The Imperfect but Real Effects of International Institutions on LGBT Rights in Europe

by Erik Voeten on August 1, 2013 · 3 comments

in Comparative Politics,Foreign Policy,Institutions,International Relations,Judicial

So far I have made two points in my mini-series on human rights: that international human rights institutions at best have modest effects in countries where the opportunity structure permits international influence and that I am not persuaded by the argument that human rights institutions would work better if they focused only on the “ideals that inspired the human rights movement in the first place.”

It makes sense to follow up with some evidence that an international human rights institution did successfully influence non-traditional rights, albeit in a limited way. The case that I will discuss is the impact European Court of Human Rights (ECtHR) rulings have had on LGBT rights in Europe. It is based on a paper (forthcoming in International Organization) that I wrote with Duke law professor Larry Helfer. With Europe we mean all 47 Council of Europe member states, which includes Russia, Turkey, Albania, Azerbaijan, Georgia, Serbia and many other countries where public acceptance of homosexuality is among the lowest in the world.

The document that the ECtHR is supposed to interpret, the European Convention on Human Rights, makes no mention of homosexuality or sexual orientation. Thus, states have not explicitly delegated the Court the authority to uphold LGBT rights. Yet, we find that ECtHR rulings that find a certain practice, such as criminalizing homosexuality, a violation of the Convention have a substantial effect on policy change. However, this effect is only manifest in countries where public support for homosexuality is low and where a government is in power that does not draw its primary support from rural, religious, or nationalist bases (see graph below).

In countries with high levels of public acceptance and an urban and non-religious government, policy change happens without international legal action. Rural, religious, and nationalist governments tend to resist liberalization regardless. Yet low public support but a government that is not necessarily ideologically opposed to liberalization creates an opportunity for an international intervention to make a difference. We estimate that a substantial number of countries, especially in Eastern and Southern Europe, have more liberal LGBT rights laws than we would have expected in the absence of international court action.

Presentation1
Let me back up a bit and explain how we got there and what all of this means in the context of the current backlash against LGBT rights in some of these countries, most notably Russia.

The ECtHR has grown to allow about 800 million citizens to file claims that their government has violated one of the rights protected by the European Convention on Human Rights. If the Court declares an application admissible, it can find that a government pay monetary compensation to a victim or take some other compensatory measure, such as reopen judicial proceedings or release an individual from prison.  Larry and I did not look at whether countries complied narrowly with such rulings. Instead, we studied whether they actually changed the laws or policies that led to the violation in the first place. More ambitiously: we did not just look at the country against whom the Court found a violation but at all countries in the Council of Europe system. So, we wanted to know whether after the Court found that the UK military could not discriminate based on sexual orientation Germany and other countries also changed their practices? After all, this is how a well-functioning legal system should work.

We found that the ECtHR system is far from perfect. Some countries still have discriminatory practices that the Court has found illegal a decade or more ago. Yet, we did find that an ECtHR ruling on average sped up policy change. We were able to identify this effect (imperfectly) because in all issue areas we looked at the Court had on multiple occasions found that a practice was not a violation of the Convention before it found that it was a violation. Sometimes these contradictory judgments were only a year apart. Yet, after the Court found a violation it never reversed. This obviously opens the Court up to the criticism that it is “inventing” rights. Nonetheless, we found that these judgments still mattered albeit selectively.

Our most notable finding is that the marginal effect of Court rulings is largest in countries with low public acceptance of homosexuality; suggesting that the institution matters most where it is needed most. Among the 25 countries with the lowest public acceptance of homosexuality 9 are in Europe (according to the World Values Survey). All of these have decriminalized homosexuality, mostly in the 1990s. Among the 16 other countries, only 4 do not have laws on the books that explicitly criminalize homosexuality.

There is little evidence that the legal changes in these European countries have noticeably changed public attitudes or that they have eliminated discrimination in society. Nonetheless it matters that you cannot be thrown in jail for consensual sex. Still, the existence of more liberal laws than a society might have supported creates the opportunity for backlash. I have not seen any evidence linking the backlash in Russia or elsewhere to overly liberal ECtHR rulings. Nevertheless, Russia’s propaganda law likely violate existing ECtHR jurisprudence and Moscow continues to ban gay pride parades in direct violation of a Court ruling. Ultimately, then, I view this as a case of real but also limited change resulting from international court action.

 

{ 3 comments }

rvman August 2, 2013 at 9:30 am

Just as a note, if you have probability as a dependent variable, and it isn’t strictly bound by zero and one, you are using the wrong model. Find a logit, logistic, or probit based way of modeling what you want to model.

Erik Voeten August 2, 2013 at 10:48 am

It is a logit model. I was just a bit sloppy with the axes and the display of the confidence intervals here.

rvman August 2, 2013 at 1:36 pm

Good, Thanks for clarifying that.

Comments on this entry are closed.

Previous post:

Next post: