makes a vote for the extreme right/radical right much more likely. There is, however, a potential problem with this argument: if radical right support is stable in the medium term, and if other parties react to past successes for the radical right by modifying their manifestos, this relationship might be spurious. In my paper for the ECPR conference at Potsdam, I use a time-series model to address this problem: I estimate a Vector Auto Regression (VAR) of radical right support and issue salience in France (while controlling for immigration and unemployment). As it turns out, salience is independent of previous radical right success. This finding provides some support for my original argument, though the analysis preliminary and restricted to France (at the moment).