In a multiperiod setting, decision-makers can learn about the consequences of their decisions through experimentation. In this paper we examine how in a two-party system polarization and political instability affect learning through experimentation. We distinguish two cases: the decision to be made is not salient and does not affect the outcome of the following elections (exogenous elections) and the decision is salient and the election outcome depends on it (endogenous elections). We show that while the possibility of learning increases activism, the existence of political instability distorts learning. Furthermore, in contrast to the existing literature, we demonstrate that, when elections are exogenous, polarization between political parties does not always decrease active learning. In the case with endogenous elections we find that electoral concerns may induce candidates not to experiment, even if the majority of voters prefers activism.

, ,
,
hdl.handle.net/1765/6870
Tinbergen Institute Discussion Paper Series
Tinbergen Institute

Ossokina, I., & Swank, O. (2001). How Polarization and Political Instability affect Learning through Experimentation (No. TI 01-040/1). Tinbergen Institute Discussion Paper Series. Retrieved from http://hdl.handle.net/1765/6870