Discretionary mechanisms and cooperation in hierarchies: An experimental study

https://doi.org/10.1016/j.joep.2019.102193Get rights and content

Highlights

  • We study the effect of managerial discretion over inputs and outputs on cooperation.

  • We gather experimental data from 368 subjects playing a hierarchical public goods game.

  • Different discretionary mechanisms have different effects on cooperation and managerial opportunism.

  • We find that managers nudge workers into making higher contributions in certain settings.

  • Discretionary mechanisms are therefore critical for the design of hierarchy.

Abstract

This paper experimentally investigates the effects of managerial discretion over organizational inputs and –outputs on cooperation levels in a novel hierarchical public goods game. We observe treatment differences suggesting that while the introduction of hierarchy improves contribution levels vis-à-vis a baseline without hierarchy, certain combinations of discretionary contribution (the extent to which managers can contribute) and discretionary rewards (the type of share accruing to the manager) are significantly more efficacious than others. We further detect evidence in certain treatments of managers engaging in “strategic luring” i.e. nudging workers into contributing more by sacrificing their own returns in early rounds, only to expropriate the value of the public good in later stages of the game. Discretionary mechanisms may thereby constrain the scope of managerial opportunism.

Introduction

The installation of managers as a means of fostering worker cooperation is a common feature of hierarchical organizations (Liu et al., 2012, March and Simon, 1958). Managers are ordinarily granted discretionary rights over inputs (i.e. the allocation of resources) and outputs (i.e. the distribution of benefits) in order to stimulate the cooperation of workers, who are otherwise inherently predisposed to shirk in joint production settings (Alchian and Demsetz, 1972, Gill et al., 2013). As such, managerial discretion may promote the greater good of the organization (Van Vugt, Jepson, Hart, & De Cremer, 2004), and ultimately, its residual claimants (Drouvelis, Nosenzo, & Sefton, 2017).

However, the allocation of such discretionary rights creates opportunities both for managerial shirking on the input side (by choosing to contribute incommensurably to group production) and/or misappropriation on the output side (by disproportionately remunerating themselves at the expense of others). The prospect of self-serving managers has indeed long been recognized by principal-agent theorists (Jensen & Meckling, 1976) and organizational scholars concerned with the ramifications of opportunistic leadership on worker morale and performance outcomes (Einarsen et al., 2007, Liu et al., 2012). The effects of managers advancing their own interests at the expense of their subordinates is increasing coming under the purview of experimentalists in both the economic- and social psychological literatures (e.g. Balliet et al., 2011, Drouvelis et al., 2017). Preliminary experimental studies suggest that endowing delegated managers with such discretionary rights would appear to provoke resentment, incite disobedience, and so forth (Cappelen et al., 2016, Nikiforakis et al., 2014), which not only undermines worker cooperative propensities, but ultimately sabotages the interests of the organization itself, and thereby, its residual claimant shareholders (Einarsen et al., 2007). This raises the question of how to design discretionary mechanisms in hierarchical organizations in order to circumscribe the scope of managerial self-seeking behavior vis-à-vis their subordinates.

While the topic is meriting increased interest among experimental economists, prior studies have either focused solely on organizational inputs (e.g. Falk and Kosfeld, 2006, Harrell and Simpson, 2015) or -outputs (e.g. Drouvelis et al., 2017, Van der Heijden et al., 2009, Van Leeuwen et al., 2018, Wit and Wilke, 1988). Accordingly, prior experimental designs do not necessarily reflect real-world hierarchical settings where managers are often entrusted with discretionary powers over both inputs and outputs in an effort to boost team output. The aim of this study is therefore to design an experiment which allows us to study these discretionary managerial rights on cooperation. We do so by combining analysis that tests predictions, based on the partial preliminary results of prior research, with an exploratory analysis that examines in greater detail the role of discretionary mechanisms.

To study the effect of managerial discretion on rank-and-file group member cooperation, we devised a controlled laboratory experiment, the novel “hierarchical public goods game” (HPGG) as an extension to the canonical public goods game (PGG).1 In effect, the public good represents the boundaries of the organization through which inputs and outputs flow. The HPGG modifies the PGG by incorporating two hierarchical parameters: First, “discretionary contribution” which can be either “present” (i.e. managers2 are allowed to contribute to the public good) or “absent” (i.e. managers are prevented from contributing), which allows us to examine workers reactions to a situation in which the manager can benefit from the proceeds of the public good, without contributing to it. Second, “discretionary return” which can either be “static” (i.e. managers either receive a fixed share of the proceeds) or “dynamic” (i.e. managers have full discretion over the distribution of the public good and can self-determine their own share), which enables us to investigate the effects of allowing managers the opportunity to disadvantageously self-remunerate. This results in four HPGG treatments: PD (present, dynamic), PS (present, static), AD (absent, dynamic), and AS (absent, static). The HPGG thus allows us to disentangle two central aspects of organizational hierarchy i.e. the discretionary mechanisms of contribution and return, on the ability of managers to elicit more cooperative outcomes.

The treatments are based on observations that reflect real-world organizational realities. The PD treatment can represent franchise agreements, where franchisees (effectively independent contractors) have high-powered incentives to work hard (to recoup lump-sum fees to the franchisor), yet the capacity for self-remuneration often has resulted in noncompliance with labor market practices, as manifest in, for example, docking hours and wage theft of workers who are typically low-paid, highly fragmented and poorly-organized (Adams, 2018, Ji and Weil, 2015). Archetypical organizations facing PS conditions could be entrepreneurial start-ups, where founders/owners characteristically work long hours (inputs of “sweat equity”), yet frequently exercise self-sacrifice (only extracting modest amounts out of the business) in order to have the surplus to hire talented workers (e.g. sales managers, software developers) necessary for new venture growth (Bhide, 1996). AD reflects a situation where managers may have the inclination to shirk and have the opportunity for self-remuneration. This characterizes family owner-manager firms. There is a body of literature (e.g. Dyer and Whetten, 2006, Lubatkin et al., 2005) which argues that, on the one hand, the prevalence of the “dark side of parental altruism” in family firms, where children are hired as managers (often during succession) without having the necessary skills or qualities, results in managerial free-riding and shirking in family owner-manager firms. On the other hand, family owner-manager firms, based on concentrated ownership where owner-manager residual claimants are less constrained in their actions (typically unaccountable to an independent board/shareholders), and thereby have unchallenged discretion over the use of the firḿs resources. Finally, the AS treatment encapsulates the situation of managers in public-sector organizations, whose salaries are typically fixed by the respective ministry and are not entitled to retain cost savings or revenue increases for personal use (Frant, 1996), and thereby may be more inclined to reduce own effort expended by hiring support staff, etc. (Khuntia and Suar, 2004, Van der Heijden et al., 2009).

The paper offers four findings. First, our experimental results show that the introduction of hierarchy clearly matters. Average contribution shares in all four HPGG treatments exceeded those of the (non-hierarchical) baseline, and indeed were even insensitive to typical end-game effects in three of the four treatments. Second, while there was considerably less variation between the various HPGG treatments, we show that certain combinations of discretionary contributions and -rewards are more efficacious than others in fostering cooperation. Specifically, discretionary contributions are more effective than discretionary returns at sustaining cooperation in endgame situations. Furthermore, AD was the worst-performing HPGG treatment, indicating the marked reticence of workers to contribute in situation where managers were afforded distributive discretion without contributing to the public good. Third, we find that managers in the dynamic treatments appear to engage in “strategic luring” in which they either self-sacrifice in earlier rounds to nudge workers into making higher contributions, only to expropriate the value of the public good in later rounds, or leverage their discretionary powers to misappropriate throughout the game. Finally, we show that managers’ opportunistic behavior, i.e. the way they shirk and disproportionately reward themselves by enjoying on average higher receiving shares than their subordinates, alters depending on the specific discretionary mechanisms with which the HPGG operates.

Our study contributes to the debate as to whether managerial discretion has a galvanising or detrimental effect on the willingness of workers to cooperate by examining the effect of affording managers discretionary rights over both inputs (discretionary contributions) and outputs (discretionary returns) on contribution levels in a multi-period HPGG, and elucidating how specific combinations of hierarchical discretionary mechanisms facilitate cooperation and alter the scope of managerial opportunism.

The remainder of this paper is structured as follows. In Section 2, we present the related literature, while in Section 3, we describe the HPGG experimental design and procedures. In Section 4, we present our experimental results, which are discussed in Section 5. Section 6 concludes.

Section snippets

Related literature

Over recent years, there has been a burgeoning number of experimental studies into the effect of leadership on group cooperation. The field is characterized by an ongoing controversy as to whether organization design that allocates certain discretionary rights to managers has beneficial or deleterious consequences for group-level cooperation. On the one hand, the findings of a number of experimental studies suggest that the introduction of leaders (altering the structure of a simultaneous PGG

Experimental design and procedures

We conduct a baseline PGG and four HPGG treatments. All are dynamic public good games

Experimental results

We structure the results section as follows. First, in Section 4.1, we report findings on the effect of hierarchy on average contribution shares of managers and workers over time across the various treatments. Second, we present results for accumulated earnings and receiving shares for both managers and workers temporally and by treatment (in Section 4.2).

For the analysis we use the following variables. Discretionary contributions and Discretionary returns are binary variables indicating the

Discussion

The findings of the present study strongly suggest that, irrespective of HPGG treatment, hierarchy is more capable of sustaining worker cooperation levels and ameliorating the characteristic decay in contributions than in the manager-less baseline condition (Andreoni, 1988, Cinyabuguma et al., 2005). Managers endowed with discretionary powers over organizational inputs and outputs are thereby able to lead workers to more efficient outcomes. These findings corroborate prior experimental findings

Concluding remarks

This study reports a laboratory experiment which manipulates managerial discretion over inputs and outputs to systematically study their effect on group member cooperation levels. We introduce a novel modified HPGG in order to disentangle the effects of hierarchy on contribution shares, receiving shares, and accumulated earnings of group members.

This paper presents three overarching findings. Firstly, the introduction of hierarchy through the implementation of discretionary mechanisms sustains

Acknowledgements

We are grateful for comments from participants at the 13th Nordic conference on Behavioural and Experimental Economics 2018 and the JOD Conference in Sandbjerg 2018. This research was funded by a grant issued by the Danish Research Council (FSE).

References (50)

  • W. Güth et al.

    Leading by example with and without exclusion power in voluntary contribution experiments

    Journal of Public Economics

    (2007)
  • S.D. Haigner et al.

    To lead or not to lead: Endogenous sequencing in public goods games

    Economics Letters

    (2010)
  • B.K. Jack et al.

    Leadership and the voluntary provision of public goods: Field evidence from Bolivia

    Journal of Public Economics

    (2015)
  • M.C. Jensen et al.

    Theory of the firm: Managerial behavior, agency costs and ownership structure

    Journal of Financial Economics

    (1976)
  • M. Keuschnigg et al.

    The dark side of leadership: An experiment on religious heterogeneity and cooperation in India

    Journal of Socio-Economics

    (2014)
  • N. Nikiforakis et al.

    Hierarchy, coercion, and exploitation: An experimental analysis

    Journal of Economic Behavior & Organization

    (2014)
  • E. Van der Heijden et al.

    Hierarchy and opportunism in teams

    Journal of Economic Behavior & Organization

    (2009)
  • M. Van Vugt et al.

    Autocratic leadership in social dilemmas: A threat to group stability

    Journal of Experimental Social Psychology

    (2004)
  • A. Wit et al.

    Subordinateś endorsement of an allocating leader in a commons dilemma: An equity theoretical approach

    Journal of Economic Psychology

    (1988)
  • R.J. Adams

    McDonald́s versus NLRB: The end of franchising, or an overdue restoration of countervailing power?

    Business and Society Review

    (2018)
  • A.A. Alchian et al.

    Production, information costs, and economic organization

    American Economic Review

    (1972)
  • A. Ambrus et al.

    Imperfect public monitoring with costly punishment: An experimental study

    American Economic Review

    (2012)
  • K. Aquino et al.

    The effects of resource distribution, voice, and decision framing on the provision of public goods

    Journal of Conflict Resolution

    (1992)
  • E. Arbak et al.

    Voluntary leadership: Selection and influence

    Social Choice and Welfare

    (2013)
  • D. Balliet et al.

    Reward, punishment, and cooperation: A meta-analysis

    Psychological Bulletin

    (2011)
  • Cited by (0)

    View full text