“Collaborator” is a loaded term. It’s used by insurgents to condemn those who work with the state, and counterinsurgents avoid it in favor of ally, patriotic citizen, or local partner. Regardless of rhetoric, collaboration is an integral process of civil conflict: information is hugely valuable and so local informers are essential to the war effort of the state and its foes. Collaborators, and those accused of being collaborators, of both sides are liable to be found dead in ditches.
There are many reasons for ordinary people to become collaborators – blackmail, settling scores and rivalries, political commitments, contingent events, economic gain, etc (for human faces of collaboration in Vichy France, watch The Sorrow and the Pity). But what I want to focus on here is a specific type of collaboration with the state, by former insurgents. These are insurgent “ethnic defectors” who jump ship and become an arm of the counterinsurgency apparatus waging war against their former comrades. These are a particularly useful type of collaborator because of their detailed knowledge of the insurgency.
This type of defection is a central goal of American policy in Afghanistan. Flipping parts of the Taliban and/or tacitly allying with Gulbuddin Hekmatyar’s Hezb-e Islami have been broached as possible ways out of the American predicament. But defection is difficult and rare – actively and publicly turning on former comrades is a dangerous choice, even more so than dropping out of the fight, surreptitiously informing, or cutting a de facto separate peace. “Selling out,” with lethal consequences for all involved, is not a decision taken lightly.
Given these challenges, what do we know about why it happens?
There are at least three non-mutually-exclusive mechanisms that seem to be at play. The first are government policies aimed at attracting “moderates” to its side. By offering political programs satisfying the war aims of particular factions, the goal is to exploit internal ideological differences to create splits (see related work by Stedman, Bueno de Mesquita, and Kydd and Walter). One problem with deploying this mechanism in practice is that it can be awfully hard to tell ex ante who the moderates are. Insurgents willing to cut a deal, for instance, are glowingly lauded as moderates by peace processers after the fact, but before the fact they may have been assessed by observers assessed as hard-line militants (a la Brian Keenan, Slab Murphy, and Martin McGuinness in Northern Ireland)
A second mechanism is some variant of state divide-and-conquer. Counterinsurgents exploit personal rivalries, blackmail and extortion, informants, revenge, personal gain, bribery, and misinformation to sow mistrust and division within the insurgent milieu (see work by Johnston on Sudan, Pearlman on the Palestinians, Kalyvas on Greek counterinsurgent militias). The difference with the first mechanism is that this does not necessarily involve big political concessions, but instead a nitty-gritty, unpleasant exploitation of internal difference.
Third, in some preliminary recent work I’ve been using historical cases and fieldwork to trace out a third mechanism that focuses less on state policy and more on the internal politics of insurgent movements. Serious, lethal fratricide between insurgent factions, whether over politics, cash, drugs, or personal prestige, can create a point of no return beyond which insurgent rivals become open to lethally allying with the state against their enemies in the insurgency. Internal violence makes defection more likely, since the stakes are higher, other choices are restricted, and the powerful emotions of betrayal, anger, and fear integral to feuding create a willingness to take the final step. States can then exploit these internal contradictions, but their policies tend to be reactive rather than the causal trigger for defection.
Ultimately, while I think the state can be hugely important, it’s crucial to remember two things. First, insurgents are political actors too, with their own political interests and rivalries. The state’s policies are not a deus ex machina driving variation in outcomes. Second, states (especially foreign ones) are often pretty awful at clearly identifying the salient cleavages within insurgencies and the right incentives to appeal to. Insurgent movements are often opaque and fluid, while counterinsurgents tend to spend much of their time Seeing Like a State. Only after years of bloody trial and error can states start to accumulate the intelligence they need. As a result, it’s hard, though not impossible, to play clever games of divide-and-rule in any fine-grained way.
The implication, at least for me, in Afghanistan is that the US probably doesn’t have a reliable sense of exactly what the structure of incentives, divisions, and loyalties is, and so expecting ISAF to be able to easily split off defectors or manipulate the Taliban is unrealistic. It’s more likely the Karzai regime has this knowledge, but it appears that Karzai is playing a different game than the US. Instead, the goal should be to be prepared to exploit internal strife as it appears, with guns, money, and protection; if feuds between local Hezb-e Islami and Taliban units in Afghanistan escalate to a broader inter-organizational clash, Americans and Afghans bearing gifts should be ready to parley with Gulbuddin Hekymatyar, or whichever side most needs help. This may involve extremely unlikely bedfellows and morally problematic alliances. As Giustozzi has noted in his study of warlordism in Afghanistan, many of the defectors to the Kabul regimes of 1979-1996 jumped ship because of blocked promotion paths, feuds, and rivalries within the insurgency, not just manipulation by the regime or sudden ideological conversions.
This is all well and good as an exercise in armchair counterinsurgency. But encouraging and exploiting the rise of unofficial pro-state militias can create severe problems of political order and basic human decency, since they often engage in horrific rights abuses and escape accountability. “India’s Secret Army” in Kashmir, the defectors who joined with the Indian security forces in Kashmir in the mid-1990s, are still referred to with a shudder; Tamil pro-state paramilitaries have been installed as quasi-warlords in parts of Sri Lanka. There are long-run consequences of defection and collaboration that remind us how bloody, nasty, and morally compromising counterinsurgency can be.