Partially Revealing Campaign Promises

• Published date: 01 April 2017
• DOI: http://doi.org/10.1111/jpet.12189
• Date: 01 April 2017
• Author: ELENA PANOVA

PARTIALLY REVEALING CAMPAIGN PROMISES

ELENA PANOVA

Toulouse School of Economics, University of Toulouse-1, Capitole

Abstract

Candidates competing for political office give promises to voters. There

are no legal restraints preventing incumbents from breaking their elec-

toral promises. This paper models campaign promises as pure cheap

talk and asks why is it influential. We propose that a candidate’s cam-

paign promises are a partially revealing signal of her policy preference

type. The incumbent’s policy choice is yet another signal of her type.

Policy choice is a costly signal (unlike campaign promises). The in-

cumbent keeps her electoral promises in order to preserve ambiguity

about her type, which is necessary to assemble a winning majority for

reelection. She keeps her promises regardless of information about the

efficiency of different public policies which she receives upon taking

office. Therefore, campaign promises generate inefficiencies in public

policy.

1. Introduction

The candidates competing for office give promises on various policy issues: corporate

taxes, healthcare policy, minimum wage, immigration, gay marriage, abortion, and so.

The winners face no legal restraints preventing them from breaking their word to the

voters. Yet, they deliver most of their promises (Lederman and Pomper 1980; Krukones

1984; Fishel 1985; Budge, Robertson, and Hearl 1987; Petry 1995; Royed and Borrelli

1997). How can we explain this behavior without assuming any direct cost of breaking

promises? Conventional wisdom suggests that incumbents keep their electoral promises

in order to win reelection. However, why do incumbents who keep their promises win

against challengers?

Elena Panova, Toulouse School of Economics, University of Toulouse-1, Capitole, 21 Allee de Brienne,

31000 Toulouse, France (e panova@yahoo.com).

I am grateful to the editor and anonymous referees for their constructive comments and insightful

suggestions. I also thank Francis Bloch, Oliver Board, Stephen Coate, Gary Charness, Stefan Krasa,

Jean-Marie Lozachmeur, Nicolas Marceau, John Morgan, Michael Peters, Julia Shvets, and Jean Ti-

role for valuable feedback on earlier versions of this paper. My work has benefited from attention

of seminar participants at New Economic School, Universit´

edeMontr

´

eal, Universit´

eduQu

´

ebec `

a

Montr´

eal, Simon Fraser University, University of Victoria, University of Pittsburgh, and participants of

Annual Meetings of the American Political Science Association, the Canadian Economic Theory Con-

ference, and the Canadian Economic Association, and the Third World Congress of the Game Theory

Society.

Received September 6, 2014; Accepted December 20, 2015.

C2016Wiley Periodicals, Inc.

Journal of Public Economic Theory, 19 (2), 2017, pp. 312–330.

312

Partially Revealing Campaign Promises 313

A possible reason is that the voters deduce some information about an incumbent’s

policy preferences first, from her electoral promises (pure cheap talk), and then from

her policy choice (costly signal). Due to the electoral competition, the incumbent’s

promises reveal her policy preferences to a large extent. In order to win a majority for

reelection, the incumbent has to avoid any further revelation. Therefore, she keeps her

promises.

We formalize this argument by modeling first the electoral competition, and second

the incumbent’s policy choice in face of upcoming reelection (earlier literature models

these two stages of political process independently; more on this in Section 2).1

We consider a game with two successive elections: an open seat and the incumbent-

challenger.2In each election, two candidates with private information about their policy

preferences (types) compete for office via simple majority vote. The winner takes office

and chooses public policy. A voter votes for the candidate who is more likely to share

the voter’s preferences, conditional on the voter’s information.

Policy space and preferences are modeled as in Maskin and Tirole (2014).3The

voters are divided in minorities benefiting from different public policies. Public policy

benefiting one minority is costly to other minorities (no minority is “median,” unlike

in spacial models). Choosing public policy is equivalent to choosing the beneficiaries

among the minorities.4A candidate for office has the same policy preferences as one of

the minorities—which minority is her private information.

The game proceeds as follows. First, there is an open-seat election. Two candidates

for office give campaign promises regarding the policies they would undertake if in

office. A candidate can give any promises she wishes at no cost, and she has no com-

mitment to keep any of her promises if in office. The winning candidate chooses pub-

lic policy and runs for reelection against an “untried” challenger. The winner of the

incumbent-challenger election chooses any public policy she wishes.

Proposition 1 shows that this game has an equilibrium with partially revealing

campaign promises. In this equilibrium, the voters deduce that a candidate’s preferred

policies are among those she is promising. A candidate promises to implement her pre-

ferred policies plus some other policies, so as to assemble a winning majority of voters on

her side. If in office, she has to keep these voters on her side in order to win reelection.

So, she delivers the promised policies.

Because campaign promises are cheap talk, the game has equilibria in which the

voters ignore campaign messages and the candidates “babble” in various ways. We filter

these equilibria away by a refinement based on neologisms. Proposition 2 shows that

campaign promises are partially revealing in any refined equilibrium.

Proposition 3 shows that campaign promises generate inefficiencies. The reason

is that the incumbent receives information about the efficiencies of different policies

upon achieving office. However, she keeps her promises regardless of this information.

The paper is organized as follows. The next section reviews related literature.

Section 3 presents the model. Section 4 describes an equilibrium with partially

1The models of electoral competition do not study the winner’s policy choice, while the models of

electoral accountability have no actual campaign stage.

2We focus on the game with two elections which is just sufficient for our purposes. A finite horizon of

our game makes it apparent that our argument is different from the reputational argument proposed

in the literature.

3This approach is most convenient to model the candidates’ campaign promises and the incumbent’s

policy choice as signals.

4For example, cutting college tuitions is equivalent to benefiting young voters, overhauling immigra-

tion policy is equivalent to benefiting Hispanic voters, and so on.