How do institutions affect the degree to which parties act in unison? Using both the adoption and reversal of the direct elections to the prime minister (which altered executive-legislative relations) as well as divergent selection mechanisms characterizing Israel during the last three decades and utilizing three party-level measures of behavior–Rice Scores, the standard deviation of a party’s Knesset Members’ ideal points, and the mean distance of Knesset Members from the median position of their parties–I set to test whether regime type, electoral systems and selection procedures affect party behavior. I find that during the electoral reform period parties exhibited lower levels of unity compared to parties in the pre and post reform periods. However, the analysis fails to provide support for the effect of permissive selection processes on lowering unity levels. To explain these results I theorize about the conditional effect of elections and selections on parties’ behavior. Testing it using cross-temporal variation in selection processes yield sup- port for the conditional logic
