Faculty of Law, University of Hong Kong
Center for Law, Society and Justice, University of Wisconsin-Madison
Constitutional courts can play an important role in protecting minority rights and providing a forum for the non-violent resolution of constitutional disputes. Arguably, this role is especially vital in post-conflict and deeply divided societies. But if the politics that divides the society more generally also influences judges, a court’s ability to play this role – at least impartially and independently – will be compromised. Our recent LSR article, ‘Judicial Impartiality and Independence in Divided Societies’, is the first ever published study to rigorously consider the extent to which ethno-national politics influences judicial decision-making on constitutional courts.
The article focuses on the Constitutional Court of Bosnia-Herzegovina. Bosnia-Herzegovina is a post-conflict society, still grappling with the legacy of the war and ethnic cleansing that ravaged the country in the 1990s. Following the peace agreement that ended the war, a new constitution divided and decentralized power in the country, creating two territorial ‘entities’ and a power-sharing system to accommodate Bosniak, Croat, and Serb factions (the country’s three main ethno-national communities). The Constitutional Court is similarly designed to include two judges from each of these three groups, alongside three ‘foreign’ judges who are there – at least in theory – to provide some impartial expertise. For most of the Court’s existence, the judges have held long-term mandates (with mandatory retirement at age 70). For the first 5 years, they were appointed for short and non-renewable mandates.
The judges have been accused of dividing along ethno-national lines and favouring claims brought by their own group (or, alternatively, displaying bias against claims brought by other groups). Our study examines if there is an objective basis for these impressions. With an original dataset of the Court’s decisions, and using a statistical technique called logistic regression, we examine the extent to which the probability of a judge finding in favour of a constitutional challenger significantly increases when both the judge and the challenger are from the same ethno-national community. Accounting for other possible influences, including potential party-political connections between the judge and challenger, we find evidence of a dramatic effect: the judges are over four times more likely to find a constitutional violation when the challenge is brought by a member of their own group.
We also consider if the Court’s move from a short-term tenure model to a long-term tenure model had any effect on the judges’ tendency to side with co-ethnic challengers. We find that it did not; there is no significant difference between the short-term and long-term appointed judges in this respect. We argue that this result suggests that the judges’ propensity to favour co-ethnic challengers mostly reflects genuine political and/or group biases, as opposed to a strategic motivation to appease powerful politicians. That being said, we do find some evidence that the length of time a judge serves on the Court makes some difference to this tendency: as time goes on, the judges appear to become more ethnically partial. This effect may reflect a process whereby judges ‘acclimate’ to their role on a divided court over time (evidence for a similar effect has been found on the US Supreme Court).
Our article concludes by suggesting some ways in which constitutional courts might be designed to reduce the potentially damaging influence of ethnic or ethno-national politics on judicial behavior. Against the conventional wisdom, we argue that short and non-renewable mandates might be preferable to long-term tenure, at least to the extent that they reduce the potential for judges to become ‘acclimated’ to dividing along ethno-national lines. We also suggest that the practice of allowing dissenting opinions should probably be avoided in a deeply divided context like Bosnia-Herzegovina; though dissents give judges a way to express themselves (and facilitate the kind of empirical research we conducted for this study), they also draw attention to political divisions and, consequently, undermine a court’s ability to speak with a single authoritative voice.