Elsevier

European Economic Review

Volume 101, January 2018, Pages 528-545
European Economic Review

Monetary and non-monetary incentives in real-effort tournaments

https://doi.org/10.1016/j.euroecorev.2017.10.021Get rights and content

Abstract

Results from laboratory experiments using real-effort tasks provide mixed evidence on the relationship between monetary incentives and effort provision. To examine this issue, we design three experiments where subjects participate in two-player real-effort tournaments with two prizes. Experiment 1 shows that subjects exert high effort even if there are no monetary incentives, suggesting that non-monetary incentives are contributing to their effort choices. Moreover, increasing monetary incentives does not result in higher effort provision. Experiment 2 shows that the impact of non-monetary incentives can be reduced by providing subjects with the option of leaving the laboratory early, using an incentivized timeout button, or working on an incentivized alternative activity. Experiment 3 revisits the relationship between monetary incentives and effort provision using the insights from Experiment 2. Using a design with an incentivized alternative activity, we show that participants increase effort in response to monetary incentives. Taken together, the findings from the three experiments suggest that results from real-effort tasks require a careful evaluation and interpretation of the motivations underlying the observed performance.

Introduction

The impact of incentives on the supply of effort is a central question in economics. In recent years, economists have been increasingly relying on laboratory experiments, especially experiments with real-effort tasks, to study this question. Real-effort tasks require experimental participants to exert actual physical or mental effort to earn money.1 They are seen as a better representation of how people earn their income in the real world in comparison to the alternative ‘stated effort’ designs, where subjects are given information on the costs and benefits of different effort options and are asked to state the level of effort they would like to exert (see, e.g., Charness and Kuhn, 2011).

While the use of real-effort tasks may contribute to the external validity of experimental results, it may also come at a cost of loss of control over incentives.2 The main concern voiced in the literature is that in real-effort tasks, researchers cannot observe the cost of effort which differs across subjects. This makes it challenging to compare the behavior of subjects against theoretical benchmarks derived from assumed cost-of-effort functions.3 In contrast, in stated-effort experiments, the parameters guiding the optimization problem of subjects are clear.

Not knowing the (monetary) marginal cost of effort is not the only reason for loss of control in real-effort tasks. The introduction of a real-effort task also means that non-monetary incentives (such as task enjoyment) can play a larger role in the determination of effort. Hence, if the goal of the researcher is to measure the impact of monetary incentives on effort provision, clearly this becomes more challenging. The presence of non-monetary incentives may make it difficult to identify the impact of monetary incentives. This may be the reason why results from experiments using real-effort tasks provide mixed evidence on the relationship between monetary incentives and effort provision. For example, in their survey article, Camerer and Hogarth (1999) report that in some real-effort studies subjects respond to monetary incentives, while in others incentives fail to provoke a response or may even backfire (see also Jenkins et al., 1998, Gneezy and Rustichini, 2000, Pokorny, 2008, Ariely et al., 2009, Araujo et al., 2016; DellaVigna and Pope, 2017).

This evidence suggests that it is important to have a systematic exploration of the impact of monetary and non-monetary incentives in order to determine under what circumstances they increase effort. The goal of this paper is to identify the impact of monetary incentives on effort provision by reducing the impact of non-monetary incentives. We report findings from three related experiments designed to study effort provision within a real-effort contest setup. Contests are used in many economic, political, and social environments, e.g., competition for bonuses and promotions within firms, patent races, lobbying for government contracts, elections, political conflicts, and sports. Our approach is based on the conjecture that, in addition to money, the following factors may contribute to participants’ effort choices in real-effort contests: (i) task enjoyment; (ii) the desire to do what is expected of one and/or something productive/useful for others (which results in the experimenter demand effect); and (iii) competitiveness.

Experiment 1 presents the ‘puzzle’ we aim to address in this paper. Subjects participate in two-player tournaments with two prizes. We eliminate the impact of monetary incentives on behavior by setting the prizes equal to each other in the baseline treatment. We show that effort provision in this treatment is not significantly different from effort provision in two other treatments with low and high monetary incentives. Hence, non-monetary incentives by themselves are sufficient to motivate subjects to exert significant levels of effort in real-effort tasks. Moreover, providing monetary incentives does not have an impact on effort provision. This result is in line with other findings in the literature which document that behavior in real-effort tasks is relatively insensitive to changes in monetary incentives (e.g., Camerer and Hogarth, 1999, Araujo et al., 2016).

In Experiment 2, we design treatments to reduce the impact that the non-monetary incentives mentioned above have on effort. Using a simple framework, we explain that this can be done either directly by reducing the benefit subjects receive from task enjoyment, etc. (e.g., by picking a task that is likely to be less enjoyable), or indirectly by increasing the cost of working on the task. We consider two treatments aimed at increasing the opportunity cost of effort (by allowing participants to leave early and by providing them with a small payment for taking a timeout), and a third treatment with an incentivized alternative activity that aims to increase the opportunity cost of effort at the same time it decreases the non-monetary benefits subjects may receive from performing the real-effort task. We show that participation and effort in the real-effort task are significantly lower in all three treatments as compared to the baseline treatment from Experiment 1. Moreover, introducing an incentivized alternative activity that pays a piece rate is the most effective way of reducing the impact of non-monetary incentives, while allowing subjects to leave early is the least effective.

In Experiment 3, we again examine the response of agents to changes in monetary incentives. However, this time we use the design with an incentivized alternative activity from Experiment 2. We find that, in contrast to our results from Experiment 1, participants increase effort monotonically in response to monetary incentives. Hence, our results reveal that, with appropriate modifications to the way real-effort tasks are implemented in the laboratory, it is possible for the experimenter to observe changes in effort provision as the level of monetary incentives is varied.

In general, the findings from our three experiments highlight that understanding the role of incentives in real-effort tasks has important implications. We provide a framework that allows the researcher to mitigate the impact of non-monetary incentives in real-effort experiments. Beyond studying the role of monetary incentives on effort, doing this can be important if the researcher wants to examine how the process of earning one's income affects the decisions in later stages of the experiment. If non-monetary incentives (e.g., experimenter demand effect or task curiosity) play a bigger role in determining effort choices, then subjects’ behavior in the later stage may be driven by their response to these incentives rather than differences in income levels. In this case, failing to acknowledge the role that non-monetary incentives play on effort provision in the first stage may result in a misidentification of the causal effect.

The rest of the paper proceeds as follows. In the next section, we review the related literature and discuss our contributions to it. In Sections 3–5, we present the design and results of Experiments 1, 2, and 3, respectively. We conclude in Section 6 with a discussion of the implications of our findings.

Section snippets

Related literature

We build on and contribute to different strands of the literature. Our first contribution is to the literature on the impact of incentives. We examine how experimental economists can study the role of incentives using real-effort experiments in the laboratory. A central principle of economic analysis is that people respond to incentives. Hence, economists see incentives as a powerful tool for explaining and altering individual behavior. When it comes to the provision of costly effort, standard

Design and treatments

Our first experiment aims to investigate whether individuals respond to changes in monetary incentives by altering their effort levels. Subjects participate in a contest to determine how two prizes (Prize 1 and Prize 2) will be allocated between themselves and their matched partner. Each subject i chooses an effort level, ei. Subjects make their effort choice by performing a real-effort encryption task (Erkal et al., 2011) for 10 minutes. In this task, each subject is given an encryption table

Framework

Our design in Experiment 2 is motivated by the following simple framework, where subjects participating in real-effort tasks are assumed to make their effort choices by comparing the marginal cost and marginal benefit of exerting one more unit of effort. Formally, let ei denote the effort chosen by subject i. We assume that the marginal cost of effort is determined by the opportunity cost of time and is convex in effort (cei2). For example, the opportunity cost of time in Experiment 1 is

Treatments and hypotheses

Experiment 3, similar to Experiment 1, aims to detect responses to changes in monetary incentives. However, we now mitigate the impact of non-monetary incentives by using the insights from Experiment 2. Since we find treatment A1A2 to be the most effective, we choose to implement this treatment in Experiment 3 and vary monetary incentives.

Panel (c) of Table 1 provides a summary of the treatments in Experiment 3. In treatments A1A2-L and A1A2-H, we vary the value of Prize 1. In parallel with

Conclusion

In recent years, real-effort tasks have been growing in popularity. In this paper, we design three real-effort experiments spanning eight treatments to examine a fundamental question in economics about whether subjects respond to monetary incentives in their effort choice. Evidence from the laboratory has thus far provided mixed findings on the relationship between monetary incentives and effort provision. We re-visit this issue in Experiment 1 within a two-player real-effort contest and find

Acknowledgments

This work was supported by the Australian Research Council (DP1094676). Koh acknowledges support from the Melbourne International Research Scholarship. The authors would like to thank Tim Cason, Cary Deck, Nick Feltovich, Simon Gächter, Phil Grossman, Lingbo Huang, Guy Mayraz, Daniele Nosenzo, Jörg Oechssler, Roman Sheremeta, Bob Slonim, Chris Starmer, Marie Claire Villeval, Tom Wilkening, the Associate Editor, three anonymous reviewers, participants at the 10th Annual Australia New Zealand

References (56)

  • J. Abeler et al.

    Reference points and effort provision

    Am. Econ. Rev.

    (2011)
  • G.A. Akerlof et al.

    Identity Economics: How Our Identities Shape Our Work, Wages, and Well-Being

    (2010)
  • F.A. Araujo et al.

    The slider task: an example of restricted inference on incentive effects

    J. Econ. Sci. Assoc.

    (2016)
  • D. Ariely et al.

    Large stakes and big mistakes

    Rev. Econ. Stud.

    (2009)
  • N. Bardsley et al.

    Experimental Economics: Rethinking the rules

    (2010)
  • J. Berger et al.

    Can losing lead to winning?

    Manag. Sci.

    (2011)
  • T. Besley et al.

    Competition and incentives with motivated agents

    Am. Econ. Rev.

    (2005)
  • S. Bowles et al.

    Economic incentives and social preferences: substitutes or complements?

    J. Econ. Lit.

    (2012)
  • C.F. Camerer et al.

    The effects of financial incentives in experiments: a review and capital-labor-production framework

    J. Risk Uncertain.

    (1999)
  • J. Carpenter et al.

    They come to play: supply effects in an economic experiment

    Ration. Soc.

    (2010)
  • J. Carpenter et al.

    Tournaments and office politics: evidence from a real effort experiment

    Am. Econ. Rev.

    (2010)
  • G. Charness et al.

    The dark side of competition for status

    Manag. Sci.

    (2014)
  • B. Corgnet et al.

    Why real leisure really matters: incentive effects on real effort in the laboratory

    Exp. Econ.

    (2015)
  • E. Dechenaux et al.

    A survey of experimental research on contests, all-pay auctions and tournaments

    Exp. Econ.

    (2015)
  • E.L. Deci et al.

    A meta-analytic review of experiments examining the effects of extrinsic rewards on intrinsic motivation

    Psychol. Bull.

    (1999)
  • DellaVigna, S., List, J.A., Malmendier, U., Rao, G. 2016. Estimating social preferences and gift exchange at work....
  • S. DellaVigna et al.

    What motivates effort? Evidence and expert forecasts

    Rev. Econ. Stud.

    (2017)
  • D.L. Dickinson

    An experimental examination of labor supply and work intensities

    J. Labor Econ.

    (1999)
  • Cited by (43)

    • Don't pay the highly motivated too much

      2023, Journal of Behavioral and Experimental Economics
    • Because I (don't) deserve it: Entitlement and lying behavior

      2021, Journal of Economic Behavior and Organization
    View all citing articles on Scopus
    View full text