Confidence Management: On Interpersonal Comparisons in Teams

• DOI: http://doi.org/10.1111/jems.12037
• Author: Bauke Visser,Benoît S. Y. Crutzen,Otto H. Swank
• Published date: 01 December 2013
• Date: 01 December 2013

Confidence Management: On Interpersonal

Comparisons in Teams

BENOˆ

IT S. Y. CRUTZEN

Department of Economics, Erasmus School of Economics,

Erasmus University Rotterdam, DR Rotterdam

The Netherlands

crutzen@ese.eur.nl

OTTO H. SWANK

Department of Economics, Erasmus School of Economics

Erasmus University Rotterdam, DR Rotterdam,

The Netherlands; and Tinbergen Institute, Amsterdam, The Netherlands

swank@ese.eur.nl

BAUKE VISSER

Department of Economics, Erasmus School of Economics

Erasmus University Rotterdam, DR Rotterdam,

The Netherlands; and Tinbergen Institute, Amsterdam, The Netherlands

bvisser@ese.eur.nl

Organizations differ in the degree to which they differentiate employees by ability. We analyze

how the effect of differentiation on employee morale may explain this variation. We characterize

sufficient conditions for the manager to refrain from differentiation. She refrains from differ-

entiation when employees are of similar ability, especially if absolute levels are high. Avoiding

differentiation boosts the self-image of employees. To limit the negative effects of differentiation,

the manager’s strategy often relies on the coarsest message set possible. The likelihood that the

manager differentiates depends on the presence of synergies between employees and on the con-

vexity of the cost of effort function. Finally, we show that in the absence of commitment no

differentiation is chosen too often.

“There’s differentiation for all of us in our first 20 years. Why should it stop

in the workplace?” (in Welch 2001, former CEO of General Electric).

“Comparisons are odious” (Saying recorded from the mid-fifteenth century)

1. Introduction

An employee who has imperfect knowledge about his ability or his performance may try

to infer information from how he is being treated by his supervisor. A manager should

We thank a co-editor and two referees for their comments. We also gratefully acknowledge feedback from

the audience at presentations at ESEM 2008 meeting in Milan, the Erasmus School of Economics, European

University Institute, University of Essex, the University of Groningen, and HEC Montreal. Special thanks

are due to Leon Bettendorf, Pascal Courty, Robert Dur, Mike Gibbs, Maarten Goos, Marco Haan, Giuseppe

Moscarini, Karl Schlag, and, in particular, Vladimir Karamychev. Part of this paper was written while Bauke

Visser was a Fernand Braudel Fellow at the EUI.

C2013 Wiley Periodicals, Inc.

Journal of Economics & Management Strategy, Volume22, Number 4, Winter 2013, 744–767

Confidence Management 745

take this into account when deciding what information to disclose, either verbally or

through the actions that she takes.

The idea that people increase their self-knowledge by viewing themselves through

the eyes of others is well-known to social psychologists. It is known as the “looking-glass

self.” B´

enabou and Tirole (2003), BT, formalize this concept to show that rewards may

have a hidden cost, because a reward may signal that the manager does not trust the

employee. Giving a challenging task to a subordinate, by contrast, signals confidence

and consequently motivates.1In BT an employee wants to have an accurate perception

of his ability to make a proper decision on how much effort to exert.2With effort and

ability being complements, the more the employee is confident about his ability, the

more effort he exerts.3In such an environment, managers want employees to believe

they are as able as possible.

The aim of this paper is to analyze the looking-glass self in a context in which

agents gain self-knowledge by considering how they are being treated in comparison

with other agents. That is, we study multi-agent settings: the manager is responsible for

the performance of a unit in which various employees work. Total output depends on

individual contributions and possibly on a synergetic, team component.

Comparisons among employees also play a key role in the recent literature on in-

terim performance feedback in tournaments. The main consideration is whether making

differences among contestants public increases or decreases total performance of these

contestants. The existing literature has uncovered a number of important effects that the

provision of interim feedback can have on effort exerted before and after feedback is

given. The exact effects depend on whether individual output is determined exclusively

by effort or also by an employee’s ability, and on the interaction between effort and

ability, see Lizzeri et al. (1993), Ertac (2005), Gershkov and Perry (2009), Aoyagi (2010),

Ederer (2010), Casas-Arce and As´

ıs Mart´

ınez-Jerez (2009), and Goltsman and Mukherjee

(2010).

Apart from the fact that the employees in our model are not contestants, our pa-

per differs from the above papers and the literature on feedback in tournaments more

generally as it relaxes two assumptions that are commonly made in that literature.

First, that literature typically assumes that the manager commits to a disclosure policy

before she observes actual interim information.4The main reason is that this “allows

[one] to suppress considerations of incentive problems on behalf of the principal such

as misreporting the first-period output difference to elicit higher efforts or to selec-

tively announce interim information” (Ederer 2010, p. 7405). Second, the manager is

constrained to choose between truthfully revealing any and all differences in realized

interim outcomes and providing no information at all. As a result of these assumptions,

considerations on what can be credibly conveyed through feedback, and the precision of

that information are suppressed. In our paper, both the issue of incentive compatibility

and of the informational content of feedback are key.

1. See also Ishida (2006) and Swank and Visser (2007).

2. In BT, a person wants to learn about himself to maker better decisions. Psychologists have paid much

attention to a person’s desire to obtain accurate self-knowledge from evaluations. One source of information

are other persons’ appraisals (Felson, 1993; Baumeister,1998). A problem is that accurate feedback on abilities

is rare (Jones and Wortman,1973). Feedback tends to be too positive (Brown and Dutton, 1995).

3. BT also pay some attention to the case in which effort and ability are substitutes.

4. G¨

urtler and Harbring (2010) is an exception. We discuss their paper shortly.

5. Ederer does discuss the implications of relaxing the commitment assumption but only in the context

where effort and ability enter additively in the production of output. In this paper, effort and ability are

complements which raises interesting commitment issues.