Bård Harstad† CMS-EMS Discussion Paper 1442 21 June 2007

Abstract Principals, such as voters or districts, typically bene?t by strategically delegating their bargaining and voting power to representatives di?erent from themselves. There are con?icting views in the literature, however, of whether such a delegate should be «conservative» (status quobiased) or instead «progressive» relative to his electorate. I show how the answer depends on the political system in general, and the majority requirement in particular. A larger majority requirement leads to conservative delegation, but «sincere» delegation is always achieved by the optimal voting rule. The results may be interpreted as normative recommendations for the EU’s future «Constitution».Key words: Strategic delegation, collective decisions, voting rules JEL Classi?cation: D71, D72, F53, H11

? Building on my thesis, the paper has bene?ted from the advice of Philippe Aghion, Patrick Bolton,
Oliver Hart, Torsten Persson, my colleagues at MEDS and several seminar audiences. Christina Lönnblad has provided editorial assistance.

† MEDS, Kellogg School of Management,Northwestern University, 2001 Sheridan Rd., Evanston, IL
60208. E-mail: [email protected].

1. Introduction
Political decisions are made by delegates, not the citizens themselves. For example: In the European Council, as well as the European Commission, each country is represented by a delegate who, on its behalf, negotiate and vote on whether policies should be approved. Each country may havean incentive to strategically delegate to a representative that is biased one way or the other. What determines the incentives to delegate strategically? Do they depend on the political system? Can institutions be designed to ensure «sincere» delegation? Strategic delegation may be costly from a social point of view: If the delegates are «conservative» (status quo biased), they tend not toimplement projects even if they are socially optimal. If, instead, the delegates are «progressive» (public-good lovers), they implement projects even if these are too costly. Strategic delegation may thus separate voters’ preferences from those of the politicians, leading to a «democratic de?cit», a characteristic often attributed to the EU. It is thus important to understand when and how votersstrategically appoint representatives. But there is a controversy in the literature on delegation. Starting with Schelling (1956), a large bargaining literature shows how principals delegate to status quo biased agents to gain «bargaining power». Such agents are less desperate in reaching an agreement and, therefore, able to negotiate a better deal.1 On the other hand, a more recent literature in politicaleconomy argues that «voters attempt to increase the probability that their district is included in the winning coalition by choosing a representative who values public spending more» (Chari, Jones and Marimon, 1997, p. 959). The majority coalition will typically consist of the winners, i.e., the representatives who are least costly to please (as in Ferejohn, Fiorina and McKelvey, 1987). And,being a member of the majority coalition is important, since this shares
Schelling’s argument is formalized by Jones (1989) and Segendor? (2003) in two-player games. MilesiFerretti, Perotti and Rostagno (2002) compare majoritarian and proportional systems where three districts delegate to gain bargaining power. An n-person bargaining game is studied by Brückner (2003); he ?nds that the bias may bemitigated by relaxing the unanimity requirement. Besley and Coate (2003) study strategic delegation in a context where two districts maximize joint utility. In a similar model, Dur and Roelfsma (2005) show that the direction of delegation may go either way, depending on the cost-sharing rules.


the surplus and expropriates the minority whose votes it does not need. To increase the…