Who Is the Citizen in Citizen Security?

This article explores the governing logics and political implications of citizen security in Puebla, Mexico, specifically, the Programa Nacional de Prevención del Delito (National Program for the Prevention of Crime, PRONAPRED). Taking a biopolitical perspective, the article centers on how citizens are produced through the application of citizen security initiatives. This production operates at two levels: the regulation of populations and the molding of subjects. First, vital statistics and risk-related variables highlight those populations that exceed the median for risk. Second, those at-risk groups are targeted for training to reduce their susceptibility to violence. Training focuses on the individual, asking citizens to identify and minimize risk in a manner that emanates from and contributes to security governance. The political ontology of citizen security is this citizen/citizenry coincident with official measures. The democratic aspirations of citizen security are consequently muted, as security governance locates citizens within the confines of state power and, at the same time, holds them separate from this power to produce citizen existence.

homicides per 100,000 residents in 1980 to 22.8 in 2016 (Gagne 2017)-the demand from civil society for greater safety only made citizen security more urgent.
The United Nations Development Program (UNDP) is vocal in this regard. In a landmark publication, Citizen Security with a Human Face, the UNDP placed citizen insecurity as "the cause and consequence of development problems confronting Latin America and the Caribbean." Consistent with a human security framework, the report juxtaposes "work on justice and law enforcement" with "preventive measures like generating employment opportunities and promoting social inclusion" (UNDP 2013, iii). The report posits the primacy of human life over the security of the state and confirms a multidimensional approach by positioning public security alongside "economic growth, employment, health, education and social development" (UNDP 2013, 13). The World Bank and the Inter-American Development Bank have also promoted preventive measures, with the latter having realized over 130 programs throughout Latin America since the 1990s (Aguirre and Muggah 2017;Frühling 2012). Some of the more prominent measures include Fica Vivo in Brazil (Arias and Ungar 2009), Plan Cuadrante in Colombia (Rincón Morera 2018), Barrio Seguro in the Dominican Republic (Cunningham et al. 2008), the Sistema Metropolitano de Seguridad Ciudadana in Ecuador (Torres Angarita 2011), and gang-related negotiations in El Salvador (Bruneau 2014). In Peru, for example, citizen security initiatives were to replace repressive mano dura (iron fist) policies with an emphasis on social prevention and control to limit the proliferation of violence (Instituto de Defensa Legal 2003). While these programs do not operate under a single definition of citizen security-often oscillating between a UNDP-inspired human security approach (Bonner 2008) and mano-dura-like policies (Holland 2013)-they attempt to mitigate crime by strengthening the capacities of the state and encouraging citizen participation. Rather than forwarding a specific definition of citizen security, this article instead centers on the relation between citizens and the state as forged through citizen security initiatives. It does so in reference to the Programa Nacional de Prevención del Delito (National Program for the Prevention of Crime, PRONAPRED) in Mexico.
Citizen security is well established in Mexico. Although recognized by multiple state and municipal governments in the mid-1990s, with many renaming their Secretarias de Seguridad Pública as Secretarias de Seguridad Ciudadana, citizen security came to national prominence starting in 2010 (Solórzano Peña and Contreras Acevedo 2015, 859). In that year, President Felipe Calderón implemented the program Todos Somos Juárez: Reconstruyamos la Ciudad to reduce violence in Ciudad Juárez. More recently, in 2014 the government of Enrique Peña Nieto established the Regulation of the General Law for the Social Prevention of Violence and Delinquency. Article 22 of the regulation is dedicated to citizen security, stating: "The Citizen Security Obligation of the State is to guarantee the security of the individual, acting in regard to the root causes of violence, delinquency and insecurity." 1 In contrast to such guarantees, however, this article suggests that citizen security initiatives forge a particular image of both the citizen and security. The state and not the citizenry defines "the root causes" of existential threats and grants itself the power to mitigate such insecurities. Yet the obligation goes further, extending to the construction of citizens themselves. Citizen security not only makes citizens the referent object of security but also locates them as subjects within an ever-shifting interface of governmental technologies and expectations of individual conduct.
This article concerns how citizens are reproduced through citizen security. Rather than evaluating particular initiatives or offering policy recommendations, discussion centers on interrogating the rationales informing citizen security. Following suggestions by Enrique Desmond Arias and Mark Ungar (2009, 409), such an approach addresses limits in the literature that overlook the political relationships of citizen security. More than a technical problem to be solved, citizen security is a political issue concerning how citizens are made to act and the relation this engenders to government. The article accordingly focuses on the political consequences of citizen security, on how citizens operate amid security (Seri 2012, 1); how they are activated as security "stakeholders" monitoring themselves, others, and their surroundings (Jarvis and Lister 2010, 182); and how, on this basis, everyone "is potentially the police" (Doty 2007, 132). Yet, rather than policing per se (Seri 2012), and in contrast to human security approaches and the protection of all life (Bonner 2008), the article interrogates citizen security in relation to what Barbara Cruikshank (1999, 1) calls "technologies of citizenship." Citizen security does not take human life as it is but forges a particular image of life, be it via "discourses, programs, [or] other tactics aimed at making individuals politically active and capable of self-government." To focus on how citizenship is constructed through security initiatives is to appreciate how security discourses forge the objects of which they speak. Understood in relation to the work of Giorgio Agamben (1998), security governance entails the ongoing production of political existence or bios, with the citizen an effect of citizen security. Departing somewhat from Agamben, however, the interest below is not in asking which lives are secured, thereby locating bios alongside the exclusion of bare life or zoē. Rather, it explores how "at risk" individuals become targets of citizen security initiatives to be incorporated into the state, albeit in a manner that distances them from political power. This more subtle inclusive exclusion within bios is examined in relation to how citizens are activated amid government policies, with official data-rather than victimization surveys or qualitative data from interviews-the means through which this activation is charted (Bergman 2006, 220).
This political approach to citizen security overlaps with biopolitical security. Michel Foucault introduced this approach in his lecture series Society Must Be Defended; Security, Territory, Population; and The Birth of Biopolitics, and in the first volume of The History of Sexuality. Of interest to Foucault were the political rationalities and technologies of power involved when regulating life, whether understood individually or via the population. To regulate a population involves managing the circulation of elements related to life and death within a determinate milieu and, on this basis, deciphering risk. An apparatus of security that concerns ever-greater circulations and their ever-expanding relationality is established, thereby allowing statistical calculations of probabilities of violence and crime (Foucault 2003a, 246-247;2007, 20-21). If the population is administered through this "science of security" (O'Malley 2010), then individual life is located in what Michael Dillon (2015, 105) calls the modern ontology of security: citizens emerging from the uncertainty associated with living amid (and preventing) potential violence. On the basis of risk, subjects are gradually brought into the fold of the state not through repression, but a power more productive that instills life capacities consistent with preventing crime. To paraphrase Foucault (2003, 30), the citizen is "one of power's first effects." Citizens assemble information, material and practices together into strategies that identify and minimize risk, strategies that emanate from, and contribute to security governance (O'Malley 2000, 465).
The citizen of citizen security is charted through this management of populations and subjects in relation to PRONAPRED in Puebla. Puebla is emblematic of the larger, nationwide PRONAPRED initiative. Although state and local governments selected which policy options to implement, the options were designed federally, as were the metrics through which Puebla and its population were mapped. Local developments are made co-extensive with a national orientation, with PRONAPRED less working to diffuse power to local actors than centralize power consistent with official rationales. The first section details how the population is deciphered through the accumulation of data concerning nationwide patterns of crime and violent homicides. The population is made visible through security-related optics, giving the state an image of the citizenry drawn directly from the requisites of security governance. On this basis, those populations outside what is considered empirically normal become the object of further attention. This is where citizen security turns on the subject, with outlying populations targeted through specific interventions that tailor existence. The second section explores these interventions, be they training modules, workshops, or infrastructure projects within PRONAPRED that mold citizens. However, more than a focus on individual life itself, the object of citizen security also concerns the conditions from which life emerges. That is, citizen security not only focuses on immediate lives but operates within a preventive register wherein the likelihood of future crime obliges interventions into communities to condition how life develops. The article concludes by locating this construction of populations and subjects alongside the democratic aspirations of citizen security.

Citizen Security and the Population
PRONAPRED was designed and implemented nationwide between 2013 and 2016. In contrast to the militarization of the previous Felipe Calderon administration, PRONAPRED sought a social approach to reducing violence. As the incoming president, Peña Nieto, made clear upon taking office, violence and crime are "not only combated with force. It is essential that the State undertakes a comprehensive endeavor to reconstruct the social fabric" (México Evalúa 2014). PRONAPRED thus became the centerpiece of a new, preventive approach to security during the Peña Nieto administration, the importance of which was confirmed in the Plan Nacional de Desarrollo 2013-2018. On this basis, the program would subsequently work through two complementary fields: the social prevention of violence and the containment of crime. Turning to the mechanics of PRONAPRED, this approach operates on the basis of statistically folding the citizenry into the state to later enable preventive measures. In reference to the document "Anexo Único" it reveals how PRONAPRED comes together in three elements: (1) the demarcation of priority municipalities, (2) the polygonal mapping of these municipalities in relation to risk, and (3) the application of policies within these polygons. It is through the first two elements that the population as datum is constructed, composed of vital statistics and rates of crime and homicide.
PRONAPRED makes the citizenry visible through probabilistic profiling. Specifically, citizens are made visible through risk-related statistics combined with life-sustaining variables, which are then placed within a preventive logic. In contrast to rights-based appreciations, this is a statistically mediated take on the population based on risk profiling. As various authors attest, such an approach has a long history. François Ewald (1991, 196-199), for example, notes an overlap between probabilistic techniques and epidemiological knowledge that enabled insurers to identify the risk of ill health. The mapping of society involved seemingly unrelated elements being brought into a relation with one another to determine risk pools. These techniques, however, would expand beyond the insurance industry to allow the application of calculative practices when governing populations. Society became increasingly statistical, as logics of categorization underpinned classifications that, in turn, allowed for prediction and targeted interventions (O'Malley 2004). Located within a biopolitical register, the quantification of risk soon became tied to the everyday flows needed to sustain population life (Ewald 1993;Donzelot 1991). PRONAPRED is resonant with these calculative logics and predictive applications. The citizenry is made visible through risk in a manner consistent with what Louise Amoore (2007) terms "vigilant visualities." These visualities are securitized ways of seeing the population that draw together various elements to evoke an image of the citizenry and enable a preventive politics. The statistical accumulation of data does more than create an image of the citizenry; it is a visualization in relation to risk that allows citizen security to target specific populations.
The mapping of populations on the basis of risk began by selecting fifty-seven demarcations for priority intervention throughout Mexico, with a remaining forty-three classified as secondary areas for action. While this fifty-seven later expanded to seventy-three in 2014, eighty in 2015, and eighty-three in 2016, of immediate interest is how these areas were selected. Selection was based on three criteria: population, crime rates, and territory. The first criterion was compiled by the Cohesión Comunitaria e Innovación Social and Fundación Este País and required that each demarcation contain one hundred thousand inhabitants or would contain that amount by 2015 based on estimates by the Consejo Nacional de Población. This left 210 of the 2,441 municipalities in Mexico eligible for selection. One hundred of these were demarcated for intervention, fifty-seven for immediate action. Officials claim that the selection of this hundred was made on the basis of territory and crime rates. However, there is no information on the methodology behind the selection of the fifty-seven priority zones. Somewhat confusingly, there is information on the selection of the forty-three demarcations of secondary priority. Their selection was based on twenty-three variables broken into two groups: twenty-one that relate to socioeconomic factors and two that concern violence and crime (see Table 1).
Ranking the forty-three municipalities involved the average of each variable being calculated to find which municipalities exceeded the median. The program compared the municipalities according to two categories, with 50 percent of the final ranking determined by their position relative to the first group of variables (1-21) and another 50 percent with respect to the last two (22-23). An image of the population is arrived at through this statistical analysis of reproductive properties (variables 1-21) and their interconnection with indexes of risk (variables 22 and 23). Through this statistical logic, a municipality's potential to experience crime became calculable. PRONAPRED distributed specific characteristics (the rate of violent robbery and violent homicide) against certain types of individuals (citizens under the age of twenty-nine, singleparent families) and their correlation with the scarcity or abundance of particular elements (rate of traffic accidents, number of occupants in dwellings). However, to repeat, this ranking only applies to the fortythree municipalities that did not receive funds. The fifty-seven designated as high priority do not correspond with this metric. If the same process of ranking is extended to all one hundred municipalities, then the fiftyseven most pressing cases do not match those earmarked for action.
While uncertainty surrounds the selection of high-priority demarcations, it is clearer how the population is made visible and intervention is made possible through statistically patterning security-related elements. Rather than local appreciations or models developed for Puebla, citizen security turns on big-data-inflected rationales based on population-wide circulations. Making risk visible does not render the entire population subject to intervention but isolates particular areas outside what is considered normal. The constitution of population in relation to risk, then, cannot be understood in the usual political sense. It is not the inscription of identity per se but an image arrived at by reducing the infinite variations of life to governable elements and their relation to one another: rates of crime and violent homicide, territory, population, variables 1-21 and 22-23. Moreover, these images are less patterns of specific behavior than "fuzzy" patterns that work through medians from which to judge deviations (Dillon 2008, 322). The aim is then to work on particular populations to shift them from implacable abnormalities to manageable susceptibilities (Rose 2007, 154). Technologies of surveillance become entwined with modern practices of governance insofar as they regulate corrupting factors that threaten the security of the population (Bell 2006, 156).
What makes intervention possible is not these images but their location within a preventive politics. From images of risk we move to visualities, as outlined by Amoore. Specifically, these are visualizations that isolate populations exceeding the median and, on this basis, assert that there is a greater likelihood of future crime. In practice, this means that there is a shift away from traditional models of retrospective intervention toward predictive measures based on the calculative rationality of risk. That is, more than past rates of crime and violent homicide, this population-wide profiling becomes predictive in that it offers a visualization of future violence. In PRONAPRED, this visualization concerns the construction of security polygons. Specifically, Puebla, alongside the other fifty-six selected municipalities, is mapped into polygons based on the most at-risk zones within each demarcation. While the selection criteria for demarcations remain opaque, those for polygons are even more so. The arbitrary nature of polygonal selection is evident in how official documents often fail to specify the communities that make up the polygons and instead create generic maps that lack identifiable geographical coordinates. The subtleties of difference-between individuals in polygons or the multiple milieus that comprise these demarcations-are minimized with the citizenry assembled through predicted risk based on their physical location. In Puebla, this abstraction is evident in one of its four polygons named 2111401_14; 21 is the state code (Puebla State), 114 the municipal code (Puebla City), 01 the first polygon of the demarcation, and _14 the year (2014) for which the polygon exists (see Figure 1).
Rather than local topographical markings, statistical epistemologies partition municipalities and reassemble them into polygons to align with the requirements of citizen security. PRONAPRED constitutes a geo-inductive approach, with all citizens within the polygon now the object of security governance.   (2010) 11. Rate of municipal growth (2005)(2006)(2007)(2008)(2009)(2010) 12. Population density (2010) 13. Percentage of the population not born in the municipality (2010) 14. Percentage of the total population of 5 years of age or over that lived outside the municipality (2010) 15. Percentage of the population under 29 years of age 16. Percentage of homes that have a female head of household (2010) 17. Percentage of homes that have a female head of household (2010)  Populations are extracted from a statistical paradigm, with probabilities of risk determining communities, to enable intervention. The following section turns to a more detailed appreciation of this intervention.

Citizen Security and Subjection
If a statistical appreciation of population makes life visible through risk, it also enables the training of individual life. Indeed, visualities of risk are not confined to populations but allow for the promotion of responsible citizens (Amoore 2007, 141). Citizens are to become active agents of security, increasingly taking charge of reducing crime. This arrangement is examined via the three objectives of PRONAPRED as implemented in Puebla: (1) "to increase the co-responsibility with the citizenry and social actors in the social prevention [of violence] through their participation and capacity building"; (2) "to reduce the vulnerability to violence and delinquency in populations of high priority"; and (3) "to generate environments that favor coexistence and citizen security." By moving through these three objectives the aim is to reveal how citizens are constructed in PRONAPRED. Rather than an arrangement wherein individuals recognize themselves in their singularity, citizens are to emerge amid a mode of being consistent with, and generative of, security governance (Foucault 1990, 89).

First Objective: Co-responsibility
At its most basic, PRONAPRED binds citizens to a securitized identity that, although consistent with external authority, develops through how they constitute themselves when faced with risk. Through various initiatives, citizens become attentive to their conduct by bringing into play a certain relationship through which they discover themselves amid security governance. The aim is to demonstrate this relation in reference to the first objective of "co-responsibility" and its subsequent actions: objectives are broken into strategies and actions.
Responsibility is central to biopolitical security. Understood in reference to literature on responsibilization, citizens are encouraged to rationally conduct themselves within security governance. By operating amid a "web of vocabularies, injunctions, promises, dire warnings and threats of intervention," citizens discover themselves through minimizing risk (Miller and Rose 2008, 205). Developed in reference to Foucault's work on governmentality, responsibilization is more than a regulatory power imposed on life, as it also concerns how individuals recognize themselves amid such power. Translated to citizen security, the interest centers on both a politics of subjection and how individuals constitute themselves as responsible citizens. If the former are institutional codifications regarding how citizens should act, the latter are how individuals enact operations on themselves to realize a securitized way of being. As Foucault (1991, 95) asserts, governmentality is not "imposing laws on men [sic], but rather of disposing things, that is to say, to employ tactics rather than laws, and if need be to use laws themselves as tactics." To interrogate co-responsibility is to explore these tactics, not as regimenting individual behavior but as informing how individuals think of themselves as responsible citizens.
To better appreciate law as tactic and responsibilization more broadly, attention turns to the second PRONAPRED strategy within the objective of co-responsibility. Strategy 1.2 is "to develop capabilities in the citizenry and civil society organizations with respect to a culture of peace." These capabilities are promoted through action 1.2.4.1 and its workshop titled "Culture of Legality and Reporting." The action asks: "What is the state of law, and why is it important to me?" Understood tactically, educating individuals on the "state of law" is not a question of enforcement but ensures that citizens emerge from a legal understanding. Individuals are taught the "norms and laws of conduct" as well as "why the state of law is important in society" and "why the state of law is violated." This tactical deployment allows individuals to realize themselves in relation to the law-"why is it important to me"-with the "laws of conduct" not so much the object of legal oversight, but instead located within how citizens conduct themselves. The locus shifts from an "external" law to an "internal" economy that less punishes than it tethers citizens to official norms (Foucault 1995, 17-18). Co-responsibility begins by promoting a relationship that locates the self in conformity with the law, with the imperatives of security governance actualized through the ongoing work of the self with reference to "norms and laws of conduct." Co-responsibility is this discovery of the self through an officially recognized exploitation of risk. The abovementioned "Culture of Legality and Reporting" is indicative of this relation. Action 1.2.4.2, for example, "establish[es] an orientation module for victims of crime." On the one hand, the action informs individual conduct by prioritizing the act of reporting witnessed crimes to authorities. On the other hand, the emphasis placed on reporting not only trains citizens on how to respond to threats but instills caution in how they interact with their world. More than encouraging specific behavior-if I see someone suspicious I know what to do-the action makes residents mindful of their surroundings. Promoting "a culture of reporting and trust in authorities" reworks the individual based on an awareness of others: to be on the lookout for crime is to continually monitor all those around me. Responsible citizens are informed by the requirements of security governance, but co-responsibility also operates through a fashioned behavior that becomes a way of life: I am continually vigilant. By "strengthen[ing] the recognition and exercise of citizen rights and responsibilities," citizens discover themselves through this securitized disposition consistent with government.
Responsibilization is further revealed in strategy 1.1: "to implement mechanisms of inclusive participation and free of discrimination in the citizenry in local projects for the social prevention [of violence]." Action 1.1.4.1 consists of a workshop on "the prevention of criminal and high risk conduct" by "training the citizenry." Training involves citizens working on themselves according to security objectives to recognize "the importance of the prevention of criminal conduct, beginning with the factors that have their origin in the home that reach the community." Prevention is put forward as both a moral provocation and as commonsense measures to ensure safety, with "inclusive participation" providing the medium through which self-government connects with the imperatives of security governance. Similarly, action 1.1.4.2 concerns "training the citizenry with regard to situational prevention." This is to "establish a program named Alert Inhabitants which has the objective to train an organized population in schemes of self-protection and security." Citizens are to work on and evaluate themselves according to these schemes based on the personal responsibility "to respond and act opportunely before situations of risk" (Gobierno Municipal de Puebla, cited in Emerson 2020). By offering participants "diverse recommendations and means of self-protection," personal development is equated with preventing violence. Citizens gradually discover themselves through security governance, with "schemes of self-protection" becoming the matrices for individual conduct.
PRONAPRED works on the individual and the community alike. Community-based measures promote a specific type of participation through diffuse networks of actors. Action 1.1.2.1 involves the "elaboration and training of citizen networks of the social prevention of violence and delinquency." It establishes networks that are first trained in co-responsibility and then tasked with disseminating this knowledge throughout their communities. This relational appreciation of security fosters links between communities and local authorities in which a government-contrived set of arrangements not only recalibrates individual conduct but promotes relations of identification that align with official rationalities. Action 1.2.3.5 complements this approach by selecting "community leaders in crisis management and the resolution of conflicts." The aim is "to develop abilities of honest brokers between inhabitants of the polygons, reinforcing their knowledge in terms of the resolution of conflicts." Such initiatives reinforce a relation between communities and government, as state-derived knowledge on crisis management organizes the thoughts and actions of influential residents. "Honest brokers" repeat security governance through their engagement with violent contingency and redistribute its set relations by operating within the community. Just as biopolitics for Agamben (1998, 21) saw the sovereign extend its powers over individuals by entering into an alliance with "the jurist … the doctor, the scientist, the expert and the priest," such an alliance in PRONAPRED is achieved via direct engagement with community leaders. Community leaders are consistent with security governance, or, paraphrasing Didier Bigo (2002, 64), they become managers of risk who conform to a hierarchy of relations and, in the process, propagate its logic throughout communities.
Co-responsibility reveals how citizens become active partners in security, ensuring their own well-being and that of those around them. PRONAPRED feeds into literature on responsibilization insofar as, following David Garland (2001, 125), individuals are invited to regulate risk through their security choices. The program "raise[s] public consciousness, interpolate[s] the citizen as a potential victim, create[s] a sense of duty, connect[s] the population to crime control agencies, and help[s] change the thinking and practices of those involved." It does so (1) by promoting an appreciation of the self in reference to "norms and laws of conduct"; (2) by enacting a securitized disposition to oneself and one's surroundings; and (3) by forwarding a relational appreciation of security that emphasizes both self-organizing citizens and communities. Through each component a political relationship develops in how individuals come to conduct themselves responsibly in a manner coincident with government.

Second Objective: Reducing Vulnerability
The second objective within PRONAPRED complements responsibility by working through citizen vulnerability. Security governance operates through uncertainty to convert it into a means of informing securitized citizenship. To become a secure citizen one must first recognize and later implement strategies to overcome one's susceptibility to violence. Vulnerability consequently becomes an additional channel through which PRONAPRED realizes the transition from "at risk" individual to secure citizen, a channel dependent on the future likelihood of violence and the consequent need for present action.
Each strategy within the objective of vulnerability involves the "reduction of risk factors of violence and delinquency." This reduction ensures that citizens respond to potential harm according to the requirements of security governance. Again, however, regulation less disciplines subjects than it concerns a securitized disposition governing how individuals transition through uncertainty. It promotes particular interactional properties to inform how at-risk individuals conduct themselves amid contingency. These individuals are boys and girls (strategy 2.1), adolescents and youths (strategy 2.2), and women (strategy 2.3). A series of workshops target these groups not by regimenting their responses to violence but by tailoring how they emerge alongside uncertainty. That is, individual emergence amid vulnerability is conditioned so that the citizen, and its relation to risk, progressively align with official protocol. Vulnerability operates at an immanent level. While responsibilization concerns how citizens respond to already apparent threats-resolving existing conflicts, reporting witnessed crimes-a politics of vulnerability centers on how citizens act amid threats yet to materialize. Citizens are to spot the signs of crime before it is too late, and to anticipate danger before its violent consequences appear. Security governance focuses on how citizens inhabit this indeterminacy. However, this risk potential is more than simple uncertainty. Citizen security introduces a violent futurity, a futurity of potential danger that informs citizen conduct. That is, the future possibility of crime becomes the basis from which a secure citizen emerges, as future risk is sutured onto the present to inform individual conduct. Citizen security exerts a power over the future, foreclosing its inherent uncertainty by having it unfold on an already established ground consistent with its rationale. This takes two forms: how citizens appreciate a potential threat in others, and how they recognize this danger in themselves. These are far from mutually exclusive; in both instances security governance compels citizens to discover themselves amid insecurity by taking preventive action to stop threats from arising.
Attention begins with action 2.3.9.1 and the potential for risk in others. Concerned with the "psychological attention to aggressive persons and their families," the action works amid potential threats. It organizes "intervention workshops in families with aggressive persons," with these interventions to "detect problematic family members, who are to be channeled to the respective institutions to continue treatment." Intervention workshops operate through individual vulnerability to "problematic family members" by outlining the different signs, tendencies, and expressions that may appear in "aggressive persons" before they act aggressively. These elements can then be isolated, scrutinized, and measured in accordance with official appreciations. Citizens are to see the signs, discover the tendencies, and sense the expressions within violent family members. Or better, citizens draw from the indeterminacy of the looming threat, such signs, tendencies, and expressions. Secure citizens remain in contact with a potentially violent family member but are mindful of certain psychosocial signs of violence. The workshop thus imposes a degree of regularity through how individuals engage with indeterminacy. Certain stimuli are rendered problematic, thereby exerting their own pressure in terms of how risk unfolds, with this "danger" no longer conforming to its own immanent trajectory but according to these external dictates. A dynamic, open-ended future gives way to rigidity, wherein only certain factors are responded to, factors that confirm external authority as family members are "channeled to the respective institutions." This relation to uncertainty is further reflected in the emergence of the potentially dangerous self. Action 2.2.1.1 speaks of "formative activities of sexual education and reproductive health." These activities are complemented by lessons wherein experts offer "prevention campaigns on sexually transmitted diseases and on adolescent pregnancy." While targeted at vulnerable individuals, such activities do not require citizens to be mindful of specific, dangerous possibilities such as overly amorous teenagers or youths experimenting with their sexuality. Rather, it calls on these at-risk individuals themselves to emerge amid a wariness of their own dangerous although not yet fully determinate potential. Securing these citizens requires them to operate immanently to this emergent potentiality and to be mindful of the processes through which risk may materialize so that they monitor themselves and disrupt danger. From within their own being they look out for risk potential that conceivably is just working through the boundaries of sexual, physical, and emotional well-being. Through this ongoing appraisal, citizens modulate how they emerge amid uncertainty in a manner consistent with security governance (Foucault 2005, 324-326). PRONAPRED organizes the thoughts and actions of citizens. Knowledge regarding one's own conduct now contributes to the state, as citizens are made consistent with official thinking that renders them, and the indeterminacy of the threat, more governable.
Vulnerability allows security governance to extend itself, as citizens become the means for administering risk. Limits to official knowledge-an inability to see potential threats-are circumvented by citizens implementing strategies to minimize harm. Uncertainty is captured not by surveillance measures but through the conduct of the vulnerable. By working through these individuals, citizen security becomes multi-scalar. Repeated through the daily lives of citizens, security governance reproduces itself across time and space as secure citizens unfold differently depending on their circumstances. Action 2.1.4.3, for example, offers a "School for Parents" wherein ten workshops "equip parents … with tools in order to identify possible victims of abuse within the school community." On the one hand, equipping parents can be interpreted as commonsense measures to ensure safety. Citizens recognize their own responsibilities and ensure proper familial relations. On the other hand, the action inaugurates a process of complex adaptation wherein a securitized disposition operates amid local knowledge within the school community. Secure citizens, rather than government, make the signs of possible abuse visible, operating amid and responding to vulnerability beyond official intervention. Although individual conduct is normalized, this normality now includes occasional excesses associated with how citizens operate in the community. These excesses are integral to security governance because they allow citizen security to open onto new horizons, horizons exceeding official risk calculation.
When juxtaposed with the first objective of responsibility, it becomes evident how PRONAPRED operates through a grid comprised of two intersecting horizontal and vertical axes. If the horizontal axis concerns a securitized disposition forever expanding beyond its bounds, then the vertical axis ensures that citizen conduct confirms the set relations of government. The horizontal axis allows citizen security to go multiscalar, while the vertical hierarchically interlocks these new actors and elements within government. To emphasize the horizontal dimension is to recognize how secure citizens are composed through a series of movements, desires, and transitions undertaken from a particular dispositional angle. Citizen security allows for the passage between these elements and the production of new meaning. In contrast, the vertical axis ensures that this novelty is made familiar to security governance. It does so by redirecting conduct to confirm official rationales, by drawing relations of constancy from uncertainty. When combined, the result is a kind of feedback loop between citizen and government. The infinite ways in which individuals might respond to their vulnerability to violence now oscillates within the recognizable bounds of security governance.

Third Objective: Generating Environments
While the first two objectives of PRONAPRED govern risk through informing individual life, the third objective turns on the conditions enabling life, or "to generate environments that favor coexistence and citizen security." Rather than focusing on life per se, citizen security intervenes into the environment in which citizens are bound to inform how they emerge. This approach underlines the importance of creating enabling spaces for citizens to work together and ensure security (Frühling 2012;Muggah 2017). Strategy 3.1, for example, is to "realize actions of situational prevention that contribute to the reduction of opportunities for violence and delinquency." By exploring the implications of "situational prevention," this final section focuses on the mechanics of intervening into the environment when located in a preventive register.
Analysis of the third objective of PRONAPRED begins with what Foucault (2008, 260) called "an environmental type of intervention." Introduced briefly in his lectures on The Birth of Biopolitics, Foucault hinted at this recalibration in governing life. In contrast to a modality preoccupied with circulations in set territories and displays of constancy in populations, attention instead centered on an "environmentalism open to unknowns and transversal phenomena" (2008,261). While he did not elaborate on this environmentalism, in the previous year's lectures Foucault (2007, 23) put forth an expanded governing of life in relation to nature. The sovereign, he suggested, is "someone who will have to exercise power at that point of connection where nature, in the sense of physical elements, interferes with nature in the sense of the nature of the human species." These "points of connection," however, are not regulatory mechanisms to establish standards, and nature cannot be reduced to an overview of milieu from which the population is governed. Rather, this is a mode of governing that acknowledges how life is wholly implicated in nature, with the individual in dynamic assemblage with milieu. Expanded on below, this environmentalism offers an appreciation of a situated life composed of the various elements in which it is bound, and "an environmental type of intervention" concerns how this setting is conditioned to inform life's emergence. It is toward the practical implications of such environmentalism that attention turns.
Various actions are consistent with environmental intervention. Action 3.1.3.6 concerns the "construction of a community center." Rather than targeting vulnerable individuals or intervening into life capacities to make citizens more responsible, PRONAPRED focuses on the environment in which at-risk citizens live. Indeed, intervention not only concerns the construction of a community center; it also informs collective existence by changing the conditions from which it develops. Acknowledged implicitly is the interconnectedness of security-related objects and subjects insofar as spaces are created to allow safe existence to follow. Intervention is made possible on the basis of a situated appreciation of existence, of how life is tied to the milieu in which it is bound. The aim of intervening in this milieu is to affect "a multiplicity of individuals who fundamentally and essentially only exist biologically bound to the materiality within which they live" (Foucault 2007, 21). Strategy 3.2 confirms this bias, as it is designed to "foster the appropriation of public spaces to strengthen coexistence and citizen security." This strategy consists of several actions, from a "community cinema" (3.2.1.1) to "music concerts with content that reflect and reinforce local identity and cultures of peace" (3.2.1.4). In each instance, intervention turns on a regeneration of milieu, a regeneration that is to result in the emergence of secure citizens. Security governance accordingly shifts from the administration of life to what Paul Rutherford (1999) calls control over the conditions of life.
This environmental intervention also operates in a preventive register. Action 3.2.4.1 confirms this bias, "promot[ing] actions that inhibit youths from integrating into gangs." At first glance, the action appears consistent with the abovementioned techniques. It engenders spaces from which a secure existence follows: stipulating actions that "contribute to improving their surroundings and areas of coexistence." It operates according to techniques of responsibilization: the actions limit youths from "adopting negative attitudes, through activities of citizen participation." And it is directed at vulnerable individuals who detect signs of risk: actions "addressing themes like … identification of violence, prevention of conflicts between urban groups." However, rather than repeat these dynamics, the action also introduces a temporality that exceeds the administration of life per se. That is, the action is made possible by projecting future risks and the consequent need to take preventive action. If this futurity was already acknowledged, its application to environmental intervention reworks how space is understood in PRONAPRED: from milieu to a protomilieu. Previously, the abundance or scarcity of particular elements demarcated possible futures: higher instances of homicide combined with a younger population or lower levels of education meant a more dangerous polygon. This projected possible future then informed action to be taken: training modules for parents or social activities to reduce violence. Where PRONAPRED expands on these effects is that its future projections are based not only on statistical approximations of past crime but also on events that have not yet come to pass. Rather than an empirical backgrounding of future risk, there is also a foregrounding to security governance wherein risky scenarios are projected to mandate present attention. Futurity is conditioned not only by possibilities informed by the (statistically appreciated) past but also by potentialities emanating from future risk-related imaginaries. These imaginaries, in contrast to empirical projections, operate less in relation to milieu and its associated circulations than through a protomilieu: projected dangerous environments that do not exist but that nonetheless condition how life is to emerge (Massumi 2015, 39-40).
This preventive logic offers a new vantage from which to understand PRONAPRED. Actions are based on a statistical backgrounding of future risk insofar as they are premised on polygonal mapping. However, they also exceed this backgrounding by operating through a futurity conditioned by imaginaries constructed within the program itself. Action 3.2.4.1, for example, consists of "actions that inhibit youths from integrating into gangs." It begins by outlining the need to realize training on the basis of "inhibiting" the likelihood of future violence. The action is made possible by projecting dangerous futures ("participating in criminal activities") if present interventions are not realized. Present action takes shape based on the violent potentiality evident not in the actual milieu (the circulation of existing empirics) but in a projected protomilieu. On this basis, PRONAPRED enacts workshops to shape the present-day emergence of youths in ways that "contribute to improving their surroundings and areas of co-existence." Citizen security does not exclusively instrumentalize the future via past distributions of risk (Hacking 1990), but it also deploys worstcase scenarios to complement its risk calculations. PRONAPRED consequently folds future uncertainty into governmental logics. The result is an overlap between an unlived future and a securitized present in which the citizen is situated in an officially mediated "discourse of futurity" (Schott 2013, 213).
The third objective reveals how citizen security concerns not only life but the environment from which life develops. PRONAPRED regenerates the settings in which secure citizens are situated, thereby conditioning the conditions for life. Yet, this intervention is made possible less through the actual milieu of existing circulations than on a protomilieu of imagined futures. In short, security governance forecloses future uncertainty by having it unfold on its own terms. This enacted future then becomes a field of emergence, as citizens are to materialize on the basis of this risky futurity. Futures are projected with the aim of sewing the citizen into this protomilieu, an imaginary territory over which PRONAPRED has exclusive authority. Citizen security is no longer solely concerned with regulating life but with bringing into being what its apparatus has already captured. To repeat, the citizen is an effect of citizen security.

Conclusions: Citizen Security, the Citizen, and the State
This article has explored the political relation forged between the citizen and citizen security. In contrast to a citizenry comprising multiple individuals or a body politic of particular identities, the population as datum is composed of vital statistics and their relation to risk. An image of the population is drawn exclusively from this calculative method, an image that visualizes the likelihood of future violence and crime through polygonal mapping. On this basis, the subject of security is forged. Yet, rather than a citizen with civil, legal, or political rights, this is a citizen reproduced within security governance. Techniques of responsibility collapse self-government into political government as citizens discover themselves amid officially recognized conduct, techniques of vulnerability ensure this overlap by informing how citizens emerge amid uncertainty, and techniques that generate environments confirm this arrangement by intervening into the conditions necessary for life so that it unfolds according to security governance. In short, citizens are made consistent with citizen security.
The ongoing reproduction of the citizen becomes the means for realizing citizen security. How citizens emerge amid violent contingency not only repeats a citizenship coincident with government but also extends security governance to new frontiers. First, government is confirmed as responsible citizens take an active role in preventing violence and thereby complement the gradual withdrawal of the state as premised in neoliberal arts of government. Citizen participation becomes the corollary to the withdrawal of the state, with individuals activated through techniques of self-esteem and consultation to become active, informed, and responsible members of self-managing communities (Dean 2010, 196). Self-mastery as the basis for safety is framed in the values of autonomy, self-actualization, prudence, and choice (Rose 1996, 155;2007, 125). The result is security governance repeated through individual conduct. Second, vulnerable citizens extend citizen security by overcoming limits to official knowledge. An appreciation of risk is not attained through official measures per se, but through the conduct of the vulnerable. Security governance works through these individuals and in the process expands, as citizens approach new scenarios from a securitized disposition consistent with official rationales. Third, the situated citizens of environmental intervention confirm this relation to security governance. Citizens emerge amid a looming threat imagined by government, required to forestall future dangers by taking present-day action; dangers that can only be preemptively mitigated by citizens reproducing official thinking.
PRONAPRED reveals how the relation between citizens and citizen security is an onto-political one of regulated populations and subjects. A statistical appreciation of the citizenry, combined with the subjection of secure citizens, reveals the ontological foundations of citizen security. This is not to divorce such ontologies from securitized epistemologies. Indeed, official images of risk-rates of crime and homicide-make the citizenry visible and, on this basis, enable the reproduction of secure citizens. Rather, it is to highlight that the efficacy of citizen security rests on the capacity of citizens to discover themselves amid governmentauthored knowledge forms. Moreover, the expansion of security governance confirms this ontological bias. The governing of complexity associated with problematic family members or dysfunctional communities is repositioned as an ontological rather than epistemological problem. Complexity is made intelligible not through surveillance measures but on the basis of how citizens engage with uncertainty. Citizen security is less government intervention into problematic areas based on an overview of milieu than it is citizens exhibiting a self-reflexive understanding of themselves and their surroundings in relation to security governance (Chandler 2014, 56). The result is that citizen security exhibits a particular hermeneutics wherein security governance becomes the occasion through which citizens experience themselves as secure subjects.
The onto-political basis of citizens in PRONAPRED undermines the democratic aspirations of citizen security. As various authors attest, citizen security aims to forge a relation between the citizen and the state that reconciles security with democracy (Bonner 2014) and move beyond the repressive state apparatus (Gasper and Gómez 2015). While PRONAPRED undoubtedly improves on the state-sanctioned violence often associated with mano dura policies, it is democratically reductive in two ways: it locates citizens within the existing confines of institutional power and at the same time holds them separate from such power. First, although encouraged to take an active role in their own security, citizens are to emerge exclusively amid security governance. Citizen existence is made coincident with the boundaries of official thinking, with citizenship reduced to questions of complicity with the imperatives of PRONAPRED. Second, and on the basis of this reductive vision, citizens are separated from political power. Compelled to act in a submissive relation to government, citizens are perpetually at a distance from policy-making processes. Unable to inform the development of citizen security initiatives, the recognition of citizens remains tied to how they perform their government-authored responsibilities (Hardt and Negri 2009, 346-347). This is the inclusive exclusion of the citizenry. Citizens are gradually brought into the state not through repression but a productive power through which they are to discover themselves amid, but not contest, governing rationales.
The three objectives of PRONAPRED can be appreciated according to this logic. Instead of empowering citizens, co-responsibility is read as an objectifying force, ensuring a narrow appreciation of participation. Rather than a securitized life developing on the basis of certain rights, as per the definition of citizen security by Mauricio García Ojeda and Alba Zambrano Constanzo (2005, 69), life is instead confined to government requirements. Responsibilization operates in a unidirectional manner that robs individuals of democratic potential (Dunn, Kaufmann, and Kristensen 2015, 7), with citizens unable to hold state officials to account (Ungar and Arias 2012). Similarly, by working through individual vulnerability PRONAPRED reproduces docile actors who are to accept their status as perpetually at risk. The need to survive amid the likelihood of future violence becomes the basis for individual existence. The political goals of citizenship are not concerned with questioning the sociopolitical order but with the imperatives of survival. Vulnerability becomes consistent with a political nihilism seemingly anathema to any democratic project (Schott 2013, 215). And finally, the generation of environments confines citizen emergence to future imaginaries over which the state claims exclusive authority. Rather than opening politics up to various visions, citizen security advances threat-related futures that further rob citizens of political engagement. A deliberative politics that allows for citizen action and organization are consequently forestalled, and the individual capacity to transform outside of security initiatives is negated (Ungar 2009, 242). Cumulatively, the result is a lack of active citizen engagement that could potentially guide and improve citizen security (Muggah 2017, 296), with security governance unable to consolidate itself through genuine citizen participation in its planning and implementation (Nateras González and Tinoco García 2014). By asking who the citizen of citizen security is, PRONAPRED thus responds with a citizenry through which it governs risk and over which it exerts control.