The Provision of Relative Performance Feedback: An Analysis of Performance and Satisfaction

• Date: 01 March 2016
• Published date: 01 March 2016
• Author: Ghazala Azmat,Nagore Iriberri
• DOI: http://doi.org/10.1111/jems.12151

The Provision of Relative Performance Feedback:

An Analysis of Performance and Satisfaction

GHAZALA AZMAT

Queen Mary University of London

Mile End Road, E1 4NS, UK; and

Centre for Economic Performance

London School of Economics and Political Science

Houghton Street, London, WC2A 2AE, UK

g.azmat@qmul.ac.uk

NAGORE IRIBERRI

Foundations of Economic Analysis I

University of the Basque Country UPV/EHU

Av. Lehendakari Aguirre83, 48015 Bilbao Spain; and

IKERBASQUE Basque Foundation for Science

Maria Diaz de Haro 3, 6 floor 48013 Bilbao Spain

nagore.iriberri@gmail.com

This paper studies the effect of providing relative performance feedback on individuals’ perfor-

mance, under two incentive schemes. In a laboratory setup, agents perform a real-effort task. We

show that relative performance feedback increases performance when performance is related to pay

(piece-rate) but has no effect on performance when pay is independent of performance (flat-rate).

These effects are independent of the agent’s relative position. Subjects are also asked to rate their

satisfaction during the experiment. We find that under flat-rate, feedback has no effect on agents’

satisfaction, while under piece-rate, feedback about relative position affects satisfaction.

1. Introduction

Performance appraisals have become standard practice in organizations. Informing

agents about how well they are performing relative to their peers—that is, providing

workers with relative performance feedback—is a common method of implementing

performance appraisals. Given its widespread use, it is important to understand the con-

sequences of providing relative performance feedback. Managerial economics and social

psychology have devoted quite a lot of attention to the study of performance appraisals

(see Bretz et al., 1992; Levy and Williams, 2004, for reviews). Research in economics,

however, has paid little attention to relative performance feedback and its effects.

This paper studies the provision of relative performance feedback under two rel-

evant incentive schemes: performance-related pay (piece-rate incentives) and fixed pay,

independent of performance (flat-rate incentives). We examine whether thereare differ-

ences in the response to feedback under the different payment mechanisms. In order to

We thank Manel Baucells, Michael Bashshur, Gary Charness, Ada Ferrer-i-Carbonell, George Loewenstein,

Pedro Rey-Biel and anonymous referees for their comments. Ghazala Azmat acknowledges financial support

from ECO2011-30323-C03-02, Fundaci´

on Ramon Areces and the support of the Barcelona GSE Research

Network and the Government of Catalonia. Nagore Iriberri acknowledges financial support from Fundaci´

on

Ramon Areces,Ministerio de Econom´

ıa y Competici´

on (ECO2012-31626), Departamento de Educaci´

on, Pol´

ıtica

Ling¨

o´

ıstica y Cultura del Gobierno Vasco (IT869-13) y Ministerio de Ciencia e Innovaci´

on (ECO2011-25295).

C2015 Wiley Periodicals, Inc.

Journal of Economics & Management Strategy, Volume25, Number 1, Spring 2016, 77–110

78 Journal of Economics & Management Strategy

understand the extent of feedback’s effectiveness, we study the influence of feedback on

agents’ emotional state (or level of satisfaction). Moreover, given that, in practice, feed-

back is rarely provided only once, we investigate the effects of its provision over time.

By exploring all of these dimensions, we provide a comprehensive empirical analysis of

the effects of relative performance feedback and study the channels through which it is

effective.

We propose a controlled laboratory setup in which subjects perform a real-effort

task and are rewarded according to piece-rate or flat-rate incentives. There are four

working periods. Between periods, the control subjects are provided with their absolute

performance, while the treated subjects are provided with their absolute performance

and with the average performance in the session. Once feedback is provided, both control

and treated subjects’ satisfaction is elicited via questions about their happiness, arousal

and dominance levels.1

With respect to performance, under piece-rate incentives, we find that relative per-

formance feedback had a strong and positive effect on individual performance, even

after controlling for individual characteristics, such as ability. Those subjects who re-

ceived relative performance feedback increased their performance by 17% compared to

those who did not. In contrast, under flat-rate incentives, we find that the feedback does

not affect subjects’ performance. Interestingly, control subjects perform similarly under

both incentive schemes, but relative performance feedback affects performance only

when payment is related to performance. One may argue that when pay is related to

performance, the information is “relevant,” and individuals react because relative per-

formance information has consequences in terms of relative income (i.e., strong effect

under piece-rate but insignificant effect under flat-rate incentives).

With regard to the dynamic effects of providing relative performance feedback

(under piece-rate incentives), we find that in all periods, the treated subjects outperform

the untreated, although the effect becomes weaker over time. In addition, the content of

the feedback (i.e., positive (negative) feedback when agents are informed that they are

performing above (below) the average) does not affect subjects’ subsequent performance

differently because all subjects increase their performance.

When studying satisfaction, we find that the incentive scheme matters. Under

piece-rate incentives, relative performance feedback had strong effects on both happi-

ness and dominance levels. Contrary to the findings on performance, we show that the

treatment had very different effectson those who received positive versus negative feed-

back. We find that receivingpositive (negative) feedback affects subjects’ happiness and

dominance levels positively (negatively), such that when we consider only the overall

treatment effect, the opposite signs cancel out. With respectto happiness, between those

performing above and below the group average, the relative feedback leads to a seven-

percentage-point increase in the gap (or inequality) of subjects’ happiness. With respect

to dominance levels, the increase in inequality is eight percentage points. Moreover, the

inequality in both happiness and dominance increases over time with cumulative infor-

mation. We do not see any effect of relative performance feedback on satisfaction under

flat-rate incentives. This indicates that such feedback affects individuals’ emotional state

only when this information has consequences in terms of relative income (i.e., under

piece-rate).

These findings on performance and emotional response are consistent with

a competitive-preferences framework in which individuals care about income

1. See Figure 1 for a graphical presentation of the experiment.

Relative Performance Feedback 79

comparisons. Individuals get extra utility from knowing that they earn more than others

and disutility when they learn that they earn less than others—that is, individuals have

competitive preferences.

The findings from this paper have important implications for understanding

whether or not an organization would choose to provide relative performance feed-

back. We show that there are very strong effects on performance when performance

is linked to payment. This implies a strong incentive for an organization to employ

this mechanism, especially because the cost to implement it is negligible. However, we

go beyond the conventional thinking on this issue to highlight that this mechanism,

while very effective at increasing performance, also has important consequences for

individuals’ affective state or satisfaction.

Understanding how agents feel when they receive relative performance feedback

is important, as it sheds light on when the information is relevant to the agent. Bowles

(1998) argues that “economic institutions structure the tasks people face and hence

influence not only their capacities but their values and psychological functioning as

well.” In recent years, along with more-traditional outcome variables, economists and

managers have started to give more weight to individuals’ happiness as an important

outcome variable. Individuals’ well-being has been linked with economic outcomes.

For example, labor-market studies that include broader measures than happiness have

found that workers’ psyches are correlated with their morale and motivation, as well as

with other relevant issues such as firms’ retention ability (see Bewley, 1999).2

The paper is organized as follows. Section 2 includes a review of the exist-

ing literature. Section 3 describes the experimental design and procedures in detail.

Section 4 provides a theoretical framework of thinking about the expected effects. In

Section 5, we present the raw results via tables and figures. Sections 6 and 7 include the

regression analysis, quantifying the main treatment effects under piece-rate and fat-rate

incentives, respectively. Finally, Section 8 concludes.

2. Related Literature

The provision of feedback on individual effort—under either piece-rate or flat-rate

incentives—has recently been studied in several environments. Under piece-rate,

Hannan et al. (2008), using a laboratory experiment without real effort, Azmat and

Iriberri (2010) in an educational setup, and Blanes i Vidal and Nossol (2011) in a firm

setting, find that the provision of feedback has a positive effect on performance. In con-

trast, Eriksson et al. (2009), in the laboratory with real effort, find that although feedback

does not affect performance, it increases the mistake rate of the worst-performing agent.

Under flat-rate incentives, Kuhnen and Tymula (2012) study the effect of feedback in

relation to individuals’ expectations of their relative performance and find that subjects

exert more effort and expect to rank better when they are told that they may learn their

ranking.3Our paper contributes to this literature on the effect of feedback on perfor-

mance by explicitly comparing the effect of relative performance feedback under both

piece-rate and flat-rate incentives, a comparison that the existing literature does not

make. This allows us to disentangle individuals’ responses to relative income versus

relative performance comparisons.

2. Kahneman et al. (1999), Loewenstein (2000), Frey and Stutzer (2002), Krueger (2005), and McFadden

(2005) provide recent overviews of research on happiness and its relationt oeconomics.

3. Recently, Bandiera et al. (2013) and Delfgaauw et al. (2013) also considered the impact of relative

performance feedback at the team level.