Voting power and parliamentary defections : the 1953–54 French National Assembly revisited

We reconsider Riker’s [10] test of the hypothesis that inter-party migrations in the French National Assembly in 1953–54 can be explained by deputies’ tendency to maximize their a priori voting power. However, instead of the Shapley–Shubik index used by Riker, we use the [absolute] Banzhaf measure, which we argue is more appropriate for this task. The theoretical model we use differs in some key respects from that of Riker, due to the difference in the underlying notion as to the nature of voting power. Our conclusion, however, is broadly in agreement with Riker’s: the hypothesis under question is not substantiated.