The rules that are employed to pass policies in legislative bodies vary widely. It is generally argued that policies that differ in complexity or importance level should be decided under different kinds of voting rules. While this question has been examined for static legislative mechanisms, an analysis of the precise relationship between the level of policy complexity and the type of voting rule is still missing for dynamic mechanisms. We address this problem from the perspective of a preference-blind political designer. Given the level of complexity of the decision that is to be made, the political designer's goal is to select the supermajority rule that avoids (1) policy instability; (2) guarantees efficiency; and (3) minimizes institutional status quo bias. We provide an answer to this objective problem, deriving a closed-form relationship between voting rule and policy complexity. Our analysis rationalizes the use of different rules to adopt different types of policy only when preferences are weak. When preferences are strong, the optimal rule is unique, and it does not vary by level of policy complexity. These findings significantly differ from those obtained for static mechanisms. Our study also implies that more complex policies are more likely to be persistent, even after a change in political preferences.
Pongou, Roland and Tondji, Jean-Baptiste, Political Design Meets Policy Complexity (March 10, 2020). Available at SSRN: https://ssrn.com/abstract=3287477 or http://dx.doi.org/10.2139/ssrn.3287477