ECPR

Install the app

Install this application on your home screen for quick and easy access when you’re on the go.

Just tap Share then “Add to Home Screen”

ECPR

Install the app

Install this application on your home screen for quick and easy access when you’re on the go.

Just tap Share then “Add to Home Screen”

Judicial Independence and Judge Selection: Evidence from Switzerland

Courts
Quantitative
Experimental Design
Judicialisation
Gabriel Gertsch
ETH Zurich

Abstract

To guarantee judicial independence, European countries typically grant long tenure or even life tenure for judges. Switzerland marks an exception, with short and renewable tenure and a practice requiring that judicial candidates be members of a political party. To study how this affects judicial independence, this paper collects a novel dataset (N = 9,653) covering the universe of social security and immigration decisions by the Swiss Federal Administrative Court (FAC). In many judicial set-ups, it is nearly impossible to disentangle the effect of case facts and of judges on case outcomes. To overcome this inferential challenge, the paper exploits a natural experiment at the FAC where cases are assigned to judges quasi- randomly. In addition, FAC judges have a known political party affiliation, providing a proxy for judges’ political ideology. These two institutional features are unique in Europe and provide an opportunity to investigate whether impartial verdicts require full separation of the judiciary from other state powers. The paper finds that judicial preferences vary substantially in social security law, but less so in immigration law. In neither legal area do preferences correlate with political ideology in expected ways. This contrasts with previous research on Swiss asylum judges, where political ideology strongly predicts judicial preferences. To investigate the mechanism behind these differences between legal areas, the paper explores partisan citation behavior of FAC judges and estimates the relative importance of the case facts versus the law by training a classification model which predicts outcomes based on the text of the verdict.