Set them (almost) free: Discretion in electoral campaigns under incomplete information

• Date: 01 August 2019
• Author: Elena Manzoni,Vardan Baghdasaryan
• Published date: 01 August 2019
• DOI: http://doi.org/10.1111/jpet.12363

Received: 2 January 2018

|

Revised: 10 September 2018

|

Accepted: 17 January 2019

DOI: 10.1111/jpet.12363

ORIGINAL ARTICLE

Set them (almost) free: Discretion in electoral

campaigns under incomplete information

Vardan Baghdasaryan

1,2

|

Elena Manzoni

3

1

College of Business and Economics,

American University of Armenia, Yerevan,

Armenia

2

CERGE‐EI Foundation, Prague, Czech

Republic

3

Department of Economics, University of

Verona, Verona, Italy

Correspondence

Elena Manzoni, Department of Economics,

University of Verona, via Cantarane 24,

37129, Verona, Italy.

Email: elena.manzoni@univr.it

Abstract

The paper analyzes a model of electoral campaigning as a

problem of competitive delegation. We consider an environ-

ment characterized by two sources of uncertainty: uncer-

tainty on the nature of the optimal policy and uncertainty on

the candidates’biases. Although voters know whether the

candidate is left‐or right‐wing, they do not know the extent

of the bias. In this environment, discretion may benefit

voters as it allows the elected politician to adjust his policies

to the state of the world. The paper shows that the optimal

set of promises must be a closed interval, whose size is

decreasing in the expected bias of the candidate. An

example where the set of types is finite shows that an

increase in the variability of candidates’types may either

increase or decrease the voters’willingness to grant

discretion to politicians.

1

|

INTRODUCTION

“First: the reduction of the fiscal burden with total exemption for incomes up to 22 million of Italian

liras per year; the reduction of the tax rate for incomes up to 200 million to 23%; the reduction of

the tax rate for incomes above 200 million to 33%.”

1

(S. Berlusconi, 2001)

“… our first promise is that we will not steal money from the Italians and that we will reduce the

fiscal burden below 40% of the GDP.”

2

(S. Berlusconi, 2008)

J Public Econ Theory. 2019;21:622–649.wileyonlinelibrary.com/journal/jpet622

|

© 2019 Wiley Periodicals, Inc.

1

In the original language: “Punto primo: Abbattimento della pressione fiscale con l’esenzione totale dei redditi fino a 22 milioni di lire annui; riduzione al

23 per cento dell’aliquota per i redditi fino a 200 milioni; riduzione al 33 per cento dell’aliquota per i redditi sopra i 200 milioni.”From Corriere della Sera,

online version, March 3, 2008.

2

In the original language: “… la nostra prima promessa …è che non metteremo mai le mani nelle tasche degli italiani e che abbasseremo la pressione fiscale

sotto il 40% del Pil.”From Corriere della Sera, online version, February 29, 2008.

“What we need is a decisive government that says we are going to have a budget that will help small

businesses like the ones here with a cut in Corporation Tax, saying that any new business doesn’t

have to pay National Insurance on the first ten new jobs, busting open state procurement so you can

all bid for the state contracts, and make sure that you can make money for your businesses out of

what the government does.”

3

(D. Cameron, 2010)

“I want to make it absolutely crystal clear that our objective when the fiscal circumstances are right,

is to lower all taxes. We want to lower all taxes. We really are the party of lower, simpler, fairer

taxes—look at our record in government.”

4

(T. Abbott, 2013)

Vagueness of electoral campaign statements is a common element of political elections, experienced by voters

of every country and political arena. As the above‐mentioned examples show, political announcements during the

electoral campaign may have different degrees of vagueness, therefore leaving different degrees of postelectoral

freedom to elected politicians.

The issue of vagueness in electoral campaigns has attracted considerable interest in the economic and political

science literature. In most cases, the phenomenon is described as a candidate‐driven phenomenon in which

politicians are able to maintain discretionality despite facing ambiguity‐averse voters.

5

There is, however,

increasing evidence that ambiguity can, in fact, be supported by voters (Hersh and Schaffner, 2013; Tomz and Van

Houweling, 2009). Although candidates have a clear interest in conducting an electoral campaign that guarantees

the possibility of choosing their preferred policy ex‐post, voters’preferences for a vague politician are less obvious.

This paper provides a rationale for this type of voters’preferences and investigates whether and how incomplete

information on candidates’policy preferences may affect the discretionality of the policy announcements.

We introduce a model in which ambiguity of the electoral promises results from a delegation objective:

depending on how changing the environment is, the tension between control and discretion that the electorate

experiences may be resolved with voters having a preference for some degree of discretion. Specifically, voters may

prefer to elect a politician with a political program that leaves him free to operate in a given policy region once

elected, targeting the policy to the state of the world ex‐post. Notably, this result holds true even when voters do

not know the candidate’s bias. In particular, the level of discretion that the voters are willing to concede to the

candidates is inversely related to their expected bias (where the bias is defined with respect to the median voter’s

preferences). In the main model, the type structure is such that the variability and the expectation of the bias are

interlinked. Therefore, we introduce an example with a discrete‐type structure, which allows for a comparative

statics analysis on the variability of the bias for a given level of expected bias. The relation between the level of

discretion and the uncertainty over the candidates’types varies depending on the values of the parameters; in

some parametric regions, an increase in the uncertainty over the candidate’s types decreases the level of discretion

that the voters are willing to leave, as one would expect given that the voters are risk‐averse. In other parametric

regions, however, the effect is the opposite and more uncertainty implies more discretion for the candidates; this is

due to an insurance effect that the bounds of the set of promises exert: increasing the upper or lower bound

(depending on whether the candidate is right‐wing or left‐wing) has the effect of giving more discretion to good

candidates’types, and not so much to bad candidates’types, who will hit the bound more often and will be

constrained by it anyway.

The paper contributes to different strands of literature. First of all, we relate to the extensive theoretical

literature on ambiguity in electoral campaigns that started with Shepsle (1972) and Page (1976). The literature

discusses quite diverse reasons that may give rise to ambiguity in electoral campaigns. A strand of the literature

3

http://conservative‐speeches.sayit.mysociety.org/speech/601487

4

http://www.abc.net.au/news/2014‐05‐01/ fact‐file‐what‐tony‐abbott‐promised‐on‐tax/5420226

5

See, among others, Alesina and Cuckierman (1990), Alesina and Holden (2008), Aragones and Neeman (2000), Glazer (1990), and Meirowitz (2005).

BAGHDASARYAN AND MANZONI

|

623