Measuring Institutions: Independence, Authority, and Accountability, with an Application to Legislative Process
49 Pages Posted: 17 Jun 2013
Date Written: 2012
Abstract
Veto player theory has profoundly influenced political science, contributing both to theoretical clarity and to a minor revolution in defining and measuring political systems. The simplicity that makes the approach conceptually powerful is problematic for modeling and empirical analysis, however. For instance, it might not be appropriate to expect institutional or partisan veto players to face the same incentives about whether or when to wield their veto authority; moreover, some institutionally defined players often possess important powers distinct from legislative structure and process to punish (or reward) veto players. The effect, we argue, is to situate the decision whether to exercise a veto in a framework similar to principal-agent models. Veto players' payoffs thus extend beyond the policy at hand, depending on where they are in the broader institutional structure. Consequently, we believe that most empirical studies of veto points or checks are built on misleading measures that are too limited for the goals they set. We thus propose a novel approach to operationalizing the concept of checks, building on the idea that vetoes should be considered in their institutional context, where other constitutional features can strengthen or undermine a veto player's willingness to act. We use IAEP data (Regan, Frank and Clark 2009) to generate a cross-national time-series index of institutional checks, which we then use to replicate several prominent studies in order to highlight issues of validity and reliability and to illustrate the importance of taking seriously institutional authorities beyond those that affect 'how a bill becomes a law.' We also stress that our measure is purely institutional, distinguishing partisan and other preference-based concerns from institutional authority, but discuss ways to incorporate 'player preferences in recognition that outcomes result from the combination of institutions and preferences' using existing data.
Suggested Citation: Suggested Citation