Balancing the power to appoint officers☆
Introduction
Appointing people to office is one of the main ways how the powerful exert their influence in society. But the ability of any authority to appoint officers is often limited by the existence of other “de jure” or “de facto” powers.
In this paper we study a class of methods that allow several agents to share the power to appoint. We call them rules of k names, and they work as follows. The set of deciders is divided into two groups: the proposers and the chooser. Proposers consider the set of all candidates to a position and screen k of them. Then, the chooser picks the appointee out of these k names.
Indeed, rules of k names can vary, depending on the composition of the set of proposers, on the value of k, and also on the voting procedure adopted to form a list of k candidates.
Here we focus in a specific family of rules of k names that is used in many practical cases. This family adopts the following procedure to form the list of k names: each proposer votes for v candidates, and then the k most voted candidates get into the list. Though one can think of other methods to select the k names, the ones we consider are simple and frequently used. We call the resulting systems v-rules of k names.
Rules of that form have been used in the past and are still very much used in the present. They seem particularly fit to give partial decision power to different parties that are interested in the workings of an institution, and want to have a voice when selecting its officers. Historically, rules of k names were used within the Roman Church since the early middle ages, when secular rulers tried to control the appointment of bishops, while the clergy would rather decide on its leaders. And similar rules are still used to share the power between Rome and the local congregations. At present, in many countries (US, Argentina, Brazil, Canada, Chile, Haiti, Mexico, Spain, Turkey etc.), the seats in appellate courts are filled through the use of rules of k names. This is the case, for example, of the commission-selection, political appointment method to select judges to the state supreme courts of several US states, sometimes referred to as merit selection or the Missouri Plan. They are also used in several countries (Brazil and Turkey) to appoint the Rectors of public universities.
It is clear that the size k of the list to be submitted has an important effect on the distribution of power between the proposers and the chooser. In the extreme case where the proposers must submit the whole list of candidates, all power goes to the chooser. In the opposite extreme case where , it is the chooser who has no room left, and all the power stays in the hands of the proposers.1 We shall focus in the intermediate, non-degenerate cases where both the proposers and the chooser have influence on the final decision, and study how the values of v and k affect the actual distribution of power between the two parties. Specifically, we shall be interested in the kind of ex ante evaluation that a designer could make of different v-rules of k names.
The methods that are actually used in practice differ in the values of these definitional parameters: the size of the list is not always the same, nor its relationship with the number of votes that each proposer is allowed. There are instances where in order to participate in the choice of k candidates, each voter is allowed to submit k names. The rule used to elect Irish bishops or prosecutor-general in most Brazilian states are of this sort, with . Yet, in most cases we know, each proposer is asked to submit a vote for v candidates, with v less than k. This is the case, for example, when choosing public university rectors in Brazil (), members of Chile's courts of justice () or Chile's Supreme Court ().
In order to understand what is the actual power of each party under any given rule, we first study the strategic behavior of different agents operating under it. This positive point of view leads us to analyze a strategic game of interaction between the different proposers, when determining what list to submit. Much of our work is geared to identify the resulting equilibria for different values of v and k. Armed with this understanding of equilibria, and the power distribution that they imply, we can then turn to more normative questions. Can we balance the power among the parties? If so, what choices of parameters will be appropriate?
The paper is organized as follows. In the next section we provide a summary of our findings in preceding papers on rules of k names, and then discuss several related papers. Section 3 provides the general setup and a discussion of the strategic issues that arise. In view of their complexity, Section 4 concentrates on the analysis of power balance within the limits of a specific but still rich model that focuses in the case where agents are polarized. Conclusions follow in Section 5, and proofs appear in the Appendix A.
Section snippets
Literature review
There does not seem to be a body of literature specifically devoted to study appointment rules with their checks and balances. Of course there exist many voting rules that can be adapted to this specific purpose, and rules of k names can also be used for other purposes. But we think it is useful to focus in the case of appointments, and to pay special attention to methods that especially fit that purpose.
We have studied rules of k names in two preceding papers. Barberà and Coelho (2008) focuses
The setup
In this section we formally define rules of k names and the games they induce. We observe that, in addition to other structural features, like the number of proposers, the number of candidates and the size k of proposed candidates, a full specification of a rule of k names also requires to define the screening rules by which the proposers decide what names go into the list. In principle, this method could remain unspecified, or be rather complicated. But in actual practice simple and specific
Choosing among v-rules of k names under polarized opinions
In this section we study how the choice of the parameters v and k can help to satisfy different normative properties, like the equalization of the power among different groups of participants in an appointment. Before engaging in that analysis, we present a family of societies to which we restrict attention. Example 1, Example 2 in the preceding section suggest that we cannot expect to develop a full general characterization of the equilibrium outcomes when proposers are sufficiently
Concluding remarks
Rules that contemplate several stages to arrive at a final choice are widely used. Some people are in charge of screening, then others choose among those candidates that were not screened out. The very idea of dividing these tasks may arise from very diverse reasons. One of them, that we consider important but we did not follow here, would be in the line of Condorcet's Jury Theorem for a common values setup: assigning to each agent responsibility for those partial decisions that she is better
References (23)
- et al.
On the rule of k names
Games Econ. Behav.
(2010) The Condorcet criterion and committee selection
Math. Soc. Sci.
(1985)- et al.
Core of effectivity functions and implementation theory
J. Math. Econ.
(1982) Voting by veto
J. Public Econ.
(1978)- et al.
Effectivity Functions in Social Choice
(1991) - et al.
Sum of us: strategyproof selection from the selectors
- et al.
Voting as a lottery
(2010) Political individualism, positional preferences and optimal decision-rules
- et al.
How to choose a non-controversial list with k names
Soc. Choice Welfare
(2008) - et al.
Choosing how to choose: self-stable majority rules and constitutions
Quart. J. Econ.
(2004)
Binding versus final-offer arbitration: a combination is best
Manag. Sci.
Cited by (13)
Generalized Veto Core and a Practical Voting Rule with Optimal Metric Distortion
2023, EC 2023 - Proceedings of the 24th ACM Conference on Economics and ComputationApproval-based shortlisting
2023, Social Choice and WelfareOn guarantees, vetoes, and random dictators
2023, Theoretical EconomicsOrder-k rationality
2022, Economic TheoryWorst case in voting and bargaining
2021, arXiv
- ☆
The authors gratefully acknowledge the support from the Spanish Ministry of Science and Innovation through grant “Consolidated Group-C” ECO2008-04756 and FEDER, from the Generalitat de Catalunya, Departament d' Universitats, Recerca i Societat de la Informació through the Distinció per a la Promoció de la Recerca Universitària and grant SGR2009-0419. Salvador Barberà acknowledges financial support from the Spanish Ministry of Economy and Competitiveness, through the Severo Ochoa Programme for Centres of Excellence in R&D (SEV-2011-0075) and grant ECO2014 53051-P. Danilo Coelho acknowledges the financial support from the Brazilian Ministry of Science Technology and Innovation, through CNPq Research Productivity Scholarship, grant no. CNPQ/PQ-2 302018/2012-3. We thank Miguel Ballester, Alexandre Carvalho, Anke Gerber, Matthew Jackson, David Jimenez, Oscar Nupia and Carmelo Rodríguez-Álvarez for useful comments.