes)	Not important	Not important	Important	Important

As with all typologies, these are ideal types in which exant governments will fit in gradations Figure 1 illustrates these gradations.

Figure 1. Gradations of inequality and violence.

Equality/ inequality

Participation in state decision-making can be more or less equal. On the political right socioeconomic inequality is perceived as natural and good; inequality is what allows societies to accomplish great things (Bobbio 1996, pp. 67–68). Thus it is not surprising that Jones, Newburn and Smith find the New Right in Britain to prefer police reforms that apply ‘business methods to policing’ with negative consequences for equity (1996, p. 190).

From the perspective of the right, more minimalist definitions of democracy are preferred as they better accept and facilitate natural inequality. Schumpeter (2003 [1943]) offers the most widely cited minimalist and procedural definition of democracy as elections (2003 [1943], p. 269). He explicitly states that citizens should not participate in politics between elections and, more specifically, experts, not elected representatives, should decide crime policy (2003 [1943], pp. 292, 295). In this case, COP is not an option.

The purpose of policing in a democracy is to uphold the rule of law. For the political right, the law is treated as politically neutral. Since it issues from democratically elected representatives, it is assumed to embody and govern morality. Those who engage in deviant behaviour are naturally ‘bad people’ from whom ‘good people’ need to be protected (Lakoff 2002, Reiner 2016, pp. 40–42). Since such inequalities in goodness are natural, rehabilitation or socioeconomic preventive policies will not reduce crime; punishment is the logical response. Thus, for the political right, more police, with greater powers, will reduce crime and can be understood as democratic policing.

‘Broken Windows’ (BW) is emblematic of right and centre-right approaches to policing that accept inequality as natural. Kelling and Wilson introduced the concept to the general public in their 1982 article in The Atlantic. It has since sparked a large amount of (contested) research and, with the help of the Manhattan Institute and other organisations, has spread internationally – including to Latin America. For example, Rudy Giuliani, the New York mayor (1994–2001) who first implemented BW, and his consulting firm, Giuliani Partners, have made many trips to Latin America to promote and assist with the implementation of BW, ‘zero tolerance’, and this version of COP (Kubal 2012, Müller 2016).

BW juxtaposes and makes irreconcilable the concepts of ‘order’ and ‘equality’. Order is achieved through the police suppression of marginalised people such as ‘panhandlers, prostitutes, drunks, addicts, rowdy teenagers, prostitutes, loiters, [and] the mentally disturbed’ (Kelling and Wilson 1982) who are viewed as the source of disorder given that they ‘bother’ neighbours and establish an environment conducive to the escalation of violent crime.

This acceptance of inequality as natural then translates into a particular vision of community policing. The centre-right is theoretically more willing to accept liberal democratic processes of decision-making (Munck 2015) and thus is more likely to be open to COP than the right. If COP is accepted, ‘communities’ are perceived to inherently prefer ‘order’ to equality. The order desired by each community will vary but consists of informal rules that may or may not correspond to the law. The police gain legitimacy and the community's trust by enforcing these informal rules, more than by solving crimes (Kelling and Wilson 1982, Fournier-Simard, this issue).

There are also accepted inequalities between communities. This is where the BW's approach to COP is consistent with other conservative police reforms such as hot-spot policing and crime mapping (Manning 2010, pp. 168–178). Essentially, some communities need less policing because they have order or are beyond hope. Those on the cusp of disorder will benefit most from large numbers of foot-patrol police officers (Kelling and Wilson 1982). Which communities are to be policed most is a top-down police decision.

These ‘reforms’ are consistent with long existing police practices, maintain police hierarchies, and reflect ‘a preoccupation with trust and legitimacy, and a nostalgic appreciation for order and decorum’ (Sklansky 2008, p. 75, also see Gianikis and Davis 1998; Manning 2010). Rather than build trust through co-equal participation and improved police practices, preference is given to changing public perception. COP is then symbolically inclusive of the community, more than participatory. To the degree that the community does participate, it is comprised of those people police identify as ‘citizens’, in opposition to those deemed ‘criminal’ or ‘disorderly’. To the extent that social rights are considered, it is to ensure the legitimacy of elite rule (Nun 2003, pp. 52–53, Fournier-Simard, this issue). Symbolic NGO-run social programmes may be viewed as acceptable and sufficient, as the right and centre-right prefer a minimal role for the state, consistent with natural inequality and neoliberal economic policies.

Certainly, as Manning (2010) notes, these police reforms are based in conservative political ideology, not rigorous academic research on policing or crime control. Indeed, Manning reveals the supporting research to be incorrect, un-provable, contradictory, and inconsistent with liberal democratic ideals of equality (Manning 2010, pp. 132–134). He concludes that this makes BW ‘undemocratic’ (Manning 2010, p.154). However, this conclusion is not entirely correct. While inconsistent with some forms of liberal democracy and not an effective way to reduce crime, it is consistent with the more minimalist definition of democracy as electoral choice.

On the political left, socioeconomic inequality is perceived as socially created or structural and therefore part of the definition of democracy. People are born equal and then societies, particularly through the distribution of labour, make people unequal (Bobbio 1996, pp. 67–68). Since inequality is structural, it is possible and desirable for democracy to strive to become more equal, including addressing questions of political economy (particularly social rights) that prevent some members of society from full political participation (Bobbio 1996, ch.6, Nun 2003, Munck 2015, pp. 368–369). This requires a more maximalist definition of democracy, as steps must be taken to increase the range of people in society who have both the means and opportunity to participate in state decision-making.

Democratic policing can then be understood in a number of ways. First, democratic policing can involve less policing and greater attention to the preventive causes of crime, particularly those rooted in socioeconomic inequality. That is, social programmes (including welfare, public health, and public education), not police, are central to crime control (e.g. Reiner 2012). Some COP programmes give a nod in this direction with the inclusion of social services (by state or, more commonly NGOs) in selective neighbourhoods. However, many advocates of this maximalist or social democracy argue that the state needs to provide universal social programmes in order to reduce the stigmatisation of some people and limit opportunities for services to be exchanged in patron-clientelist negotiations (Nun 2003, p. 108).

At a similarly structural level, critical criminology contends that the police and criminal justice system need to redefine crime. This viewpoint holds that criminal law itself has been created by the powerful to their advantage leading to the criminalisation of deviant behaviours more common among the lower classes or other marginalised groups. In contrast, corporate or state crimes, whose harm to society can be much greater, do not face as significant punishments (Riener 2016, pp. 86–103). The law is not neutral but rather embedded with inequalities that need to be removed. Policing as the enforcement of the rule of law can then be deemed problematic depending on the laws enforced. Socioeconomic issues must be considered. Thus from the perspective of the left and centre-left, the state has a more expansive role to play that challenges neoliberal economic policies (which call for the reduction of social services). From this perspective, COP is fine but does not address these larger issues.

More centrist approaches to crime control, those Reiner (2016) labels ‘left realists’, hold that preventive socioeconomic programmes are good long-term goals. However, in the short term, increased policing is required but needs to effectively include affected community members in police decision-making. Indeed, this reflects the roots of COP in the US left of the 1960s, which advocated for a move away from more elite-focused pluralist democracy toward civil society-focused participatory democracy (Sklansky 2008, ch.3 and 4). This included the idea of police working with members of the community and community organisations in a co-equal manner to identify and prevent security issues of concern to that community (Alvarado, this issue). It also involved calls for greater participatory democracy within police institutions and a decentralisation of police power to officers so they could more easily use their discretion to address issues in ways appropriate to the community (Bittner 1970, Muir 1977; Jones et al. 1996). Community members themselves would act as the oversight that would keep officers accountable and check potential abuses of power (Trojanowicz and Bucqueroux 1990). COP is more consistent with the left-realist position than critical criminology.

In all cases on the left, if pursued, COP aims for the ideal of a co-equal partnership between community members and the police in which citizens can have a real (not simply symbolic) impact on policing that is constructive and holds police abuse of power in check (Herbert 2001, p. 448, Arias and Ungar 2009). Social programmes, as a preferred means to control crime, also remain important to all definitions of democratic policing on the political left.

Violence

As Bobbio (1996) recognises, the political left and right are not two homogenous wholes. There are important differences between what he calls moderates and extremists on both sides. He argues that the primary dividing line between these groups is violence and this includes violent policing.

Instead of violence, moderates, Bobbio explains, value ‘prudence, tolerance, calculating reason and patient mediation’ (1996, p. 25) and these values are consistent with the compromises required in liberal democratic processes. For moderates, democratic policing will include the protection of civil rights and favour checks (mechanisms of accountability) that limit the use of violence by police in both crime control and protest policing (aspects of liberal democracy). However, the definition of crime, the relative importance placed on the police in crime reduction and protest management, and acceptable forms of protest will vary between the centre-left and the centre-right. COP will also vary between the centre-left and centre-right on the scope of actors who can hold police accountable (Fournier-Simard, this issue).

Parallel with Bobbio's description of moderates, both the centre-left and centre-right in Latin America favour aspects of liberal democracy (Munck 2015). The centre-left favours equality and democracy and is manifested in the political ideologies of social democracy or ‘liberal socialism’ (Bobbio 1996, p. 79). The centre-right favours inequality and democracy (with equality in the law defined by an impartial judge, not the formation of law), and is manifested in conservatism (Ibid). However, Munck (2015) finds that in practice, conservatives in Latin America are willing to use higher levels of violence (notably police violence) and abandon aspects of liberal democratic checks on power more often than social democrats.

Extremists, Bobbio argues, interpret history as moving forward through forceful breaks with the past, not compromise (1996, p. 22). From this perspective then violence is justified, democracy is looked down upon as ineffective, and authoritarianism is preferred (Bobbio 1996, p. 25, 78). If extremists do accept democracy, then civil rights do not hold a firm place but rather can be compromised as needed to achieve the historical break with the past. Extending Bobbio's argument, extremist definitions of democratic policing (if valued at all) will favour police using high levels of violence both in crime control (although the left and right will vary on their definition of crime) and particularly in the management of public protests (notably of protests that challenge the desired break with history). In this context, policing can be violent and yet consistent with this definition of democracy.

To be sure, Bobbio's examples of extremists (Communists on the left and Fascists/Nazis on the right) are not democratic nor are they dominant in Latin America today. However, as we will see, some governments, often characterised (sometimes self-characterised) as populist or neopopulist (on the left and right respectively), do fit many of the characteristics outlined for extremists. While populism is a fuzzy concept, I define it as: a charismatic leader who evokes and claims to represent ‘the people’ by uniting disparate demands with symbolic language (what Laclau (2007) calls an ‘empty signifier’) against an ‘enemy’ with which ‘the people’ are claimed to have irreconcilable differences. Populism, as a political strategy used by all political leaders, can vary from more mundane uses to more rupturist forms (Laclau 2007). The more rupturist forms are most similar to Bobbio's extremists.

Populism is democratic in that the leader derives his or her legitimacy to speak for ‘the people’ from elections (as plebiscites) and possibly public opinion polls. On the left, populist governments also advocate for social rights. However, populists on the left and right often disregard the checks and balances of liberal democratic institutions (representing them as impeding the will of the people) and hold that the civil rights of some can be sacrificed to fulfill the will of ‘the people’. The police are often called upon to use violence to control ‘the enemy’ (however defined).

Political ideology and democratic policing in Latin America: the case studies

Moving to Goertz’s (2006) third level of concept analysis, this section explores the empirical evidence. It applies the theoretical typologies of democratic policing presented in the previous section to specific governments and examines their approaches to equality and violence as they relate to crime control and protest policing. The governments chosen were selected based on Munck’s (2015) classification of Latin American governments by political ideology. While not exhaustive, the case studies chosen aim to generate hypotheses that can later be tested or refined with additional or more in-depth case studies (for further details on the case studies presented here see Bonner 2014, 2019).

Democratic policing from the political right

The theoretical typology of democratic policing on the political right defines democracy as primarily electoral, views inequality as natural, and accepts high levels of police violence to achieve a desire break with history. In Latin America, some scholars have labelled these governments ‘neopopulist’ for their use of the populist strategy to implement neoliberal economic policies (Weyland 2003). Drawing on the examples of Fujimori's Peru (1990–2000), Menem's Argentina (1989–1999), and Collor de Mello's Brazil (1990–1992), Munck finds that the neopopulist right in Latin America has sacrificed liberal democratic decision-making in order to facilitate the implementation of these reforms. Most often this has come in the form of using, even illegal, presidential decrees, delegating decision-making to technocrats (e.g. central banks), and using police forces to repress opposition. Occasionally, the right has sacrificed even elections, as witnessed with Fujimori's self-coup in 1992, and support from the right-in-opposition for coups in Venezuela (attempted in 2002 and 2014), and Honduras (successful in 2009) (Munck 2015, p. 377). Crime control rarely includes COP, as elections (and public opinion polls as plebiscites) are generally regarded as sufficient public participation.

In Argentina, the government of President Carlos Menem (1989–1999) provides a case study of democratic policing on the political right. While not all political right governments are the same, an analysis of the case of Menem's government offers an initial test of the theory with the aim of generating new research questions. Menem accepted elections as an important part of democracy but was willing to abandon liberal democratic institutions of decision-making in order to implement neoliberal economic reforms (Ferreira and Matteo 1998). Despite a relatively recent return to electoral democracy (1983) and calls for police reform in the 1984 truth commission, Menem pursued few reforms of the federal forces under his jurisdiction, the Argentine Federal Police (PFA). Menem's first appointment to chief of the PFA was Jorge Luis Pasero, who, in his five-year term, reinforced a strong vertical structure within the forces with little internal or external transparency (Andersen 2002, p. 319). Even when the PFA were found to have been involved in the 1994 bombing of the Jewish community centre, AMIA, no action was taken to increase civilian oversight. Instead, Menem expanded the forces’ size, budget, and responsibilities.

For Menem, the police played a central role in fighting common crime. In 1998, he gave the PFA the responsibility of combating drug trafficking; a responsibility he transferred to them from the Buenos Aires Provincial Police (Andersen 2002, p. 328). He defended police institutions against accusations of police violence, increased the PFA budget by $60 million pesos, better equipped them, and increased their numbers by 5000 officers (Ibid, p. 346).² Menem also clarified that crime was not the result of unemployment.³ Echoing Giuliani, he explained that the only way to combat crime was ‘Zero tolerance. Iron fist. There is no other way as I see it. Human rights organisations can cry to the sky’⁴ It was under Menem's government that the grassroots organisations CELS and Correpi began to collect statistics on police violence. Correpi (2018) reports that security forces killed 710 people during Menem's government.

In addition to crime control, Menem's support for violent policing and acceptance of inequality as natural was also reflected in his approach to protest policing. Menem was responsible for the radical implementation of neoliberal economic policies in Argentina, which increased inequality in wealth in the country from a Gini index of 42.8 in 1986 (prior to his election) to 53.8 in 2002 (after his term in government).⁵ These economic reforms contributed to, not only a significant rise in unemployment, but also a corresponding rise in social protest, most notably by piqueteros. Piqueteros emerged in the mid-1990s and were initially comprised of workers newly unemployed as a result of neoliberal economic reforms. Throughout the late 1990s, piquetero protests became more frequent, growing from 140 in 1997 to 252 in 1999 (Epstein 2003, p. 22). Piqueteros, as well as others protesting Menem's neoliberal economic policies, were met with increasing police repression throughout the late 1990s (Amnesty International 1996, pp. 77–78, Human Rights Watch 1999, pp. 102–103).

In sum, democratic policing under Menem included police repression of protests, most notably anti-neoliberal protests. It employed a police-centred definition of crime control (COP was not pursued (Kubal 2012, pp. 193–194)) that emphasised a tough on crime approach and limited community participation in security policy primarily to public opinion polls and elections (Weyland 2003, pp. 1105–1106). It generally accepted socioeconomic inequality as natural and emphasised the role of the police in protecting the desired break with history – the implementation of a neoliberal economy. Overall, this definition of democratic policing favours electoral democracy rather than liberal or social democracy or COP.

Democratic policing from the centre-right

The theoretical typology of democratic policing on the centre-right defines democracy as liberal democracy, views inequality as natural, and favours minimual levels of police violence. Munck (2015, p. 378) finds the centre-right in Latin America (such as Chile (1990–00; 2010–14), Colombia (1990–2015), Mexico (2000–15), Venezuela (1989–99), Bolivia (1985–05), and Argentina (1999–01)) to be much more supportive of liberal democratic institutions of decision-making and electoral democracy. However, they are willing, occasionally, to use the power of the president to override Congress and more commonly to use police (as well as other security forces) to repress, most notably, anti-neoliberal protests (dramatically seen in Venezuela in 1989, Argentina in 2001, and Bolivia in 2003). Crime control can include COP but favours models that accept inequality of participation as natural and minimise the need for social programming.

Chile's President Sebastián Piñera (2010–2014) offers a case study of democratic policing on the centre-right. Piñera generally respected liberal democratic institutions of decision-making. However, he also accepted restrictions on some civil rights in the interest of neoliberal economic policy objectives and with an acceptance of socioeconomic inequality as natural. Similar to Menem, he acknowledged the police (Carabineros and Investigative Police (PDI)) as autonomous and hierarchal institutions that sell their mission to the people. Indeed, Pinochet's constitution (still in place) guarantees the institutional autonomy of the Carabineros (Candina Polomer 2005, pp. 157–158). Unlike Menem, Piñera defended, rather than implemented, the neoliberal economic policies the Pinochet regime (1973–1990) established. Democratic policing under Piñera thus accepts inequality as natural but, unlike the theory, is willing to accept high levels of police violence to defend his government's preferred economic model. This is seen in both crime policy and protest policing.

The Piñera government framed the need to combat common crime as a technocratic police role separate from socioeconomic issues of inequality and needed in order to provide a good business environment. Indeed, since the return of electoral democracy, control of common crime has been a central policy issue for the political right in Chile (Ramos and Guzman de Luigi 2000). Piñera's public security plan, Chile Seguro (Safe Chile), included a promise to increase the size of the Carabineros police force by 21% (10,000 new officers) and increase PDI forces by 1000 officers within five years (Gobierno de Chile 2010, p. 32, Frühling 2011, pp. 116, 123). Piñera explained: ‘In our country, the presence of a Carabinero officer is synonymous with security’.⁶ The increase in officers corresponded with an expansion of the existing community policing programme, Plan Cuadrante (Block Watch Plan), to many additional neighbourhoods, particularly in commercial areas. Chile Seguro also aimed to increase police powers, most notably those related to conducting identity checks, with no provisions for increased civilian oversight (Gobierno de Chile 2010, p. 39).

Plan Cuadrante is a COP programme introduced by a centre-left government in Chile and thus will be discussed in more detail in the next section. However, its continuation and expansion under the centre-right Piñera government reflects both the flexible definition of COP and the central role the Carabineros played in its application in Chile. The Chilean police have interpreted the idea of ‘community policing’ to fit their desire to maintain their centralised, hierarchal, and militarised structure (Candina 2006, p. 90). As it is for many police forces, the primary goal of COP as democratic policing for the Carabineros is improved community relations to increase public trust in and the legitimacy of the police, not community participation or changes to institutional police structures (Kubal 2012, pp. 194–195). Indeed, the Carabineros prefer the term policía de proximidad (proximity police) to community policing and adopted many of the reforms pursued by Giuliani in New York in the 1990s (such as COMPSTAT) (Ibid). As a public relations strategy it has been quite effective. The Carabineros are one of the most trusted institutions in Chile and one of the most trusted police forces in Latin America.

Piñera also supported the role of the police in maintaining the functioning of the market at the expense of the civil rights of those who oppose neoliberal economic and social policies. This was most notable during the large and regular student protests in 2011 and 2012 that opposed earlier neoliberal reforms of post-secondary education. Human rights organisations criticised police management of these protests. They reported many cases of escalated force without provocation, excessive use of water cannons and tear gas (including the frequent use of tear gas in closed spaces), beatings, arbitrary arrests, death threats, and sexual abuse (e.g. CECT 2012, pp. 211–226, Human Rights Watch 2013, INDH 2015). In 2011, Manuel Gutiérrez was killed by a police officer while watching one of these protests in his neighbourhood (Bonner 2018).

The Piñera government blamed protesters for the violence and sought to increase police powers in the management of protests. For example, in October 2011, then Minister of the Interior Rodrigo Ubilla explained that ‘The government strongly condemns these actions [of protester violence]’ as these groups are ‘extremely coordinated’ and have the sole objective of ‘producing damage and destruction’.⁷ Later that year, the government put forth a bill titled ‘Maintaining Public Order Law’ (Ley de resguardo del Orden Público). It aimed to increase the criminal punishment associated with causing public disorder from a minimum prison sentence of 541 days to three years (CECT 2012, p. 232). It also expanded the types of acts considered to cause public disorder to include not only acts of violence on people and property but also interrupting or paralysing public services and the free and unaltered circulation of people and vehicles (CECT 2012, pp. 232–233).⁸ That is, it aimed to severely limit the civil rights of freedom of expression and protest, while expanding police powers to arrest.

In sum, the definition of democratic policing found in practice during the Piñera government favoured electoral democracy, despite his greater willingness to use liberal democratic institutions of decision-making compared to Menem. His government made police principally responsible for decreasing crime. COP plays an important symbolic role aimed to increase police legitimacy, but is approached in a manner that is top-down and limits community participation. Police violence is acceptable when the civil and political rights to protest challenge government interest in protecting the (neoliberal) economy and the idea of socioeconomic inequality as natural.

Democratic policing from the centre-left

Theoretically, the centre-left supports a more maximalist definition of democracy. Ideally this involves not only elections, but also decision-making through liberal democratic institutions, and the protection of civil, political, and socioeconomic rights (social democracy). These guarantees hold police violence in check and strive to reduce the structural barriers to greater equality. The centre-left views socioeconomic rights as necessary to reduce socioeconomic inequality, to enable citizens’ exercise of their civil and political rights, and, thus, socioeconomic programmes are viewed as an important part of crime prevention (ideally more than the police).

Drawing on governments such as those of Chile (Lagos and Bachelet, 2000–10, 2014–2018), Uruguay (Vázquez and Mujica, 2005-present), and El Salvador (Funes, 2009–2014), Munck finds the centre-left to have the strongest record, of all ideological persuasions, of upholding liberal democratic decision-making processes as well as civil and political rights. However, they achieved only limited socioeconomic improvements (Munck 2015, p. 380). Indeed, an analysis of President Michelle Bachelet's governments (2006–10; 2014–2018) reveals a gap between their rhetoric and practices on equality and police violence, as they relate to crime control and protest policing.

Certainly, the Bachelet governments were respectful of liberal democratic institutions of decision-making. In addition, they made attempts to address socioeconomic inequality through programmes such as a 2007 pension reform and a 2009 law Chile Grows With You (which provides comprehensive social services to vulnerable children). In rhetoric, she defended civil and political rights, and more concretely inaugurated the National Institute of Human Rights in 2010. Yet, such changes were ‘often extremely slow and gradual’, owing to the governments’ preference to maintain an elite consensus that rests on a definition of economic success that favours growth over equality (Weeks and Borzutzky 2012, pp. 102–103). Consistent with this approach, Bachelet's governments maintained and strengthened definitions of crime control-as-policing and police repression of protests persisted.

While Bachelet recognised in her rhetoric the importance of socioeconomic approaches to crime control, her primary focus was to expand the number of police officers, their powers, and their budget. For example, in 2008, the government presented a bill titled ‘Short-term Anti-Crime Agenda’ (Agenda Corta Antidelincuencia), which was passed into law during Bachelet's second mandate in 2016 (Law 20.931). Among other punitive policies, the law increased police powers and reduced judicial oversight. Cases of unnecessary police violence reported to military courts went from 675 in 2004 to 1797 in 2011 (Diego Portales 2015, p. 134). According to the Diego Portales Human Rights Centre, judicial actors explained this increase in cases to be partly the result of this bill (and later law), as it gave police confidence that they could use higher levels of violence (Diego Portales 2015, pp. 135–136).

Moreover, in 2014, the first year of her second term in office, Bachelet passed law 20.801, which committed her to increasing the number of Carabinero officers by 6000 within four years.⁹ In 2015, she opened 13 new police stations.¹⁰ She explained: ‘they [Carabineros] should be where people need them: in work and recreation spaces, patrolling neighbourhoods, protecting people on their daily travels, in schools and plazas, in public spaces we want free and without fear for everyone’.¹¹ The Bachelet government framed these reforms as part of a ‘modernization’ of community policing efforts that she referred to as ‘Plan Cuadrante 2.0’.¹²

Indeed, like Piñera, Bachelet equated community policing with democratic policing. While the Carabineros have successfully resisted any significant internal organisational, disciplinary, or doctrinal reforms in the post-authoritarian period, they have been open to democratic policing as community policing. Plan Cuadrante (Block Watch Plan) was one of the first community policing plans. Established under President Ricardo Lagos (centre-left) in 2001 and continued under subsequent governments, Plan Cuadrante was the Carabineros’ own response to international and national pressure for them to move toward community and problem-oriented policing (Kubal 2012, p. 194). Plan Cuadrante increased the number of police on the streets; reorganised police to increase their numbers in key areas; established performance indicators to evaluate officers; and, aimed to strengthen their relationship with the community (Frühling 2003, p. 37). Thus, while a social democratic approach to policing would aim to decrease the use of police and the criminal justice system in favour of socioeconomic approaches to crime control, the centre-left, at least in Chile, have in practice reinforced the police as central actors in crime control more similar to a left-realist approach without the co-equal partnership with the community.

Certainly, the Bachelet government was more likely to condemn, or at least not encourage police repression of protests, compared to her centre-right counterpart Piñera, but she did not do so consistently. For example, in 2006, during her first term in office, high school students took to the streets to protest in large numbers. These protests were met with high levels of police repression, which the president condemned, but only those ‘excesses’ recognised by the Carabineros themselves.¹³ While her government pursued the dismissal of particular Carabinero officers deemed responsible for the excesses, she also prepared a law (passed in 2009) that would increase penalties for protesters and protest organisers that provoke ‘disturbances in the street’. ¹⁴ Police violence and repression of protests has continued and, in some cases, increased. For example, the National Institute of Human Rights surveyed people held in police custody and found that 34.33% reported arrests with violence in 2013 (under Piñera) and 82.75% did in 2014 (Bachelet's first year of her second term in office) (INDH 2015, p. 59).

To be sure, there are important structural and strategic limitations to what Bachelet could do in terms of pursuing a more social democratic definition of policing. At a structural level, these limitations stem in part from her respect for liberal democratic institutions of decision-making, as well as the restraints within these structures that are a legacy of the Pinochet regime (such as the Carabineros’ constitutional autonomy). Moreover, the positive public image of the police and the primacy of security in the dominant right-wing media, made it politically difficult to pursue social democratic policing reforms that would, for example, more extensively shift spending priorities from policing to social programming (Bonner 2019).

Context aside, democratic policing under the Chilean centre-left can be summarised as supporting, in rhetoric and some policies, a state role in reducing socioeconomic inequality (consistent with social democracy), yet in practice prioritising a police-centred response to crime control, including top-down COP aimed to increase police legitimacy. The governments were also willing to place significant limits on the civil and political rights to protest. In sum, centre-right and centre-left democratic policing is remarkably similar in practice. This finding is worth testing with additional case studies.

Democratic policing from the left

Theoretically, the left favours a definition of democratic policing that emphasises electoral democracy with social rights, views inequality as structural, and police violence as acceptable if it supports the desired break with history (left populism). Munck (2015) identifies Latin American political leaders on the left as including the governments of Venezuela (Chávez and Maduro, 1999-present), Argentina (Kirchner and Fernández de Kirchner, 2003–2015), and Ecuador (Correa, 2007–2017). These leaders place a premium on the reduction of socioeconomic inequality but, unlike the centre-left and consistent with populism, they are willing to sacrifice liberal democratic processes of decision-making in favour of alternative means of consultation, such as constituent assemblies or popular votes (plebiscites or elections as plebiscites). In practice, this has led to the concentration of power in the president, and attempts to secure indefinite re-election (Munck 2015, pp. 379–380). Democratic policing from this perspective is more likely to support socioeconomic approaches to crime control (with varied success) and accept high levels of police violence to control or repress political opposition to the envisioned break with history. COP is unlikely and if pursued would likely be top-down.

The Kirchner governments in Argentina (Néstor Kirchner, 2003–2007, and Cristina Fernández de Kirchner, 2007–2015) offer a case study of the left to test the theory. Similar to the centre-left, the Kirchners’ populist left governments in Argentina emphasised in their rhetoric and some of their policies, socioeconomic rights as central to democracy and crime control. For example, they increased the minimum wage and pensions, put in place a Universal Child Allowance (asignación universal por hijo) for the unemployed or those with black market or precarious employment, and in 2006 Néstor Kirchner stated that democracy and security ‘are not constructed with a stick in the hand, with attitudes, nor particular schemes, rather through encouraging education’.¹⁵

They were also important advocates for civil rights in their rhetoric and some of their policies. Among their many related projects, Néstor Kirchner lifted amnesty laws, which reopened trials against human rights abusers from the last dictatorship (including police). In 2010–11, Cristina Fernández de Kirchner initiated a short-lived reform of the PFA, headed by human rights activist Nilda Garré. It included a version of COP consistent with the left (emphasising socioeconomic solutions, co-equal participation of citizens and citizen oversight of the police), but prioritised purging the forces of officers involved in violence, crime, or corruption and asserting ‘political control’ over the police (Ministerio de Seguridad 2013, Bonner 2016). Political control included increased concentration of decision-making in the executive.

Yet, in tension with these accomplishments, Correpi reports that the highest levels of police violence since the return of electoral democracy occurred under the Kirchner governments.¹⁶ Of all the cases of police killings in democracy that the organisation has collected since 1983, 57.86% of these occurred during the Kirchner administrations (2003–2015) (Correpi 2018). Such violence reflects approaches to crime control and protest policing that are distinct from the above rhetoric and policies, and more similar to crime control under previous governments. Indeed, Seri and Kubal (2019) track a convergence of political leaders’ discourses on the ideological left and right on human rights and security from 2003 to 2015. This convergence led to an eventual common frame of security as a human right that can be protected through increasing the presence of police officers.

For example, justified as providing poorer communities their human right to security, the Kirchner government's 2011 Operativo Centinela (Operation Sentinel) sent 6000 members of the Gendarmería (militarised border police) to occupy shantytowns in Greater Buenos Aires, with another 4000 sent in 2013.¹⁷ The same year they launched Plan Unidad Cinturón del Sur (Southern Belt Unity Plan) that sent more than 3500 members of the Gendarmería and Prefectura (militarised coast guard) to occupy communities in the poorer southern neighbourhoods of the city of Buenos Aires.¹⁸ While some residents welcomed the influx of police as providing them greater security, human rights organisations received many reports of police abuse against residents, particularly youth, and police involvement in crime (CELS 2013, pp. 146–151).

Thus, as with the centre-left, the populist left has combined socioeconomic and police-centred approaches to crime control, with an eventual emphasis on the latter. However, this shift in policy maintained the rhetoric of the important role for the state in addressing inequality, but an inequality in security defined as common crime and with a remedy of increased institutional police force presence.

Similar to their position on crime control, the rhetoric of the Kirchner governments was always firmly against the repression of protests. Upon assuming office Néstor Kirchner announced that his government condemned police repression of protests and instead supported ‘tolerance and persuasion’ as the best approach, a position continued in rhetoric by Fernández de Kirchner (CELS 2008, p. 231). Despite this policy, in 2010, PFA officers killed two protesters in Indoamericano Park. In response, the federal position of non-violent protest policing was re-emphasised in a new national protest policing protocol in 2011, which included a ban on police bringing firearms to manage protests (including lead or rubber bullets and tear gas), reaffirming Kirchner's 2003 position (Bonner 2016).

However, the ombudsman's office (Defesoría del Pueblo) explained that police repression of protests tends to focus on smaller protests (that attract less media attention) and those without political connections (i.e. oppose the Kirchner government).¹⁹ Indeed, the president of Correpi (a civil society organisation critical of the Kirchner government) contends that she never received as many attacks against her as she did under the Kirchner governments.²⁰ Thus while not openly encouraging high levels of police violence and sometimes even condemning it (an important distinction from the right and centre-right), the left has been open to using the police to enforce support for their socioeconomic policies, programmes, or government.

Thus the populist left favours a definition of democratic policing that, in rhetoric and some policies, supports community participation and socioeconomic approaches to crime control that aim to reduce socioeconomic inequality and condemns police repression of public protests (all consistent with social democracy). Yet, in practice, the record on each has been mixed at best. Thus the practice of democratic policing from this perspective does not diverge significantly from what is found in practice under centre-right governments. Crime control is primarily police-centred and levels of police repression of protests can be high (albeit targeted), all consistent with more minimalist definitions of democracy.

Conclusion

In sum, democratic policing is not a singular concept. It can be as minimal as: regular, free, fair, and competitive elections that allow voters to choose between leaders with different ideas on policing policies and practices. Or, it can be more robust, requiring liberal democratic checks and balances or even non-police focused socioeconomic reforms (social democracy). Political ideology, and its corresponding perspectives on equality and violence, provides the values or justifications political actors can use to defend their use of more minimalist or maximalist definitions.

Of course, as the case studies illustrate, political leaders do not always pursue policing policies that are consistent with the purported political ideology of their party. This does not mean that the definition of democratic policing is not connected to political ideology. Rather, a focus on political ideology opens up new questions that might be overlooked if democratic policing is considered to be a singular and politically neutral concept. For example, it encourages us to ask: why has there been a convergence across political parties in favour of a definition of democratic policing that is on the political right/centre right? Possible explanations might include international pressures, changes in media coverage of crime and security, or other issues in need of more research. It also encourages us to explore how political leaders, civil society, or international actors interested in pursuing liberal or social democratic police reforms can better incorporate an understanding of political ideology into their negotiations with those who favour more minimalist definitions of democratic policing.

While the typologies here eschew normative evaluations of democratic policing, they do not preclude the establishment of an ideal. However, the typologies do reveal the importance of situating such ideals within political debates on the meaning of democracy. Democratic policing, including COP, reflects particular political and economic ideas about governance that are not necessarily shared across the political ideological spectrum. Moving from minimalist to maximalist definitions of democratic policing then requires a deeper understanding of politics. That is, we need more research on the structures, discourses, and power inequalities that favour the dominance of some definitions of democratic policing over others.

Acknowledgments

Thank-you to Guoguang Wu, Reeta Tremblay, Tina Hilgers, Arturo Alvarado Mendoza, and the anonymous reviewers for their valuable feedback on earlier versions of this article. Thank-you also to the Social Sciences and Humanities Research Council for the funding for this and related larger projects.

Disclosure statement

No potential conflict of interest was reported by the author.

Additional information

Funding

This work was supported by Social Sciences and Humanities Research Council of Canada [Insight Grant].

Notes

1 Author interview. Victoria, BC-Washington DC via Skype. 12 June 2015.

2 Clarín. ‘Menem: no queda otra salida que la mano dura frente a la inseguridad’ 13 September 1998. URL: https://www.clarin.com/politica/menem-queda-salida-mano-dura-frente-inseguridad_0_BylWSkWkL3g.html [last accessed 15 September 2017].

3 Ibid.

4 Ibid.

5 World Bank's GINI Index (where 0 is perfect equality and 100 is perfect inequality) URL: http://data.worldbank.org/indicator/SI.POV.GINI.

6 La Tercera. ‘Presidente Piñera firma decreto que aumenta dotación de Carabineros’. 15 Februrary 2011. URL: http://www.latercera.com/noticia/presidente-pinera-firma-decreto-que-aumenta-dotacion-de-carabineros/ [last accessed 6 October 2017].

7 ámbito.com ‘Gobierno de Piñera denucia complot para generar violencias en marchas estudiantiles’ 19 October 2011. URL: http://www.ambito.com/607271-gobierno-de-pinera-denuncia-complot-para-generar-violencia-en-marchas-estudiantiles [last accessed 6 October 2017].

8 The law was ultimately rejected by the legislature in 2013.

9 EMOL ‘Presidenta Bachelet en aniversario de Carabineros: La sensación de seguridad debe ser igual para todos’. 27 April 2016. URL: http://www.emol.com/noticias/Nacional/2016/04/27/800022/Presidenta-Bachelet-en-aniversario-de-Carabineros-La-sensacion-de-seguridad-debe-ser-igual-para-todos.html [last accessed 19 September 2017].

10 Ibid.

11 Ibid.

12 La Nación. ‘Bachelet aumenta en 6.000 los Carabineros y anuncia Plan Cuadrante 2.0’ 25 April 2014. URL: http://lanacion.cl/2014/04/25/bachelet-aumenta-en-6-000-los-carabineros-y-anuncia-plan-cuadrante-2-0/ [last accessed 19 September 2017].

13 Molina, P. ‘Zaldívar reinvindica rol de Carabineros’. El Mercurio. 2 June 2006, p. C6. Also see Bonner (2014).

14 Vergara, Rodrigo and águila, Francisco. ‘Gobierno prepara dura ley contra violentistas’. El Mercurio, 7 June 2006, p. C2.

15 Ámbito.com ‘Kirchner: La seguridad no se construye con un palo’ 7 September 2006. URL: http://www.ambito.com/294816-kirchner-la-seguridad-no-se-construye-con-un-palo [last accessed 19 September 2017]

16 At the current rate, the centre-right Mauricio Macri presidency will oversee a far greater proportion of police killings (Correpi 2018).

17 Lantos, Nicolás. ‘Agentes por gendarme’ Página/12, 28 February 2014. URL: https://www.pagina12.com.ar/diario/elpais/1-240797-2014-02-28.html [last accessed 3 October 2017]. In 2014, the 4,000 gendarme added in 2013 were replaced with Buenos Aires Provincial Police officers.

18 LaPolíticaOnline ‘Operativo Cinturón del Sur’ 4 July 2011. URL: http://www.lapoliticaonline.com/nota/53565/ [last accessed 3 October 2017].

19 Author interview. Buenos Aires. 2 December 2009.

20 Author interview. Buenos Aires. 9 November 2009.