Superiors’ Discretionary Bonus Pool Allocations when
Agents Face Disparate Performance Risk
Michael Majerczyk
Georgia State University
Tyler Thomas
University of Wisconsin-Madison
November 1, 2017
We would like to thank In Gyun Baek, Conner Blake, Jing Davis, Rachel Martin, Ella Mae Matsumura, Nathan Mecham, Kristian Mortenson, Steve Smith, Todd Thornock, Kimberly Walker, Dimitri Yatsenko, Flora Zhou as well as participants of the Managerial Accounting Experimental Research Brownbag, the Brigham Young University Accounting Research Symposium, The Georgia State University Critical Thinking Seminar, the 40th European Accounting Association Congress, the 2017 AAA Annual Meeting, and the 2017 AAA Midwest Region Meeting for their helpful comments. We gratefully acknowledge funding for this project from the BRITE Lab Research Grant and the Department of Accounting and Information Systems of the Wisconsin School of Business, as well as from the School of Accountancy at Georgia State University.
2
Superiors’ Discretionary Bonus Pool Allocations when
Agents Face Disparate Performance Risk
Abstract: We examine how decision makers adjust their allocation decisions in risky environments, and how organizational context influences these decisions. Superiors often have discretion over bonus allocations, allowing them to make adjustments for the risk faced by agents. It is thus important to understand how superiors use this discretion and the organizational factors, specifically decision responsibility and availability of information, that affect their allocation decisions. We provide evidence that when superiors are confronted by inequality in agents’ performance risk (i.e., uncertainty as to how effort translates to performance), they tend to sympathize through their bonus allocation decisions with those agents with greater performance risk. However, superiors are often charged with decisions, including initial resource allocations, which affect or create disparity in agents’ performance risk. In this case, we find that superiors are not more, and potentially less, sympathetic to the disadvantaged agents compared to the setting without additional decision responsibility. Further, when information is available to superiors that can justify the initial resource allocation decision, they will actually favor the advantaged agents through their bonus allocation. Our results provide insight into organizational factors that affect how superiors use bonus discretion and make allocation decisions, and we demonstrate that available information can have a significant biasing effect in how superiors make these decisions. These results have direct implications for organizations, given the pervasiveness of discretion in bonus allocation decisions and organizational concerns for fairness and the job satisfaction of employees and their effects on overall performance.
2
I. Introduction
Fairness and equity concerns have received widespread consideration across academic disciplines. In business and economics there has been a focus on distribution preferences examining how third-parties make allocation decisions. While the default in settings with non-implicated stakeholders is an equal-split payoff to all parties, prior research has found that context can affect such outcomes (Konow 2009). Our paper examines how organizational context, specifically decision responsibility and information, affects decision makers in a setting in which inequality exists in payoffs. Specifically, this study examines how decision makers adjust the welfare of others in risky environments, and how such decisions are affected by organizational context introduced into the setting.
Discretionary bonus allocations and subjective performance evaluation are prime settings where decision makers affect payoffs and welfare within organizations. Given superiors’ awareness of agents’ environmental factors including risk, superiors can be granted discretion over bonus allocations for their agents (Bol 2008; 2011). Often these bonus allocations can be tied to subjective performance evaluation that allows for consideration of non-contractible, or difficult to measure, factors to adjust for performance risk. We define performance risk as the uncertainty that a given amount of effort will translate to an anticipated level of performance. Performance risk can result from, among others, such factors as differences in regional and market conditions, the level or quality of training, and the amount or quality of capital and other resources at an agent’s disposal. Agents’ performance risk can also be affected by superiors, as superiors are often responsible for the initial allocation of resources that can fundamentally affect the disparate performance risk confronted by agents. Thus, as superiors can have discretion in both the planning and reward stages, it is important to understand how they respond to risk disparity and the corresponding effects on their judgments and decisions during these stages.
Organizations provide unique contexts within which decisions are made. First, within organizations superiors are often responsible for making decisions that affect both the risk that agents are exposed to and overall agent welfare. Second, organizations provide superiors with information. As a greater variety of key non-financial performance indicators are recorded and reported, superiors are likely to have even more information at their disposal upon which to base their resource and bonus/compensation decisions. However, it can also be the case that superiors possess or are provided with outdated or task-irrelevant information, which can influence their decisions. Bol and Smith (2011) find that, manipulating controllability, information in the form of objective performance measures can influence superiors’ subjective evaluation of a separate and independent task performed by the agent. The authors find high (low) performance on the independent objective measure significantly increases (decreases) the superiors’ rating on the subjective measure. Thus, it is important to understand how superiors use information and the corresponding, potentially biasing, effects on superiors’ decisions.
In our study, we evaluate how superiors’ bonus allocation decisions for their agents are affected by disparity in agent performance risk, and how the addition of organizational context, specifically additional superior decision responsibility and available information, further influence these decisions. In order to more clearly identify the effects of organizational context, we assume the superior is a non-implicated stakeholder, in that their decision does not directly or indirectly affect their own personal welfare. As in Guo, Libby and Liu (2017), this allows us to focus on superiors’ nonfinancial preferences and more directly evaluate their preferences for equity. Given that standard economic theory does not allow for predictions when the decision maker is a non-implicated stakeholder, we base our predictions on theories from the behavioral literature.
We first predict that if superiors (decision makers) are devoid of organizational context, then they will sympathize with agents (individuals) that are confronting greater downside risk in their payoff (disadvantaged agent). Comparatively, when superiors have responsibility for the risk confronted by agents, as with allocating initial resources to agents, this responsibility could make the risk more salient to superiors, leading to greater sympathy for the disadvantaged agent, as suggested by restorative justice theory. However, following consistency theory, these superiors might be less likely to favor the disadvantaged agent, as they will be somewhat influenced to show consistency in their actions by continuing to favor the advantaged agent, as the superior’s decision led to the advantage. Thus, we do not make a directional prediction for bonus allocation decisions by superiors with responsibility over the initial resource allocation.
We more concretely predict that additional context from available information within organizations will lead superiors to favor one agent and assign that agent lower risk. These superiors will be motivated to be consistent with this initial decision, and can justify consistent actions, as the initial decision was based on available information. Thus, with increased decision responsibility and available information (both common components of organizational context), superiors are predicted to not sympathize with the agent assigned greater downside risk.
To test our predictions, we conduct a laboratory experiment consisting of a lottery drawing in which participants are grouped into triads, with one participant acting as the superior and the other two as agents.[1] The superior is either responsible for the payoff allocation for their agents or for both the payoff allocation and the initial probability of their agents achieving the bonus. The assignment of initial probability operationalizes an initial resource allocation to agents, which would affect agents’ performance risk. Prior to the payoff allocation and lottery drawing, participants perform a trivia task, in which they answer 12 multiple choice questions and an opened-ended trivia question. Performance on this task provides information related to the agents for the superiors’ decisions.
Our results show that superiors without context (i.e., without additional decision responsibility or information) strongly sympathize with the disadvantaged agent over the advantaged agent in their payoff allocation. Once responsibility over the initial allocation designating agents’ performance risk is introduced, superiors continue to sympathize with the disadvantaged agent in their bonus allocation decisions. These bonus allocation decisions by the superior, however, show no signs of greater sympathy, contrary to a restorative justice explanation. Rather, the results provide some support for a consistency story, in which superiors are somewhat influenced to be consistent in their decision-making between their initial allocation decision and their final bonus allocation decision. Most notably, the presence of information has a significant effect on superiors’ behavior. First, when the information is available, it leads superiors to allocate greater resources (i.e., a higher likelihood of winning the lottery) to the agent reflected better by the information. Second, superiors allocate a greater payoff to these advantaged agents, demonstrating significantly less sympathy to the disadvantaged agent, and a bias toward the advantaged agent and consistent decision-making to justify previous decisions.
We also preform supplemental analyses. First, we show that when information is available to superiors without responsibility over resource allocation, the information has a strong effect on superiors’ payoff allocations, such that superiors favor the agent reflected more favorably by the information in their payoff allocations, regardless of the agents’ risk situation. Second, we provide evidence that when the superior has responsibility over resource allocation and information available when making bonus allocations, then there is a disparity in agents’ fairness and satisfaction perceptions, with disadvantaged agents believing their pay to be less fair while also being less satisfied with the process compared to advantaged agents.
Our study contributes to the literature on discretionary bonus allocations and subjective performance evaluation by focusing on superiors’ behavior when confronted with, or responsible for, disparate agent performance risk. Despite the large literature in this area, few studies in management accounting have focused on the behavior of superiors. Bol, Keune, Matsumura, and Shin (2010) suggest that further research examining superiors’ use of discretion concerning agent performance is needed. Our findings help address this paucity of research, and provide insight into decision making by superiors in the evaluation and bonus allocation process. Specifically, we examine how the introduction of greater organizational context can affect superior’s decision making and our study suggests a need to further examine how discretionary decisions related to fairness vary in organizational contexts.
We provide evidence that the presence of disparate risk among agents devoid of organizational context can lead superiors to sympathize with the disadvantaged agent through a greater payoff allocation. Further, we show that this sympathy toward the disadvantaged agent does not increase when the superior makes an initial allocation decision that affects the agents’ risk disparity, with some evidence that sympathy potentially decreases when the superior makes a decision that creates the disparate risk. This is in line with superiors having the desire to be consistent with their initial decisions and organizational context deteriorating natural sympathy. Ultimately, we show that in an environment with additional decision responsibility and information, more representative of the typical organizational context, superiors will actually make decisions that perpetuate inequality rather than correct for it, as their payoff allocations are biased more toward advantaged agents.
We also add to the accounting literature by showing how the availability of information can affect superiors’ decisions which subsequently have ex post motivational implications. While it is often assumed that additional information will improve decision-making, prior studies provide evidence that this is not always the case (Luft 2009; Ramalingam 2012; Luft, Shields, and Thomas 2016). Strikingly, we demonstrate that information (even without other responsibilities) significantly drives superiors’ allocation decisions, biasing them in accordance with the information. This becomes more noteworthy when considered in light of the findings of Maas, van Rinsum, and Towry (2011) that superiors will often seek out information to make decisions, suggesting that more information is likely to be consciously acquired over time.
Our results are also important in adding insight to the potential effects on agents. Agents might be unsure ex ante about their superiors’ intentions when they have discretion over evaluations or bonus allocations, but following an allocation decision by the superior, the agent obtains new information, which can potentially affect their beliefs about the superior and their motivation ex post. We show that superiors having resource allocation responsibility and available information leads to disparity in agents’ perceptions of fairness and satisfaction, such that disadvantaged agents believe they are less fairly paid and are significantly less satisfied with the bonus allocation process compared to advantaged agents. This disparity in agents’ fairness and satisfaction perceptions could lead to subsequent adverse motivational effects and remedial actions.
In the next section we develop our formal hypotheses related to the research question. Section III outlines the experimental method used. Section IV provides the empirical results and Section V concludes.
II. Hypotheses Development
When considering general decision makers, superiors in organizations are in a unique position because, as a product of their role, many of their decisions (e.g., resource and bonus allocations) directly impact the welfare of their agents, even when not necessarily carrying direct personal stakes for the superior themselves. In fact, Rohde and Rohde (2014) specifically identify managers (superiors) making allocation decisions as an example of an impartial spectator. In a laboratory setting, Aguiar, Becker, and Miller (2013) find that impartial spectators give less biased distributions compared to implicated stakeholders, suggesting that impartial spectators exhibit different behavior than implicated stakeholders that heretofore have been the primary focus of the distribution literature in economics. The findings in Aguiar et al. (2013) and suggestion of Rohde and Rohde (2014) propose a need to understand impartial spectator behavior in organizations and how organizations themselves may affect decision-making behavior.