39
governments that have ratified both treaties. The solid black line represents predicted
probabilities, the dots are the coefficient estimates, and the dotted lines are their 95%
confidence intervals. We discover consistent effects. Among repressive states that have
ratified both core treaties designed to protect human beings from violations of personal
integrity, we cannot say with any degree of confidence that domestic institutions (what
we use here as a proxy for motivation, following Moravcsik), make any difference. “Bad”
states that have done “good” things are no more or less likely to reform when
governments rule under incoherent institutional arrangements than when they are
governed by either democratic or autocratic regimes.
36
Conclusion
We began with a simple but important political question: are repressive governments,
ruling “bad” states, that make commitments to international human rights laws, doing
“good” things, making good on their promises to defend the rights of people? We
appreciate that a growing number of concerned human rights advocates have begun to
study compliance dynamics more generally, trying to understand how legal norms
cascade globally and how international laws can be effective tools of reform. This article
participates in that larger process by using observations from the world to gain
understanding about the behaviors of repressive states that ratify two human rights
treaties in particular.
Our findings are both alarming and revealing. They are alarming because they
suggest real and systematic limitations to the human rights legal regime. We do not
36
We note with some interest, however, that the spread of the confidence intervals is
smaller for incoherent regimes.
Convention