James Hollyer, a Leitner Fellow at Yale University, sends along the following response to Kathryn Sikkink’s New York Times op-ed from this weekend:
I read, with considerable interest, Kathryn Sikkink’s op-ed regarding the positive consequences of human rights prosecutions for nascent democracies. She argues that prosecutions in Latin America and elsewhere lead to lower subsequent human rights abuses. This piece drew my attention, in part, because of the contrast between her findings and my own work – joint with Peter Rosendorff – which was featured in The Monkey Cage some time ago.
Peter and I argue that the threat of prosecution for human rights violations creates a perverse reaction in autocracies. Autocratic leaders are unlikely to face prosecution for human rights violations so long as they remain in office. Leaders who know they will only relinquish power in a body bag – either because of natural causes or as the result of violent opposition – face few repercussions from increasing probability of prosecution, or the severity of punishments for human rights violations, should they be evicted from office. Those more willing to depart from power, on the other hand, face potentially dire consequences from doing so.
This perverse logic has important implications for international commitments to hold leaders accountable for human rights violations. Those autocratic leaders most likely to make such commitments are the ‘tough’ types most willing to cling to office; while those more willing to negotiate with their domestic opponents will not be willing to run such a risk. Since ‘tough’ types are more willing to employ repression to hold onto office, autocratic regimes prone to violate human rights will actually be more willing to enter into international human rights commitments – and this tendency will become stronger as the threat of prosecution rises.
We test this view using the accession of states to human rights treaties. Some of these treaties mandate prosecution for human rights abuses. We show that it is the more repressive, hard-line autocrats that sign more of the more punitive treaties than the weaker, insecure autocracies. Moreover, such leaders are likely to survive longer in office than those that do not sign such treaties. This holds for three reasons: (1) It is the relatively ‘bad-ass’ leaders who sign these treaties (self-selection), (2) the threat of prosecution makes relinquishing office more costly at the margin (commitment), and (3) in subjecting themselves to prosecution in the event of removal, autocratic leaders credibly signal their determination to hold office, deterring the opposition (information). These three effects influence the decision to repress while in office: The first two work to increase the amounts of current repression; the third could act to reduce current repression – because the opposition learning that it faces a tough autocrat lessens its efforts to unseat the current rulers. The net effect could increase or decrease the amount of repression undertaken by autocrats while in office. The three effects however work in the same direction when it comes to survival in office: acceding to human rights treaties enhances the survival in office of the worst, torturing autocrats, and reduces the chances of regime change.
Increasing the threat of punishment for human rights violations serves to strengthen these mechanisms. As a given institution increases the punishments for human rights violations, the pool of autocrats acceding to that institution will increasingly be dominated by repressive types. These types will survive for increasingly long periods in office. And the first two mechanisms will increasingly come to dominate the third, such that any positive effect of the human rights institution on reducing levels of repression will diminish.
This is not to cast doubt on Sikkink’s findings, which pertain to nascent democracies. Human rights commitments may well ‘lock in’ democratic politics in such countries, constraining leaders –who face a large risk of removal – from repressing opposition groups. Our arguments pertain only to autocracies.
But, these contrasting findings seemingly have policy implications that may be more nuanced than those widely emphasized in the human rights literature. The punishment of human rights violations – and, presumably, institutions that commit countries to pursue such punishments – have positive effects in nascent democracies. But, these same institutions have negative – or at least normatively questionable – effects when applied to autocracies. It would therefore seem that international human rights agreements are best implemented at the bilateral or regional level, where membership can be restricted to target countries that stand to benefit from these institutions the most; whereas, broad-based treaties raise substantial problems as a result of governments self-selecting into membership.