UC Berkeley

Why do drug cartels fight states? Episodes of armed conflict between drug cartels and states in Colombia, Brazil and Mexico have demonstrated that `criminal wars' can be just as destructive as civil wars. Yet insurgents in civil wars stand a reasonable chance of winning formal concessions of territory or outright victory. Why fight the state if, like drug cartels, you seek neither to topple nor secede from it?

Equally puzzling are the divergent effects of state crackdowns. Mexico's militarized crackdown in 2006 was intended to quickly break up the cartels and curtail incipient inter-cartel and anti-state violence; five years later, splintered cartels are an order of magnitude more violent, with over 16,000 homicides and 600 of attacks on army troops in 2011 alone. Conversely, in Rio de Janeiro, a massive November 2010 invasion by state forces of a key urban zone that had been under cartel dominion for a generation failed to produce the grisly bloodbath that even the government's defenders predicted. Instead, it heralded what appears to be a decisive shift by cartels away from confrontation. Why do some crackdowns lead to violent blowback, while others successfully curtail cartel-state conflict?

The key to both puzzles lies in a fundamental difference between cartel-state conflict and civil war. Cartels turn to anti-state violence, not, as in civil war, in hopes of conquering mutually prized territory or resources, but to influence state policy. Like many interest groups, cartels expend resources to influence policy, usually acting at the level of policy enforcement, through corruption, but sometimes also at the level of policy formation, through lobbying. Yet licit interest groups are not targeted for destruction by the state, and generally possess no means of physical coercion. Cartels always face some level of state repression, but fighting back usually provokes even greater repression. Often, this leads them to `hide' rather than `fight', using anonymity and bribes to minimize confrontation; under certain conditions, though, violence may seem the best pathway to policy influence. The decision to turn to violent forms of policy influence is thus highly sensitive to what the state is doing; shifts in state policy, especially crackdowns, can trigger sharp variation in cartel-state conflict.

This study first distinguishes the logics of violent corruption and violent lobbying, as well as dynamics deriving from turf war among cartels, then identifies the conditions that make each logic operative. Violent corruption---epitomized by drug lord Pablo Escobar's infamous phrase "plata o plomo?" (bribe or bullet?)---is central; it occurs, in all three cases, prior to and with greater consistency than violent lobbying or other mechanisms. States face a dilemma: they cannot crack down on traffickers without inadvertently giving corrupt enforcers (police, judges, etc.) additional leverage to extract bribes. A formal model of bribe negotiation illustrates the cartel's choice: simply pay the larger bribe, or use the threat of violence to intimidate enforcers and reduce the equilibrium bribe demand. The central finding is that blanket crackdowns in a context of widespread corruption can increase cartels' incentives to fight back, whereas more focused crackdowns that hinge on cartel behavior induce non-violent strategies.

Conditionality of repression--the degree to which repressive force is applied in proportion to the amount of violence used by cartels--is thus a critical factor behind the divergent response of cartels to crackdowns across cases. A move toward conditional crackdowns occurred both in Colombia, after Escobar's demise and the fragmenting of the drug market, and in Rio de Janeiro, with its innovative `pacification' strategy. In both cases, cartels have shifted away from confrontation and toward non-violent `hiding' strategies. In Mexico, by contrast, the state has insisted on pursuing all cartels without distinction, leading to sharp increases in cartel-state violence.

Other, less central logics help explain contrasting modalities of cartel violence. 'Violent lobbying', in the form of narco-terrorism and direct negotiation with state leaders, is dramatic and chilling, but only makes strategic sense when there is an open policy question that cartels can realistically hope to influence. Moreover, if the benefits of policy change are 'public' or non-excludable, violent lobbying is subject to the free-rider problem, and only likely to occur if cartels can cooperate. Thus violent lobbying has been intense in Colombia, where cartels were initially united and extradition remained an open policy question for a decade; occasional in Brazil where a dominant cartel uses it to influence carceral policy, and relatively rare in Mexico, where cartels are fragmented and the president's high-profile `ownership' of his crackdown creates overwhelming audience costs to policy change.

Inter-cartel turf war is far more intense in Mexico than elsewhere, driving logics of reputation-building and false-flag attacks, and contributing to the prominence of `propagandistic' violence like mutilation and `narco-messages'. These turf-war dynamics are reinforced by the government's kingpin strategy and its splintering of the cartels. Moreover, fragmentation has a general-equilibrium effect on the maximum pressure the state can apply to any one cartel, given its unconditional approach. This further reduces the sanction cartels face for using violence, and drives the escalatory spiral presently gripping Mexico.

The study concludes by asking why leaders do or do not adopt conditional strategies. Even when leaders would like to do so, they face both 'logistical constraints' arising from low capacity and fragmented security institutions, and 'acceptability constraints' deriving from the negative optics of `going easy' on less violent cartels (a necessary component of conditional repression). Case evidence helps identify political circumstances that minimize these constraints. Coalitions or partisan hegemony can mitigate institutional fragmentation, while the 'Nixon-Goes-to-China' effect allows leaders perceived as hardliners to overcome acceptability constraints, particularly if they present conditionality as a tactical, operational imperative.