Today’s Conflict Consortium Virtual Workshop paper, “How does Human Rights Law Work? Institutions, Norms and Focal Factors” by Tiberiu Dragu & Yonatan Lupu, suggest a third mechanism by which international human rights law reduces violations. The first two mechanisms–sanctions (aka the logic of consequences) and norms (aka the logic of appropriateness)–are well known and documented, though a number of recent widely publicized books have challenged their impact. This paper points us to an overlooked third mechanism: the logic of expectations.
The idea is straight forward to anyone familiar with coordination games involving lots of players. Lots of potential solutions exist, but the players have difficulty to communicating which solution they should pool on. I was reminded of the discussion during the 2004 Democratic primaries for US president in which many voters said that the liked a candidate other than John Kerry better, but planned to vote for Kerry because he was the only one who could win.
Why is that a coordination problem? Because rather than simply act on their personal preference these voters expressed concern for wanting to pool their votes with those of other voters (mistakenly) believing that doing so was important for beating the Republican nominee that November. A more common example of coordination games involves which side of the street to drive cars on: either the left of the right side would work. What is critical is that everyone adopt the same standard, thereby dramatically reducing collisions.
Dragu & Lupu argue that this logic applies to agents of coercion in situations where the executive has to call on large numbers of them to put down a popular uprising. Consider, for example, countries like Algeria, Bahrain, Egypt, Morocco, and Tunisia during 2011. All of those regimes faced large scale popular protests, and each government called upon police and/or military to play a role in policing/suppressing the demonstrations.
Dragu & Lupu suggest that the agents of repression face a coordination problem in such situations: if all other personnel repress, than each agent wants to join his/her comrades. However, if they are going to balk (as the military did in Egypt), then each individual soldier does not want to repress. The difficulty in such situations is that repressing is a better choice if the leader is going to survive (and punish anyone who sat out) but not repressing is a better choice if the leader falls (and the future regime or a court takes action against those who committed violations). Yet the success of the executive depends on how the agents of coercion respond.
That’s where international human rights law, and perhaps the lobbying efforts of watchdogs and activists, can play a role. Not by persuading soldiers that they will be punished (logic of consequences) or that it is wrong to repress (logic of appropriateness), but instead that their comrades in arms won’t repress. The third mechanism by which international law and activism can reduce violations is by providing information that suggests a focal point solution to a coordination problem among large numbers of actors who cannot readily communicate with one another.
Ann Marie Clark, Amanda Murdie, Keith Schnakenberg and Ryan M. Welch joined me to provide Lupu feedback. It was a fun and productive session. But of particular interest, I had asked Scott Edwards if he might “pinch hit” for Christian Davenport, my co-convener who could not make it due to a scheduling SNAFU.
Edwards, who has worked with Amnesty International USA for a number of years, was unable to join us, but he read the paper and caught part of the discussion on YouTube. Afterwards he fired some comments to Dragu, Lupu and I that are quite interesting.
I was a fan of the model because it formalizes a causal path that I’ve witnessed (and directed) in minimizing compliance with illegal orders; which is essentially to create doubt among rank and file about whether compliance or non-compliance will be the norm…we’ve done this through big, highly visible media presence, with talking points that stress that there will be costs for commanding officers, inter alia.Importantly, I don’t have enough fingers to count the number of times I’ve gone on AJ or BBC, with my intended audience being security forces and combatants, rather than general viewers. Intent was to create the very doubt/belief that is the causal nexus in the model.
If it sounds interesting, you can watch it on YouTube here.