Does direct democracy foster efficient policies? An experimental investigation of costly initiatives

This study investigates the effects of the provision of costly initiatives on policy efficiency in a laboratory experiment where a policy setter implements a status quo affecting the utility of the constituency. I vary treatments regarding the political institution (either purely representative or direct democracy where the status quo may be contested by the costly proposal of an alternative) and the appointment of the policy setter (either random or by election). In accordance to theoretical predictions, the experimental data reveal a substantial indirect effect of direct democracy inducing higher efficiency levels by serving as a credible threat towards the policy setter without actually being used. Moreover, the initiative impedes excessive candidate competition during elections reducing campaign costs and thus increasing overall efficiency. In contrast to theoretical predictions, the initiative is actually employed frequently, so there is also a sizeable direct effect of the initiative. However, this effect is generally overcompensated by the costs induced by the process.