The Armed Arena: Arms Trafficking in Central America

To explain the high rates of violence in Central America’s Northern Triangle, this article argues that it is rooted in an armed arena comprised of state and nonstate actors with roles in security, ranging from the military to organized crime. In this space of collaboration, those actors bring resources and reap rewards, strengthening their relationships in the process. Of the many activities within the armed arena, one of the most destructive is the trafficking of firearms. Based on fieldwork in Honduras, Guatemala, and El Salvador, this article focuses on two sets of relationships that demonstrate the armed arena’s strength. The first set is the area of criminal justice, centered on investigations, ballistics, inventories, and destruction of arms. The second is the arms market: regional trafficking, internal arms sales, and the state’s licensing system. Evaluating governmental responses, the final section shows that arms control policies are most effective when they target and disrupt these relationships.

Why are Latin America's rates of violent crime persistently high? Nowhere is the region's crisis of insecurity more acute than in the countries of Central America's Northern Triangle-Guatemala, El Salvador, and Honduras-which have had three of the world's five highest murder rates over the past decade. Such entrenched criminality, this article argues, is rooted in what it calls an "armed arena," defined as a space comprised of state and nonstate actors with roles in security, ranging from the military to organized crime. Like civil society, this is an area of collaboration continually reinforced by the benefits to and contributions of each participant. While state security forces contribute legal authority, for example, criminal groups contribute transnational supply lines.
Of the many activities within the armed arena, from extortion to narcotrafficking, perhaps the most underestimated but powerful is the trafficking of firearms, broadly defined here as the movement and sales of lethal weaponry, from small arms parts to military materiel. Arms are the currency of carnage. Unlike other contraband, they do not just flow out or through Central America but engulf it in storms of violence. The vast majority of murders are committed with firearms. Approaches such as social network analysis (Morçol 2012) strive to identify the underlying dynamics shaping the relationships among actors involved in such activities. Of the many characteristics shaping the arena's innumerable relationships, this article

The Armed Arena
The armed arena is defined as the network of connections among the entities with a role in security. The state agencies range from executive policy makers to local police agents. Nonstate actors, meanwhile, range from narcotics cartels to neighborhood gangs. While some of the connections among these many entities are open and legal, such as state contracts with a private security agency, the majority are informal and illicit. They are also highly adaptable, due to the wide range of possible alliances (such as with different gangs in an urban area), the need to quickly respond to state policy (such as militarization of an area) and their wide realm of physical operations (from local storehouses to overseas air routes). The key dimensions of the armed arena, though, are its origins and catalysts. It is formed and strengthened by two consistent conditions: a lack of coordination among state actors involved in security, and coordination among nonstate actors involved in crime. One the on hand, state security agencies are continually undermined by ongoing changes in policy and personnel, along with a lack of resources and training. In contrast, nonstate actors thrive on uncertainty: the increasing ability to collaborate, share resources, move contraband, and contract out specific tasks make nonstate actors-primarily those involved in crime-more effective. Over time, these two patterns lead to a reconfiguration of relationships, in which state actors, from elected official to street police, gravitate toward nonstate actors through relationships of corruption and collaboration. Criminal actors, in turn, use those connections to infiltrate and shape the state. In the process, an armed arena emerges and establishes the conditions for its own reinforcement.

State Actors
A lack of coordination characterizes relations among the executive, legislative, and judicial state agencies that oversee security. While dominant presidents are seen as a threat to democracy in Latin America, in the realm of security their ability to impose their will is widely seen as beneficial. But such control obscures fundamental weaknesses, rooted in a lack of clarity and consistency. As a top public concern, crime has been a central electoral issue in nearly every country. 2 From Mexico to Chile, presidential candidates have proposed increasingly bold anticrime platforms, seen most recently in Brazil's 2018 election. Needing quick results more than long-term institutional change, winning candidates then give security forces more authority. The appeal of that choice is enhanced by their reliance on the police to control unrest that can culminate in government-threatening protests. In fact, in all seven collapses of democratically elected governments in Latin America since 2000, security forces played a pivotal role by participating in, aggravating, or failing to curb unrest. Destabilizing protests in Brazil, Venezuela, Guatemala, Honduras, and other countries since then further illustrate this dependence on police to maintain order.
Such uncertainty is further aggravated by appointments and policy. In addition to postelection changes in officials or party control, the ineffectiveness of different policies leads executives to continually replace security ministers and police chiefs, who as a result have little time to initiate new policies. Security policy is further destabilized by fickle foreign governments and international organizations that bankroll security reforms and often promote approaches whose translation into feasible policy is always underfunded and usually unclear. They are also sometimes contradictory, such as tensions between anti-violence programs and antidrug operations, and often politicized, when perceived to be treating on national sovereignty (Efrat 2012).
Worsening executive coordination on security throughout Latin America is the proliferation of security agencies, which undermine coordination as they outdo each other with more aggressive operations and political networking. Embodying the popular mano dura, militarization is a particularly popular but virulent strain of proliferation that takes many manifestations, such as anti-gang joint forces; elite units like Guatemala's Kaibiles; military-police drug units, as in Mexico; and military police, as in Honduras. There are few checks in most countries on such proliferation or on executive policy overall, particularly by the legislative branch, such as by monitoring military arms inventories.
Democratic Latin America's extensive decentralization makes such proliferation exponential. Transferring security authority helps regional officials to better respond to local conditions. But security decentralization, which has been advancing even in centralized states like the Central America states, is inconsistently implemented, further undermining coordination and leaving poorer high-violence areas more vulnerable to crime. Urban Central American gangs broken up by the police regroup in smaller municipalities, which are then overwhelmed by them, for example, when the fragments of dismantled drug cartels-lacking their leaders' directives, controls, or funding-turn toward crimes like armed robbery and extortion. Decentralization also aggravates conditions more pronounced at the local level, such as corruption and alliances or organized crime with local family, landowning, and extractive entities. All these forms of state fragmentation allow the arms market to expand, which of course further complicates arms control.
Relations among agencies that investigate crime-police, prosecutors, judges-are also uncoordinated. The first step in the criminal justice process is taken by the police, who, throughout Latin America, lack necessary training, equipment, and interagency coordination. They are directed by the prosecutors who formally lead investigations. But even when prosecutorial power expands, such as through anti-racketeering laws (see below), most Latin American countries suffer from a lack of detectives, specialized prosecutors, and local offices to acquire citizen trust and information. Crime reports then lack the evidence and witnesses needed for trial, contributing to the region's dismal average of under 5 percent of homicides leading to conviction. Such weaknesses are amplified with organized crimes like money laundering.
This lack of coordination characterizing intrastate relationships on security then makes it difficult to tackle crimes like arms trafficking. That difficulty is compounded as the resulting frustration leads officials to develop links with nonstate entities.

Nonstate Actors
That same lack of coordination play into the strengths of relations among the increasingly diverse set of nonstate actors involved in crime, which include not just criminal syndicates, but also rural militias, neighborhood watches, social services, self-defense units (as in Mexico's Tierra Caliente), residential property guards, "social cleansing" groups, errant community policing programs, and Central America's two highly organized, transnational gangs or maras, MS-13 and 18th Street. These groups' expansion is fueled in part by how easily they both fuse and conceal their simultaneous roles as vigilantes, law enforcers, service providers, and criminals. State-funded social work groups in countries such as Venezuela become drug pushers, for example, while corner-store businesses launder extortion cash. Other nonstate actors are formally legal, particularly the huge private security sector: Latin America has over 3.8 million private armed guards, dwarfing its 2.6 million police officers (though with much overlap). But the real size of this sector is far larger, since the majority of firms do not register. In 2015 alone, for example, Guatemala's Office of Private Security Services (Dirección General de Servicios de Seguridad Privada, DIGESSP) shut down forty unregistered firms, which had 5,000 guards and over 8,500 firearms.
And of course fueling and helping fund these groups are transnational syndicates, from narcotics cartels to migrant coyotes. Most scholarship on organized crime studies these actors through five dimensions long used to identify and distinguish among them. A first is structural, stressing organized criminal groups' centralized or hierarchal nature. A second is their crimes' complexity, from "craft" crimes like carjacking not requiring extensive planning, moving up to "market" crimes like narco-trafficking with greater volume, more cooperation (Albanese 2014), and longer supply lines. Greatly facilitating such operations is racketeeringinfiltration of state institutions in order to shape laws, alter regulations, and dilute policies. Aiding such infiltration is the dimension of social embeddedness, which is a group's integration in a society and economy, such as by providing security, jobs, and other services. A final broad dimension of organized crime is of the goods it traffics, which can have many subdimensions. One is of goods' legality, which range from absolute contraband (never legal, e.g., contract killings); relative (legal under restricted circumstances, e.g., firearms); and commercial (legitimate but marketed illegally, e.g., cigarettes).
But nonstate groups' rapid integration has shattered the categories within and distinctions among these dimensions. Throughout the world, growing markets, decentralized states, and technological leaps in communications open up unprecedented opportunities to coordinate physical, financial, operational, and electronic resources. From low-income barrios to high-end markets, the armed arena produces, transports, and sells an increasingly diverse array of products: narcotics, pharmaceuticals, firearms, property (ranches, auto shops); humans (emigration, prostitution); and natural resources (timber, fauna, minerals). Neighborhood extortion operations become part of cross-border narcotrafficking, for example, while lands cleared by loggers become airstrips for cocaine cartels. Studies of the arms trade, in fact, map out collaborate supply chains, such as those feeding the Rio de Janeiro market (Gaspirini Alves 2001). Such cooperation also eases entrance into the legal economy, particularly through the investment of contraband profits in licit goods and services like tourism and ranching. More than most crimes, arms trafficking embodies the fusion of these categories, since it extends from neighborhood "craft" crimes to high-level extortion, from local deals to the global economy.
In regions like the Northern Triangle, this coordination is further boosted by state weakness, which it turn heightens racketeering. Uncertainty or frustration over their own positions and agencies, in particular, leads many officials to reach out to nonstate groups. Police officers, for example, can leverage their legal powers of detention to acquire support from local gangs. As this article shows through arms trafficking, the weaknesses in the two main state areas in which the state creates and applies security policy-executive policy and criminal justice, as discussed above-lead to the kinds of links with nonstate agencies that form the armed arena.

Conceptualizing and Analyzing Arms Trafficking
Among the armed arena's most consequential interactions is the trafficking and use of firearms. While homicides committed with firearms account for 42 percent of homicides globally and 60 percent in South America (Castellanos 2015), they account for 75-85 percent of homicides in the Northern Triangle, killing over two thousand people each year, mostly with high-power arms like Uzis, AK-47s, assault rifles, and fragmentary grenades, used in about 70 percent of assaults and robberies. 3 This elevated regional rate reflects the vast number of circulating weapons. Guatemala has well over 2 million legal and illegal civilianheld arms; in Honduras estimates range from 800,000 to 1.7 million, of which approximately 650,000 are illegal; in El Salvador, nearly 740,000 firearms are in civilian hands, of which fewer than 200,000 are registered. 4 Such wide-ranging estimates reflect statistical unreliability rooted in poor documentation of renewed or registered weapons; in 2011, some Salvadoran officials documented under 7,000 registered weapons and others over 25,000. More fundamentally, such numbers reflect bonds among state and nonstate entities in the two main areas of criminal justice: the lengthy judicial process, in which the state marshals its enforcement and investigatory power against crime, and the arms market, in which illegal traffic is the main pattern in the robust regional weapons trade, from individual purchases to military aid.

Criminal Justice
This section shows how poor coordination through the entire criminal justice process increases arms trafficking and impunity for crimes committed with them. It focuses on the key steps in that process: investigations, ballistics, inventories, and destruction. Focus on each step brings out characteristics of criminal justice agencies that leave them vulnerable to nonstate criminal actors.

Investigations
Criminal justice begins with investigations, usually initiated by the police. But the region's officers are ill-trained for basic steps in crime scene investigation such as shell collection and witness interviews. Such weaknesses are rooted in poor internal coordination within the police and with other state bodies, as characterized by continual changes in policy, rotation of chiefs, promotion favoritism, management disorganization, unplanned switches in daily shifts, and uneven documentation of officers' weapons use. Such uncertainly is deepened by the long-politicized policy clash between human rights and pro-gun advocacy. El Salvador's vice president said that the police "must use weapons and should do so without fearing consequences for their actions" (La Prensa Gráfica 2015). This lack of consistency undermines policy, which in turn undermines citizen confidence in it. Cruz and Kloppe-Santamaria (2019) argue that support for extralegal violence increases in inverse relation with support for the existing political system. They cite a survey about support for extrajudicial violence of twenty-one Latin American countries, in which Honduras, Guatemala, and El Salvador are in the first, fourth, and fifth spots, respectively. In addition to undermining arms control, such conditions push police officers toward the nonstate actors with which they have more regular contact, further undermining criminal justice.
Directing investigations are the attorney general's (Fiscalía General, hereafter FG) prosecutors (fiscales). But Northern Triangle FGs, rife with inefficiency and corruption, are incapable of adequately investigating arms crimes. The broadest reason is sparse coverage: between 40 percent and 90 percent of Guatemala lacks a prosecutorial presence-a wide estimate that itself reflects a lack of basic information. 5 In Honduras, the 96 percent impunity rate for murders is rooted in the lack of detectives. Of the 391 homicides in the city of Comayagua between 2010 and 2012, in just 34 did investigators complete a report (Alianza para la Paz y la Justicia 2014). Guatemala's and El Salvador's FGs lack adequate management, data analysis capacity, and procedural clarity, from the admissibility of evidence to interrogation of police officers. A similar lack of procedural clarity at the FG is blamed for El Salvador's fall in seized weapons from an average of 4,602 between 2007 and 2011 to 2,991 in 2015. Even specialized units on arms crimes are understaffed. Although arms trafficking has been documented in all of El Salvador's fourteen departments, its FG arms trafficking unit (Unidad contra el Tráfico de Armas) has just four officers. It is part of the Organized Crime Unit, which has just twelve fiscales and a chronic lack of resources. 6 And nor is there much clamor for more; in surveys of public confidence, the FG ranks thirteen out of fifteen entities, with just 8 percent having confidence in it (IUDOP 2014). In Honduras, an overhaul of the FG since 2013 has boosted transparency and efficiency, but the chief of the Investigative Police Office (DPI, Dirección Policial de Investigación) says that investigation remains limited amid a lack of fiscales, operational equipment and ballistics labs-as emphasized by the 2018 reform, discussed below. 7 Such inadequacy continues in the courts. As in Guatemala, many Salvadorans say that "judges do not investigate because they know" too well the links among gangs, armed traffickers, state officials, and businesses like transport companies. As a result, many of them release suspects or annul charges (casación) on arms trafficking. The comparative rate of casación on arms with other crimes is unavailable in each Northern Triangle country, but every judicial official is certain that it is consistently and considerably higher. Criminal arms investigation is also obstructed by the military, such as Salvadoran army officers blocking FG investigators from entering facilities in several investigations. Such obstruction allows illegal networks to flourish, as in the case of retired military official Miguel Pocasangre, which involved serial number alteration, license falsification, and illegal sales over the span of a decade to customers such as state agencies and narcotraffickers. That case exemplifies how much investigation can be disrupted by deficient internal state cooperation, which in this case included the military's refusal to document the number of arms removed from its stockpiles.
In Honduras, two agencies-the DPI and the Technical Criminal Investigation Agency (Agencia Técnica de Investigación Criminal, ATIC)-were established to cut through such impediments. But their officials complain about terrible coordination among the many other entities still involved in criminal investigation: the FG, Supreme Court, the Procuraduría, several police divisions, and the National Interinstitutional Security Force (Fuerza Nacional de Seguridad Interinstitucional, FUSINA), which oversees all security agencies. One Honduran official stated that "even personnel within agencies do not talk to each other because of corruption, impunity and politics." Resulting impacts include poor crime scene investigation and bottlenecks in police reports. In interviews, fiscales complain of their inability to process the sixty thousand reports filed each year, leading to delays of eighteen months to three years. Poor statistical coordination extends those delays even more. Unlike FUSINA, the FG and ATIC say that they do not use the US-funded NACMIS (National Access Case Management System), while the DPI uses a police database, and the Forensics Office (Centro de Ciencias Forenses) yet another.

Ballistics
Police and prosecutorial weakness is aggravated by inconsistent ballistics testing by underfunded ballistics labs, located in the FG's Forensics division. El Salvador's Instituto de Medicina Legal has been hamstrung by fraud, inflated purchases, and institutional "collapse," in the view of a former director. Guatemala's Instituto Nacional de Ciencias Forenses (INACIF) has just eighteen specialists and a budget that was the same as in 2011, when it employed twenty-four specialists. Even though the approximately 16,500 arms seized since 2014 are a small number circulating illegally, they are still too many to process. Receiving over 100 arms monthly, say INACIF officials, the lack of personnel creates a backlog of 4,000 arms. This bottleneck is also rooted in a more regular use of weapons. Most firearms are discharged in Guatemala once a week, compared to the average US frequency of nearly six months. Honduras's Centro de Ciencias Forenses has just nine specialists, handling fourteen murders per day. The main ballistics program adopted around the region is the Integrated Ballistics Identification System (IBIS), based on the ATF-directed National Integrated Ballistic Information Network. But it remains underutilized, and so-like NACMISdoes not provide reliable gun data. In Honduras, officials complain that investigators are "not utilizing" the technology because "it is not important to them," even with external prodding. But police officers blame poor coordination, saying that "we can't get shell casings to" the IBIS team. The need for such testing is underscored by Honduran ballistics technicians, who say that for every 10,000 arms examined from crime scenes and police raids, they discover tens of thousands of cartridges of various calibers; bullets from different crime scenes; and a range of accessories like clips. So much potentially incriminating materiel is the foundation for, and demonstrates the need, to investigate connections among criminals-that is, the extent to which crime is organized. In Honduras, the last examination of this kind was in 2010 by the predecessor of the Centro de Ciencias Forenses, the Instituto de Ciencias Forenses (Sánchez 2013), an agency whose collapse is emblematic of the country's institutional uncertainty.
The final stage in the criminal justice process, the penitentiary system, continues such ineffectiveness and impunity. Throughout Latin America, arms fuel criminal organizing within prisons. In September 2016, for example, 750 prisoners were transferred out of San Pedro Sula prison in Honduras for arms and drug trafficking, but in July 2017, FUSINA found 50,000 shells of different calibers in that and nearby facilities. Similar trafficking plagues Guatemala. A July 2017 raid in its notorious Infiernito prison found grenades, munitions, and other arms. Much easier to detect than drugs, the movement of arms into prisons indicates complicity by state officials. El Salvador shows how such complicity undercuts state policy. In the 2017 trial of those who negotiated a truce between the two maras, witnesses testified that the political parties, in attempts to sway public opinion, paid the maras hundreds of thousands of dollars that was then used to buy M-16s and AK-47s.
Cooperation with nonstate entities extends across the state's institutional spectrum. In Honduras, most information on narcotrafficking was carried out by the Nacional Anti-Narcotrafficking Council (Consejo Nacional contra el Narcotráfico, CNCN), whose chief stated in 2010 that "all state organizations working on drugs are infiltrated by drug traffickers." 8 The council worked closely with the police's Information and Communication Center (Centro de Información y Comunicación, CEINCO), which registered arrests, operations, and seizures related to illegal trafficking in weapons, ammunition, explosives, and other material in connection with drug trafficking. After many years of police themselves saying that organized criminal groups infiltrated CEINCO, it was eventually disbanded, as was the CNCN in 2014, a few months after being charged by the Anti-Corruption Council of embezzlement.
As with other agencies, CNCN and CEINCO raise questions about the veracity of state data. How much do they collude with the nonstate groups they are supposed to control? Even if they do not, are their reports discredited by internal rule-breaking like accounting improprieties? In one of the Anti-Corruption Council's final annual reports, for example, it compiled statistics with CEINCO on narcotrafficking and arms seizures. Among the report's findings was a jump in seized weapons from just over 1,000 in 2010 to over 1,500 in 2011 and nearly 3,300 in 2012. Accounting for much of the 2011-2012 increase was a near doubling of seized pistols and revolvers, and of about 250 shotguns compared to no shotguns the year before. Whether such increases reflected a shift in arms trafficking and crime, or instead in police efficiency or statistical accuracy, was not investigated. Neither was analysis done of geographic overlap of arms seizure patterns with markers of organized criminal activity, such as locations of drug-processing labs and banks used to launder money. Even a cursory examination of the report shows a marked expansion in water-based narcotrafficking routes in areas with some of the highest arms seizures, such as in Colón (CNCN 2012). Without functioning and secure channels of interstate communication, however, such valuable information is not incorporated into policy responses.

Inventories
Although state arms arsenals should be closely monitored, there is endemic uncertainty over their numbers, types, and background. One reason is military control. The Salvadoran military oversees inventories of firearms by most other agencies, while Guatemala's Defense Ministry oversees all agencies that have arms, from the police to banks and private firms. In Honduras, the Policía Militar del Orden Público (PMOP), ATIC, and National Anti-Extortion Force (Fuerza Nacional Anti-Extorsión, FNA) each has its own inventory. Their weapons are often "lost," says a Security Ministry advisor, 9 which, as in El Salvador, has led the FG to investigate suspected caches. One of the region's biggest busts was a warehouse in San Pedro Sula owned by Mexican narcotraffickers, which had rocket launchers originating in a Salvadoran inventory earmarked for destruction. Within El Salvador, weapons held by the military are regularly sold to criminal groups. Military officials were convicted for trafficking in 2014, the same year a raid on the home of ex-defense minister Atilio Benítez found twenty-nine arms that should have been destroyed. Trial proceedings against Benítez began in 2019 for the sales of those arms, which were worth six million dollars (Calderón and Arbaiza 2019).

Destruction
In each country, weapons seized in crimes are earmarked for post-trial destruction. But that process is rarely transparent. In October 2011, for example, El Salvador's Diario Oficial announced impending destruction of three thousand arms, an amount later reduced without explanation. Arms destruction is public, in addition, but the military often bars the media. It often sells those arms as well; over seventeen cases have been opened of sales of high-caliber weapons that should have been destroyed (Cawley 2014). Military officials have allowed, encouraged, and participated in the falsification of permits in the army's arms registry; the fabrication of destruction documents for grenades that were later discovered in caches of organized crime groups in Mexico and Honduras; and complicity with the criminal group Los Perrones, which moves weapons into neighboring countries. Guatemala has a judicial agency to coordinate arms decommissioning and destruction but only two machines to actually destroy arms. In Honduras, poor coordination among the twenty agencies involved in arms destruction has led to a backlog of 3,000 weapons in 2017 even though just 1,551 arms were decommissioned between 2013 and 2017. Few of those weapons are tested to certify origin, owners, or previous use. According to officials speaking on condition of anonymity, that failure to test stems from several conditions. First, arms are brought in by at least five different agencies, with different protocols and forms, and held in warehouses spread out among the Security Ministry, FG, and judiciary. Second, most of those arms lack identifying marks like serial numbers, or are registered under fictitious names. Of the approximately two-thirds of decommissioned arms that are "artisanal" or modified, "identification" is thus nearly impossible. Table 1 summarizes key weaknesses in the main criminal justice stages-identified by fieldwork in each case, but likely characterizing the whole region-that leave agencies vulnerable to criminal groups.

The Arms Market
The fuel of the armed arena, of course, is a robust arms market. This section focuses on three dimensions of that market-uncontrolled borders, the arms business, and arms registration-to specify the mutually reinforcing ways in which institutional incapacity and politicization in the state and the highly lucrative nature of the arms trade make the market a nexus of the armed arena. This approach also adds to the scholarship on arms trafficking, which focuses on identifying the drivers of global and regional markets. Some scholars typologize the markets, such as on a continuum from white to black markets (Rothe and Collins 2011), while others draw causal connections, such as the between US gun control policy and violence in Mexico (Chicoine 2016). Much scholarship focuses on societal anxiety (Arsovska and Zabyelina 2014); in Central America, pervasive fears of crime and lack of confidence in the police are a constant impetus of citizen arms sales, as discussed below. 10 This article adds to each of these approaches-typologies, causal relations, and societal demands-through its focus on the armed arena's relationships.

Borders
Uncoordinated border control fuels regional arms trading, demonstrating how cooperation among armed arena actors takes advantage of weak interstate coordination. Nearly every regional frontier, such as in the three-border Trifinio zone, has cooperative trafficking routes used by organized crime. In Guatemala's tropical northern border, FG officials say that Agricultural Ministry staff ship illegal timber to the US in trucks that return with firearms-an example of armed arena synergy. On its southern frontier, Victor Manuel Díaz Morales, aka "El Rojo," trafficked drugs and arms for at least thirteen years across land, water, and air borders until his 2017 arrest. Ports are also large portals of entry for contraband, in part because of chaotic administration. In Honduras, the Finance Ministry's Executive Import Office (Dirección Ejecutiva de Ingresos) officially ran the ports through 2016, then created the stricter Rent Administration Service (Servicio Administración de Rentas). The National Anti-Extortion Force (FNA) took control of customs in 2014, but was displaced by FUSINA and the Presidential Commission of Tributary Administration.
By air, land, and sea, arms enter in bulk, individually, and in parts. Blind spots riddle borders, with an estimated fifty on El Salvador's frontiers. Guatemala has seventeen official border entry points but ninetyseven informal ones, of which at least forty-five are known entry points for AR-15s, Galils, AK-47s, M-16s, submachine guns, and grenades. Mostly arriving in private vehicles, about half were purchased in Texas and 25 percent in California. Parts are especially difficult to control since they are frequently sent in other 10 According to the most updated polling data, in Honduras, 28.7 percent of have confidence in the police and in Guatemala 35.2 percent (Yagoub 2017). In El Salvador, where 53 percent have confidence in the police, 40.1 percent approve use of torture and 34.6 percent favor extrajudicial executions in fighting crime. imports like microwaves and auto parts and then easily reassembled in the region's burgeoning artisanal arms industry. With help from online tutorials, for example, it takes only about an hour to assemble an AR-15 (Ocaña 2014). In areas with a low state presence, such as the Honduran department (state) of Olancho, residential arms workshops are well integrated in the country's trafficking network. 11 FUSINA's directors describe relentless parallel flows of firearms, such as narcotraffickers bringing in AR-15s for sale to maras and private firms serving as covers for light-arms trafficking. 12 They add that when they tighten controls in the main port, Puerto Cortés, arms are easily rerouted to other ports like La Ceiba. The long border with Nicaragua, whose population includes indigenous groups distrustful of both governments, is also crisscrossed with overland and sea routes. In the north, organized crime groups buy land on each side of the border with Guatemala, thus freeing themselves from all state control. According to ATIC's director, 13 large amounts of weapons are brought in by cartels from Mexico and Colombia, including AK-47s from Nicaragua that sell for around $250-cheaper than the $2,800 online or the $1,200 price tag in northern Mexico (McCarthy 2017). About one hundred thousand arms are believed to have entered Honduras between 2011 and 2013. Honduras has seen a sharp rise in legal imports, with value rising from US$5.7 million in 2000 to $16.6 by 2011 (Vargas 2014). Many arrive in postal packages, courier services, airport luggage, and noncommercial deliveries-a range of legal supply lines for illicit means. Numbers since then have been unavailable, due to Resolution 093-2012 of the Institute of Access to Public Information (Instituto de Acceso a la Información Pública, IAIP). It authorized the Instituto de Previsión Militar, the armed forces social welfare agency that runs the Armería and gets part of its funding from its sales, to withhold documents on the acquisition of weapons, arms supplies, and ammunition.
Nearby, the US and Panama constantly replenish the armed arena. Half of arms in El Salvador 14 and 35 percent in Guatemala are US-sourced, many through straw purchasing and the "ant" trade in which individuals buy arms at arms shows and through other loopholes in US law. Many enter through Panama's freetrade ports, in a pattern that accelerated under the 1983-1989 rule of Manual Noriega, as part of his support for armed groups in Central America and Colombia. As Colombia's conflict wound down, this arms flow then shifted to the Northern Triangle. With its huge global traffic and corrupt maritime licensing, Panama has been an ideal testing ground for arms traffickers in the use of methods pioneered by drug traffickers and sanctions busters in Iran and North Korea. In addition to hiding weapons inside legitimate goods and in sealed shipping containers, those methods include using circuitous routes to elude surveillance technology. Such methods take advantage of the fact that shipowners and customs officers rarely look beyond cargo documents to verify contents (Griffiths and Jenks 2012). In 2015, Panama lifted its official ban on arms imports in order to promote personal safety amid a rise in crime-further boosting the regional market.

Business
As discussed above, societal fear generates demands for private arms. In Honduras, rates of licit and illicit gun ownership per 100 people jumped from 9.9 in 2012 to 14.1 in 2017; in El Salvador from 9.7 to 12; in Guatemala from 11.5 to 12.1 (Karp 2018). Monitoring and documentation of sales is lax, weakened by unclear regulations, state contracts that obscure supplier channels, irregular physical inspections, exemptions or relaxation of import documentation, and large numbers of shell businesses. Such conditions then moot basic regulations such as requirements in all three countries of certification of the intended end user of any small arms or ammunition transfer.
Guatemala's constitution prohibits limits on arms possession, and arms importers only need a license from the arms registry, the military's Dirección General de Control de Armas y Municiones (DIGECAM) (Arms Law article 32) (see also ATF 2010, 11). In Honduras, only the state-run Armería is authorized to import and sell weapons, which it does at outlets around the country. Oversight is a notoriously weak; buyers can purchase weapons without showing a license or proof of training. The selection is wide. One San Pedro Sula branch on October 3, 2016, offered ten brands of pistols along with many rifles and automatic weapons. A .38 caliber costs around $500 and a 9 mm around $800-both on the lower end of global market prices (based on websites ClicAds; GunsInternational.com; Anuncioo). The biggest seller is the 9 mm ($45-$350 on the black market and about $900 at the Armería), which make up 63 percent of pistols and used in approximately 40 percent of homicides (Díaz 2013). The military also imports weapons through the Armería, as well as through channels outside the Defense Ministry. A mutual interest in bolstering sales among the Instituto de Previsión Militar, military, and the Armería is an example of cooperation in the armed arena. Of the country's eighteen departments, per capita sales are highest in Olancho, Colón, and Choluteca; not coincidentally, Olancho and Colón have two of the highest seven departmental rates of firearms homicides. In Guatemala, per capita sales are highest in Petén, which has the sixth highest homicide rate among the country's twentytwo departments. El Salvador's Defense Ministry reports that the twenty-five authorized importers brought in 125,703 weapons between 2006(Rosales Martel 2016, and the money Salvadorans spent on arms rose from US$1.2 million in 2009 to just under US$2 million in 2014 (Ministerio de Hacienda). Over 40 percent of capital city residents wants arms for self-defense, even though people using arms to defend themselves were four times more likely to die than those who did not.
The armed forces are another channel of weapons that make their way into organized crime networks. In each Northern Triangle country, military and police officials sell prohibited arms such as grenades, plastic explosives, and M-16s. In Honduras, a special elite military body sold weapons to organized crime networks, while police report as "lost" the arms that they sell. According to police chiefs overseeing the northern regions of Honduras, munitions trafficking forms a strong link between military officials and gang leaders.
But the armed arena's biggest arms conduit is the private sector, like the 700 registered and over 300 unregistered firms in Honduras, which together employ at least 70,000 guards. Guatemala has 1.5 private security officials for every public officer, and 30 percent of private firms in the capital are illegal. 15 In El Salvador, up to 40 percent of illegal weapons are connected to an estimated 460 private security firms (Wolff 2015, 10), many of which are not registered and which do not report arms sales or cooperate in robberies of their merchandise, according to police officials. Since at least 2000, many of those firms have been owned and run by current and retired Ministry of Defense officials. Nationally, businesses fork out $140 million annually for private security, along with US$30 million to the gangs, with some overlap (Economist 2016). In 2015, ninety persons connected with a private security firm in San Salvador were arrested as part of a gang extortion scheme (Chávez 2015). Another arms trafficking network in San Salvador involved small businesses, informal vendors, and both MS-13 and 18th Street.

Licensing and Registration
The primary way in which the state tries to control the arms market is licensing and registration. But verifications of licenses and inspections of businesses are weak and uncoordinated in all three countries. The combination of registration's weakness and importance, in fact, has made it the focal point of recent reforms, as discussed below. El Salvador's 1999 Ley de Control y Regulación de Armas, Municiones, Explosivos y Artículos Similares places no limits on individual possession. Of the four tests for registration, only the theoretical one is required, while the medical and psychological tests are applied haphazardly. The fourth test is of the gun's use, but officials say it is rarely given. Laxity by the Security Ministry's Registro de Armas de Fuego, according to one of its former officials, has caused over 70 percent of arms registration to expire. 16 The Defense Ministry's Dirección de Logística, which oversees civilian-held arms, reported that of the 22,000 weapons sent by the police to be investigated, just 64 percent were registered. In Honduras, registration requires only basic documentation like ID and proof of ballistic training. Its National Arms Registry (Registro Nacional de Armas) was established in 2004 as part of the Dirección Nacional de Investigación Criminal (DNIC), which collapsed as hundreds of its agents were purged for corruption. Guatemala's DIGECAM is a military agency but lacks the capacity to fully implement fines or license suspensions. Of the approximately 547,000 weapons registered at DIGECAM, many of them are clustered in the border areas. Its database includes thirty-three individuals with over 50 arms, including one family that earns 3,000 quetzales (just under US$400) a month but owns 130 firearms. 17 As in other countries, munitions are even harder to control. DIGECAM registered the sale of 13.5 million bullets just between January and June 2017. Though the main arms law, the 2009 Ley de Armas y Municiones, limits monthly possession of munitions to 200-250, there is virtually no control of them. Many of the exchanges occur in shooting ranges, which are required to record them and are checked twice yearly by DIGECAM. But there is no control of, or reliable data on the number of bullets used, the users, or the use of bullets purchased. The lack of markings on munitions further deprives investigators of criminal evidence.
An aggravating undercurrent of such weak control is of laws and administrative structures of registration and licensing that are outdated, disorganized, and limited. Even basic regulatory classifications are unclear, often failing to adequately define basic categories such as "rifle," armor, or "arms of war." In El Salvador, the Congressional Security Commission "protects the police" and the Legislature overall "can't agree on anything" 18 -as seen by a lack of action since a 2004 penal code prohibition on arms circulation. In Guatemala, the Comisión Nacional Multidisciplinaria en Materia de Armas y Municiones (MDN) advocates legislation to clarify arms classification, registration periods, penal processes and arms marcation. 19 But its members lack confidence that their weak president or divided Congress will enact change. In fact, it took ten years to approve the 2009 law.
In Honduras, similarly, the Congress has been promising reform to the three arms laws since 2007: the 2003 Ley Especial para el Decomiso de las Armas AK-47 y Otras no Permitidas, which limits ownership of semiautomatic revolvers and pistols (gauges 45 and 11.5); the 2004 Ley de Tenencia y Control de Armas, which punishes possession of "arms of war" like the AK-47 with up to six years imprisonment; and the 2000 Ley de Portación de Armas Municiones, Explosivas y Otros Similares, which limits the number of registered arms to five per person. These laws do little to control purchases: they include no rule for good conduct, do not require the Armería to carry out criminal or psychological background checks for buyers, and allows sales to ill-defined "juridical" as well as "natural" persons. Promises have long been made to tighten these laws, particularly to reduce personal possession limits and increase penal sanctions. But conflicting priorities and interests, especially over the control of arms agencies, sank proposed bills in 2008, 2009, 2010, 2011, 2013, and 2014. 20 Arms control has also been caught up in the clash between harsh and preventative policing that characterizes Latin America security policy. In El Salvador, the 2003 Plan Mano Dura and 2004 Plan Super Mano Dura were followed by 2004's Plan Mano Amiga and Plan Mano Extendida. An even bigger clash came when the 2012-2014 truce gave the maras the time and opportunity to acquire more and heavier weapons, while the whiplash to crackdowns in the truce's wake led to the arrests of over 12,000 pandilleros (but not to a drop in murders). In the 2012-2013 arms-destruction campaign accompanying the truce, only around six hundred weapons were turned in, and most were nonfunctional. During that same time, say officials, the police probably sold a much larger number of weapons back to nonstate groups-an indication of how links between state and nonstate groups can be immune to policy shifts. Table 2 summarizes the weaknesses in these three dimensions of the arms market, of which the armed arena takes full advantage.

Policy and Legislation: Change through Coordination
All these intertwining relationships in the armed arena make reform difficult. But since those relationships constantly shift, they also open up opportunities. Targeting the relationships among the arena's actors is a starting point for politically and administratively feasible change. This section outlines how officials are taking that approach and presents examples of how coordination among state actors is enabling them to weaken and shrink the armed arena.

Criminal Justice
Better coordination among criminal justice agencies has shown effectiveness against arms trafficking. Promising new intelligence agencies, such as Guatemala's Departamento de Investigaciones Criminalísticas, have helped curb criminal violence. But on arms, those efforts will stall unless backed up by consistent enforcement on illegal possession and use, as specified in Penal Code articles 141-143 and articles 110-121 18 Author's interview with Janette Aguilar, Public Opinion Institute director, San Salvador, January 22, 2016. 19 MDN meeting, April 28, 2016 Author's interviews with Leyla Diaz, Tegucigalpa, April 18, 2017, andMario Pérez, Security Commission president, 2014;El Heraldo 2014b.  In Honduras, the entire police structure was shifted to focus on investigations by ATIC and the DPI; among the DPI's first operations-using mobile laboratories-was against 18th Street arms trafficking. Intelligence agencies are finalizing a manual to coordinate criminal investigation among the many agencies involved in it, while the police are adopting a new Fiscal Management System (Sistema de Gestión Fiscal) to manage police reports. Integration of arms information into this database, though, requires a clearer set of protocols for storage, handling, and documentation of seized arms to fully utilize technology like Automated Fingerprinter ID System (AFIS), in conjunction with IBIS. Such analysis is bringing improvements in El Salvador, whose Policía Nacional Civil Statistics Division monitors 125 variables on firearms-up from 25-such as categorization, model, serial number, and investigative status.
Because the lack of updated and comprehensive arms data prevents officials from uncovering connections in crime cases or broader trafficking patterns, efforts are being taken to address it. For example, Guatemala's DIGECAM is sharing more information with civilian agencies, in part through the MDN, formed by the Defense Ministry to improve collaboration with the National Police, the FG, and INACIF. The MDN has advanced initiatives like a new Information Platform and a Center for Arms and Munitions Control (Centro de Control de Armas y Municiones), formed in 2012 and bolstered in 2016 with a National Strategy of Control Arms, Munitions, and Explosives centered on coordination among criminal justice agencies. In Honduras, criminal data is now directed by the new Police Statistic System (Sistema Estadístico Policial en Línea, SEPOL) which assesses it in conjunction with the FG, police, and other agencies. 22 But what is not analyzed are connections among the crimes it reports. And according to its own reports, there were 2,048 "undetermined" deaths between 2013 and 2015.
Criminal justice action against arms has also been boosted by the growing realization of the need for greater community involvement in order to elicit critical information, such as local arms sellers or trafficking routes. That approach is centered on community-oriented policing (COP), a shift in the security paradigm throughout Latin America that integrates citizens into policy through committees, meetings, and programs. COP is one of the longest-running reforms in the Northern Triangle, adopted since the late 1990s in El Salvador and Guatemala and since 2003 in Honduras. As shown by many studies (Arias 2019), such action can bring results, even in the highest-crime areas. But its application has usually been haphazard, applied as more complementary than integral to daily policing and dismissed by officers as "social work." Half-hearted COP is then easily co-opted by neighborhood cliques, vigilantes, or even organized crime groups.
But the security crisis has pushed governments to more fully adopt this approach, led most recently by Honduras. With Inter-American Development Bank funding, its new police academy, the Instituto Técnico Policial, is based entirely on COP, 23 while all serving officers are being retrained under a Modelo Nacional de Servicio de Policía Comunitaria. National COP has also doubled the country's thirty violence observatories, which provide invaluable information on local weapons sales. 24 COP programs have shown similar results in Guatemala and El Salvador, with approaches taken up by anti-arms programs. For example, in El Salvador's Proyecto Municipio Libre de Armas, gun-free zones created by Article 62A of the 1999 law have reduced violence as the information they gather is relayed to officials through the Plataforma Virtual. Amid decentralization, municipalities are also adopting COP. The Honduran city of San Pedro Sula credits it for taming an urban homicide rate that was long the world's highest. In its most violent colonia, Rivera Hernandez, the police chief has backed effective church-based peace-building programs. 25 Another Honduran city, Choloma, has reduced crime by requiring disparate agencies, from education to transport, to work together. 26 In El Salvador, the fifty-municipality Programa Prevenir is centered on school-based prevention and employment, with intersectorial cooperation and channels to relay information to the police. The 2014 Plan El Salvador Seguro, which has brought COP to 150 municipalities, includes campaigns for the voluntary surrender of weapons.

The Arms Market
Cooperative strategies among the Northern Triangle countries, beyond steps like the 2016 formation of the Elite Security Group, can close gaps in borders and business that fuel the arms market. One strategy is on markings. Although a unique identifying mark on each firearm in required in all three countries, they are not consistently verified, while insufficient markings are regarded as the weakest point by border and customs agents. Marking technology should thus be adopted through the OAS marcation program (Promoviendo el Marcaje de Armas de Fuego en América Latina y el Caribe), backed up by biometric identification systems used to identify gun ownership in countries like Peru. Through its Central America Regional Security Initiative, the US has provided key equipment, particularly IBIS machines and training by the Bureau of Alcohol, Tobacco, Firearms and Explosives (ATF) on the use of eTrace, a program that tracks the commercial trails of firearms through the National Tracing Center and the National Instant Criminal Background Check System, which runs arms serial numbers. Even though ATF stations just one arms specialist in the region, in 2014 over 8,000 guns seized Central America, nearly 3,300 of which originated in the US, were traced by the ATF (Meyer and Seelke 2015, 27).
Most significantly, years of documentation and pressure by NGOs and their elected allies can lead to breakthroughs. In 2017, Guatemala reformed the Ley de Armas y Municiones to strengthen licensing through creation of a digital and physical databank of ballistic fingerprints, into which all arms must be registered by 2020. In September 2018, more ambitiously, Honduras's legislature passed a new arms law whose ten main provisions address many of the gaps and weaknesses this article highlights: its establishes registration offices outside the two main cities; widens IBIS capacity; tightens Armería sales, such as by requiring presale ballistic tests of all weapons; strengthens regulations on munitions purchases and public carry; transfers all seized arms directly to the forensics office; requires registration of all arms held by security officials; creates new categories of permits based on intent, such as for sport or private security; increasing the age to twentyone for possession; and reducing the individual possession from five to three. As documented in several studies (Freire 2018), gun control can help bring down crime. Officials and activists credit the legislation in part to the existence of eight different parties in Congress-rather than the traditional two that long thwarted reform-and tenacious efforts by influential legislative groups like the Congressional Commission on Security and Citizen Prevention, which spearheaded a 2014 bill to restrict weapons carrying (El Heraldo 2014a) that was shot down by the private security sector. At the same time, some fear that the 2018 reform is unrealistic, since the state lacks the institutional means to adopt it.

Broader Implications
This article develops the concept of an armed arena, comprised of state and nonstate entities, which can stretch a country's policies and laws beyond capacity. Although it is particularly susceptible to such impacts, the Northern Triangle holds key lessons for other regions. Connecting the article's introduction and empirical work, this section reinforces two wider applications of the model's relationships among security policy, policing, and crime within the larger context of state weakness.
The article's introductory discussion of scholarship on democracy and the state, first, stresses how their weaknesses obstruct needed security reform. The conceptualization of the armed arena deepens and adds to the literature's discussion of that challenge. On one level, the lack of physical or institutional controls over the influx of arms increases the power and autonomy of state security forces, in turn hurting efforts to make them more accountable. In the charged political debates over crime, arms marketing amid societal fear fuels incendiary political rhetoric, tipping the balance away from COP and nonviolent policies (which often involve disarmament) to citizen self-defense and the mano dura's repressive armed tactics. From Mexico to Argentina, security reform has been continually hindered by this policy shift, which fortifies bonds in the armed arena among state and nonstate entities-seen most recently in the vigorously renewed support by the government in Brazil of the urban militias, often directed by active and retired security officials. In every country, elected officials' normative preferences and short-term political calculations, as Mainwaring and Pérez-Liñán (2014) discuss, further push policy toward the mano dura. Such patterns feed into what Bailey (2014) characterizes as a "security trap" in which state corruption and weakness aggravate conditions that impede reform. Altogether, they also show that nothing lays bare more starkly than do arms the ongoing security crisis of contemporary Latin America.
Underscoring this reality is a second dimension of the article's applicability. As discussed above, the armed arena reflects the boost to organized crime by globalization and states' susceptibility. In particular, as discussed in works on drug and human trafficking (Naylor 2004;Zhang 2014), organized crime operations have become more fluid and transactional, with temporary contractors filling roles in flexible networks whose increasing adaptability facilitates infiltration of political institutions and formation of "grey" markets