scispace - formally typeset

Journal ArticleDOI

Why Join a Team

18 Feb 2021-Management Science (INFORMS)-Vol. 67, Iss: 11, pp 6980-6997

TL;DR: When communication is possible, the choice of high ability workers to join teams is driven by expected future financial gains from teaching rather than some variety of prosocial preferences, which has important implications for the role of adverse selection in determining the productivity of teams.

AbstractWe present experiments exploring why high ability workers join teams with less able coworkers when there are no short-term financial benefits. We distinguish between two explanations: prosocial pre...

Summary (2 min read)


  • The authors present laboratory experiments designed to study why high ability individuals might join revenue sharing teams in the absence of immediate financial benefits.
  • High ability participants who join teams can still benefit from the joy of teaching, enjoyable social interactions, and financially helping their low ability teammate, but long-term financial gains from teaching are eliminated.
  • 15 of subjects to pairs is based on their relative ability at the task; the Ice Breaker would need to take place after Stage 1. 3.c. Result 4: Consistent with H4, high ability types choose Option 2 less often in the NP treatment than in the BL treatment.


  • At the end of the allotted time, you will automatically see the results screen where you can review your answers for 60 seconds before starting the next period.
  • Before you start the second part, you should know that participants in the session ranked in the top half in terms of the number of solved puzzles (i.e., those who solved more puzzles than the median) have been selected to receive hints.
  • If you have been selected, you should now see the hints displayed in the center of your screen.
  • Please do not press Continue until you have received your sheet of paper.
  • You will keep the same partner for the 5 periods.

With option 2:

  • The total number of puzzles solved by your partner and by you will be calculated and €1.25 will be paid for each puzzle solved.
  • This group income will then be divided in two equal parts between you and your partner.
  • You should now see a side-by-side comparison of the results screens for both options.
  • In each option, you will receive information about your partner's puzzle in the middle of the screen, below your own results.
  • Please press Continue to start the first period.

Figure A2. Ice Breaker - Versions 1 and 2

  • A person can sign only one box and you cannot sign your own sheet.
  • Reflecting complementarities, (3) in the main text is modified into (A3).
  • Second, the interaction between complementarities (k) and teaching (τ) implies that the effects of k and τ are not separable; the marginal effect of k on the probability of choosing a team is an increasing function of τ and vice versa.

Different Corrections for Session Effects

  • Robust/clustered standard errors are reported in parentheses.
  • 71 Table A3 summarizes the reasons given by high ability types for choosing Option 2 in the post-experimental questionnaire.
  • This is not the order in which participants saw the reasons listed.


  • To measure whether help is being given, the authors coded all the messages sent between partners who were assigned Option 2 in BL, PR, NP, and NH.
  • Each message was coded independently by two research assistants.
  • They were instructed that they could check as many or few categories as desired.
  • Suppose a high ability type writes, “According to the hints, a 7 means you can fill in the whole row.”.
  • Cohen’s kappas are high for the categories of greatest interest - those relating to teaching - exceeding .60 in all cases.

Coding Categories

  • Explicitly referring to the hints (asking to hear about them for low ability types or talking about their content for high ability types) 4. Emotionally supportive comments related to the puzzles (e.g. celebrating getting a puzzle right, commiserating getting a puzzle wrong, encouraging each other, etc.).
  • Based on feedback from the coders, Categories 4 and 5 were difficult to differentiate because it was often not clear whether or not a comment was related to the puzzles.
  • The dataset includes all observations where the pair is assigned to Option 2 and chat is possible.
  • For giving and receiving help, the authors report the frequency for only one type since they are interested in high ability types giving help to low ability types.

Table A4: Frequency of Chat (per Period)

  • Table A4 addresses this concern by reporting OLS regressions that use a Heckman correction for selection.
  • All three are potentially correlated with the option a subject is assigned.
  • Data from the NI treatment are dropped since there is no interaction.
  • Model 1 considers data from low ability types while Models 2 and 3 use data from high ability types.

Table A5: Number of Messages Sent per Period

  • These regressions are based on 187 observations, including 107 participants assigned to Option 2, and use a Heckman correction.
  • Robust standard errors are reported in parentheses.
  • 32 Looking at first stage results, the difference between the Heckman draws for the low ability type and the tie breaker are significant at the 1% level in all four models.

Did you find this useful? Give us your feedback more

Content maybe subject to copyright    Report

Why Join a Team?
David J. Cooper
, Krista Saral
, Marie Claire Villeval
This version: 12 May 2020
Abstract: We present experiments exploring why high ability workers join teams with less able co-workers
when there are no short-term financial benefits. We distinguish between two explanations: pro-social
preferences and expected long-term financial gains from teaching future teammates. Participants perform a
real-effort task and decide whether to work independently or join a two-person team. Treatments vary the
payment scheme (piece rate or revenue sharing), whether teammates can communicate, and the role of
teaching. High ability workers are more willing to join teams in the absence of revenue sharing and less
willing to join teams when they cannot communicate. When communication is possible, the choice of high
ability workers to join teams is driven by expected future financial gains from teaching rather than some
variety of pro-social preferences. This result has important implications for the role of adverse selection in
determining the productivity of teams.
JEL Codes: C92, D23, M52, M53, J24
Keywords: Teams, teaching, revenue sharing, social preferences, self-selection, experiment
Florida State University Department of Economics and University of East Anglia School of Economics,
113 Collegiate Loop, PO Box 3062180 Tallahassee, FL 32306-2180, USA. E-mail:
University of North Carolina at Charlotte Department of Economics, 9201 University City Blvd. Charlotte,
NC 28223, USA; Webster University Geneva George Herbert Walker School of Business and Technology,
Geneva, Switzerland; GATE, Ecully, France. E-mail:
Univ Lyon, CNRS, GATE UMR 5824, 93, Chemin des Mouilles, Ecully, F-69130, France; IZA, Bonn. E-
Acknowledgments: We are grateful to participants at seminars at Florida State University, the University of Arkansas,
the University of Sydney, the University of New South Wales, the University of Geneva, the University of Lausanne,
the ASFEE conference in Nice, the COPE conference in Aachen, the ESA North American meeting in Antigua, the
EWEBE workshop in Lyon, and the VBEN meeting in Vienna for useful comments. Financial support from the
LABEX CORTEX (ANR-11-LABX-0042) of Université de Lyon, within the Programme Investissements d’Avenir
(ANR-11-IDEX-007) operated by the French National Research Agency (ANR), is gratefully acknowledged. This
research was performed within the framework of the INDEPTH program of IDEXLYON from Université de Lyon
within the Programme Investissements d’Avenir (ANR-16-IDEX-0005).

1. Introduction: Team production and team-based compensation are ubiquitous in organizations (Bandiera
et al., 2013; Owan, 2014).
The successful use of teams is considered a major component of ‘high
performance work organizations’ (Ichniowski et al., 1997; Ichniowski and Shaw, 1999). Potential benefits
include synergies in problem-solving, complementarities between team members’ inputs, better task
coordination, knowledge transfer, and employee empowerment. However, the preceding implicitly assumes
that using teams and/or team-based compensation does not affect the quality of the workers employed by
an organization. In reality, workers choose between jobs and can avoid teams through choice of
employment. It is well-known that a significant share of incentive pay’s impact on productivity is due to
selection (e.g., Lazear, 2000; Dohmen and Falk, 2011). By the same token, we cannot understand how team
production and team-based incentives affect productivity without understanding who joins teams.
There are obvious gains to joining a team if strong complementarities are present, as the increased
productivity from being in a team can offset any losses due to moral hazard or revenue sharing with lower
ability individuals. But suppose workers have heterogeneous ability and perform a task with minimal
complementarities between workers’ efforts. Why would a high ability worker want to join a revenue
sharing team? Standard economic arguments suggest they should prefer working individually for a piece
rate because they are financially harmed by an incentive system that pools their output with the production
of less able peers (Meidinger et al., 2003). Moreover, moral hazard should reduce team productivity
(Alchian and Demsetz, 1972; Holmström, 1982), leading to lower earnings for team members. If myopic
financial concerns drive the choice to join a revenue sharing team, only low ability individuals should be
willing to join and the resulting adverse selection should harm the productivity of teams.
Yet, contrary to the preceding arguments, the limited empirical evidence on this question indicates that
high ability workers are surprisingly willing to join revenue sharing teams. Hamilton et al. (2003) provide
In the U.S., a survey of Fortune 1000 companies reveals that 70% of large firms use some form of team incentives
(Lawler et al., 2001; Lawler and Mohrman, 2003). The share of large firms with more than one fifth of their employees
in self-managed teams increased from 37% to 66% between 1987 and 1996 (Lazear and Shaw, 2007). In Europe, the
incidence of teamwork varies across countries, from less than 55% of the companies in the south of Europe to more
than 75% in the UK (EWCS 2000/2001; European Foundation for the Improvement of Living and Working
Conditions; Kyzlinková et al., 2007). Similar trends can be observed in academic research (Wuchty et al., 2007).

a striking result. They study a garment company that switched its payment scheme from individual piece
rates to team compensation. The most skilled workers were disproportionately more likely to join a team
voluntarily even though team compensation often led them to lose money. Hamilton et al. suggest that
“teams offer nonpecuniary benefits to workers” (p. 469), possibly due to socialization.
We present laboratory experiments designed to study why high ability individuals might join revenue
sharing teams in the absence of immediate financial benefits. Laboratory experiments are a useful
complement to field studies of the economics of organizations because of the high degree of control
available in the lab and the ability to observe process data, specifically conversations between teammates
including attempts to teach. We designed an environment where we systematically change single elements
of the environment between treatments, isolating specific explanations for high ability participants joining
revenue sharing teams. To accentuate adverse selection, we study an intentionally tough environment with
minimal complementarities and full information about relative ability. High ability participants who join
teams have no reason to expect their productivity to increase and know they are sharing revenue with a
lower ability participant. Our conclusions are presumably not specific to our setting, as adverse selection is
a general phenomenon that should be present in less extreme environments (i.e., settings with stronger
complementarities and less certainty about being high ability).
We explore two explanations for why high ability workers might join revenue sharing teams in spite of
short-term losses. The first is that high ability workers are motivated by pro-social preferences. They might
join a revenue sharing team to increase their low ability peers’ earnings at the cost of their own, consistent
with models of other-regarding preferences (e.g., Fehr and Schmidt, 1999; Bolton and Ockenfels, 2000).
High ability workers might also enjoy the opportunity to interact socially with other workers, derive
pleasure from teaching their less able peers, or view teaching as an alternative means of increasing the
earnings of these peers. The second explanation is that high ability workers are motivated by long-term
financial benefits. If they believe they can improve the performance of their less able teammates by teaching
them, they might take a short-term financial hit due to revenue sharing in exchange for improving the long-
term productivity of the team, especially if they anticipate mandatory team membership in the future. To

separate these possible explanations, we study participants working on nonograms, a type of logic puzzle.
Nonograms provide a challenging real-effort task where teaching is possible. Participants always solve the
problems individually, eliminating any complementarities in production (i.e., positive synergies from
having a team solve the puzzle).
The experiment contains three stages. In Stage 1, participants attempt to solve a series of nonograms
with piece rate compensation for correct solutions. Prior to Stage 2, they are matched into heterogeneous
pairs by ability, as measured by Stage 1 performance, and decide whether to continue working for an
individual piece rate or to join a team with their partner (i.e., their potential teammate).
The Stage 1
performance of the partners is common knowledge, so the high ability partner (i.e., the partner who did
better in Stage 1) knows that their potential teammate is relatively poor at solving nonograms. Treatments,
described below, vary how teams are paid and how teammates interact in Stage 2. In Stage 3, participants
are forced to join a revenue sharing team with no direct interaction. Forced team membership in Stage 3
implies the existence of long-term benefits from teaching in Stage 2.
In the Baseline treatment, teams are paid in Stage 2 via revenue sharing. Teammates work on the
puzzles individually but are paid based on the number of puzzles solved by the two teammates. Between
puzzles, teammates can chat, providing an opportunity for social interaction as well as teaching. High ability
participants are given hints about solving nonograms that can be passed on to their less able teammate.
Partners know that hints will be available prior to choosing whether to join a team, making it transparent
that teaching is possible.
The other treatments vary how attractive teams are relative to the Baseline treatment. In the Piece Rate
treatment, teammates are paid a piece rate based on their individual output, rather than team output,
eliminating the short-run financial disincentive for high ability subjects to join a team. The No Interaction
treatment eliminates the chat between puzzles. Teaching is not possible, but a desire to increase the payoffs
of their low ability peers can still motivate high ability participants to join teams. The New Partner
We use the term “partner(s)” to refer to pairs of participants who are potentially teammates and the term
“teammate(s)” to refer to participants who actually join a team together.

treatment differs from the baseline by not holding partners fixed between Stages 2 and 3. High ability
participants who join teams can still benefit from the joy of teaching, enjoyable social interactions, and
financially helping their low ability teammate, but long-term financial gains from teaching are eliminated.
Finally, high ability participants do not receive hints in the No Hints treatment, limiting the possibility of
teaching and its associated benefits, pecuniary or otherwise.
A reason why a high ability participant might join a revenue sharing team involves pro-social
preferences. The literature on group identity (e.g., Chen and Li, 2009) suggests that these preferences are
stronger when social distance is low. A potential concern with our lab setting is that social distance may be
higher than in field settings. Therefore, in all five treatments we vary whether subjects participate in an ice
breaker task before beginning the experiment. This manipulates social distance, checking whether our
conclusions are sensitive to the social distance between potential teammates.
Turning to the results, 32% of high ability participants are willing to join teams in the Baseline
treatment. This doubles to 71% in the Piece Rate treatment and falls to 10% in the No Interaction
treatment. High ability participants apparently regard revenue sharing as a negative aspect of joining a team
and put little weight on improving the earnings of their low ability peer when no interaction is possible. At
13% and 15% respectively, their willingness to join teams is also quite low in the New Partner and No
Hints treatments. Neither a desire for social interaction nor joy of teaching seems to motivate high ability
participants to join teams. The Ice Breaker condition generates 31% more conversation between partners
when chat is possible, consistent with lowered social distance. However, this increase is confined to social
interactions rather than teaching, and the Ice Breaker condition does not increase the willingness of high
ability participants to join teams. Overall, the data indicate that high ability participants are almost never
willing to join revenue sharing teams unless there is a possible future financial benefit from teaching.
Consistent with the importance of teaching, “wanted to teach partner” is by far the most common reason
given by high ability participants for joining a team. Subsequent to joining teams, high ability participants
with access to the hints often try to teach their teammate by using these hints. Surprisingly, teaching does
not improve the performance of low ability teammates but instead improves the high ability teammates’

More filters

Journal ArticleDOI
TL;DR: The results show that, Physicians’ group joining behavior and the person-group fit will positively affect their online demand and reputation and it is found that physicians’ clinic title has a negative moderating effect.
Abstract: Group service in online health communities (OHCs) is a novel form of service where physicians can share knowledge and collaborate with others to provide better services to patients through the Internet. Although many literatures have explored the impact of individual group joining behavior on group performance, few researches have been done on the relationship between individual group joining behavior and their personal performance. To fill this gap, we examine the effect of individual group joining behavior on their online demand and reputation using a unique panel dataset of 8,069 physicians from a large online health community in China. In addition, how the effects differ among physicians with different clinic titles are explored and several robustness checks including propensity score matching analyses are conducted to control for self-selection among physicians. The results show that, physicians’ group joining behavior and the person-group fit will positively affect their online demand and reputation. Furthermore, we find that physicians’ clinic title has a negative moderating effect.

1 citations

More filters

Journal ArticleDOI
Abstract: z-Tree (Zurich Toolbox for Ready-made Economic Experiments) is a software for developing and conducting economic experiments. The software is stable and allows programming almost any kind of experiments in a short time. In this article, I present the guiding principles behind the software design, its features, and its limitations.

9,067 citations

Journal ArticleDOI
Abstract: There is strong evidence that people exploit their bargaining power in competitivemarkets butnot inbilateral bargainingsituations. Thereisalsostrong evidence that people exploit free-riding opportunities in voluntary cooperation games. Yet, when they are given the opportunity to punish free riders, stable cooperation is maintained, although punishment is costly for those who punish. This paper asks whether there is a simple common principle that can explain this puzzling evidence. We show that if some people care about equity the puzzles can be resolved. It turns out that the economic environment determines whether the fair types or the selesh types dominate equilibrium behavior.

8,167 citations

Journal ArticleDOI
Abstract: We demonstrate that a simple model, constructed on the premise that people are motivated by both their pecuniary payoff and their relative payoff standing, organizes a large and seemingly disparate set of laboratory observations as one consistent pattern The model is incomplete information but nevertheless posed entirely in terms of directly observable variables The model explains observations from games where equity is thought to be a factor, such as ultimatum and dictator, games where reciprocity is thought to play a role, such as the prisoner's dilemma and gift exchange, and games where competitive behavior is observed, such as Bertrand markets (JEL C78, C90, D63, D64, H41)

5,131 citations

Journal ArticleDOI
Abstract: This article studies moral hazard with many agents. The focus is on two features that are novel in a multiagent setting: free riding and competition. The free-rider problem implies a new role for the principal: administering incentive schemes that do not balance the budget. This new role is essential for controlling incentives and suggests that firms in which ownership and labor are partly separated will have an advantage over partnerships in which output is distributed among agents. A new characterization of informative (hence valuable) monitoring is derived and applied to analyze the value of relative performance evaluation. It is shown that competition among agents (due to relative evaluations) has merit solely as a device to extract information optimally. Competition per se is worthless. The role of aggregate measures in relative performance evaluation is also explored, and the implications for investment rules are discussed.

3,820 citations

Journal ArticleDOI
Abstract: This paper introduces a three-item "Cognitive Reflection Test" (CRT) as a simple measure of one type of cognitive ability—the ability or disposition to reflect on a question and resist reporting the first response that comes to mind. The author will show that CRT scores are predictive of the types of choices that feature prominently in tests of decision-making theories, like expected utility theory and prospect theory. Indeed, the relation is sometimes so strong that the preferences themselves effectively function as expressions of cognitive ability—an empirical fact begging for a theoretical explanation. The author examines the relation between CRT scores and two important decision-making characteristics: time preference and risk preference. The CRT scores are then compared with other measures of cognitive ability or cognitive "style." The CRT scores exhibit considerable difference between men and women and the article explores how this relates to sex differences in time and risk preferences. The final section addresses the interpretation of correlations between cognitive abilities and decision-making characteristics.

3,299 citations