منافع و صداقت اشتراکی در گزارش بودجه

This study uses two experiments to investigate the honesty of managers’ budget reports when the financial benefit resulting from budgetary slack is shared by the manager and other non-reporting employees. Drawing on moral disengagement theory, we predict that the shared interest in slack creation makes misreporting more self-justifiable to the manager and, therefore, leads to lower honesty. Consistent with our prediction, the results of our first experiment show that managers report less honestly when the benefit of slack is shared than when it is not shared, regardless of whether others are aware of the misreporting. Our second experiment investigates whether the preferences of the beneficiaries of the slack affect managers’ honesty. We predict that managers’ honesty will be improved when the beneficiaries of the slack have a known, higher-order preference for truthful reporting. Consistent with our prediction, the results show that managers report more honestly when other employees have a known preference for honesty than otherwise. The implications of our findings for management accounting research and practice are discussed.

Budgeting plays an important role in organizations for planning, coordinating activities, allocating resources and providing appropriate incentives (Covaleski, Evans, Luft, & Shields, 2003). Typically, lower-level managers have superior information about their subunit’s conditions, such as costs and productive capabilities. Due to this information asymmetry, upper management in the organization often relies on subunit managers to communicate such information during the budgeting process. This information is useful to the organization for improving the efficiency of resource allocation decisions (Antle & Fellingham, 1990) and the design of budget-based performance incentives (Shields & Shields, 1998). Subunit managers often submit budgets that include slack, defined as the “intentional underestimation of revenues and productive capabilities and/or overestimation of costs and resources required to complete a budgeted task” (Dunk & Nouri, 1998, p. 73). To the extent that budgetary slack results in unnecessary expropriation of resources by the subunit manager, it is not in the best interests of the overall organization.3 This study investigates how shared interests in budgetary slack affect the honesty of budget reports. Specifically, we investigate how the sharing of the benefits from budgetary slack between the subunit manager making the report and other non-reporting employees affects the honesty of such reports. Broadly speaking, benefits from slack can be obtained by reporting dishonestly during the budgeting process in two ways. First, costs can be overstated so that the subunit receives excess resources (Merchant, 1985), and the subunit benefits because excess resources may be consumed as perquisites and/or as leisure. Second, targets against which subunit performance will be evaluated can be understated (Fisher et al., 2002 and Young, 1985), and the subunit benefits because lower targets may result in higher performance-based pay and/or more leisure. Importantly, variation across organizational control systems, including incentive pay policies, is likely to affect the degree to which benefits from slack are shared between subunit managers and other employees. For example, the delegation of decision rights varies across organizations, suggesting that the ability of subunit managers to approve expenses that could be consumed as perquisites by other employees also varies. Control systems influence how leisure could be shared with other employees through mechanisms such as outsourcing, hiring excess workers or granting time off. Finally, organizations vary in terms of how deep into the hierarchy budget-based performance pay reaches. Organizations are increasingly using group-based incentive plans, defined as incentive plans in which “compensation varies as a function of performance achieved by a group of employees” (Hollensbe & Guthrie, 2000, p. 846). A distinguishing feature of such plans is that each group member has a share in any benefit resulting from the improvement of group outcomes (Bohlander & Snell, 2007), suggesting that, if subunit managers understate targets, resultant benefits would be shared with employees in their subunit. The first purpose of this paper is to investigate whether a shared interest in slack creation affects the honesty of managers’ budget reports. This question is important to management control scholars and practitioners because it provides insights for understanding when managers are more likely to include slack in their budgets, thereby informing when it may be beneficial to invest in control systems such as audits of budget reports. This question is also important because it potentially identifies when a control mechanism that is useful in one domain imposes a negative externality on a different domain. Specifically, if group-based incentive plans decrease the effectiveness of budgeting, it is important to understand this effect because it may change the optimal design of the overall management control system. That is, management may need to weigh this potential cost against the benefits of group-based incentives when designing the most effective management control system. The second purpose of this paper is to investigate how firms can mitigate the potential adverse effect of shared interest on honesty. Specifically, we investigate whether managers’ reporting behavior is influenced by other employees’ preferences regarding how budgets should be made. We conduct two laboratory experiments to address these issues. Experiment 1 examines managers’ reporting behavior when the benefit resulting from budgetary slack is shared. Moral disengagement theory (Bandura, 1990, Bandura, 1999 and Bandura, 2002) suggests that an important precondition for managers to act opportunistically is the ability to disengage moral responsibility from their action by self-justifying the action so as to make it compatible with moral standards. Therefore, we predict that, compared to settings in which misreporting only benefits the manager, a shared interest (i.e., the fact that misreporting also benefits others) provides more “legitimate” self-justification for misreporting and, as a result, leads to less honest reports. We also examine whether other employees’ awareness of the misreporting influences managers’ behavior in a setting where the awareness has no economic consequences. We predict that such awareness does not affect managers’ behavior when the benefit of slack is shared because the misreporting can be self-justified by shared interest. In contrast, we predict that such awareness increases honesty when the benefit of slack is not shared because the manager may be concerned about other employees’ impression about misreporting and such misreporting cannot be self-justified by shared interest. In Experiment 1, participants act as a division manager or an assistant to the manager. The division manager makes a budget report to request funding to finance the division’s production costs, whereas the assistant’s role is completely passive. We use this hierarchical arrangement, in which the manager has full authority for budget reporting, because it precludes potential confounding effects of “diffusion of responsibility” (Darley and Latane, 1968 and Mynatt and Sherman, 1975) or the assistant’s specific input (if allowed) on the manager’s reporting behavior. Two factors are manipulated: whether the benefit resulting from budgetary slack is shared with the assistant (yes versus no) and whether the assistant knows about the misreporting (yes versus no). Consistent with our prediction, manager–participants report significantly less honestly when the benefit of slack is shared than when it is not shared. Supplemental data suggest that this effect is not driven by managers’ concerns about payoff disparity. Also as predicted, the assistant’s awareness of misreporting does not affect managers’ behavior when the benefit of slack is shared. However, contrary to our expectation, such awareness also does not affect managers’ behavior when the benefit of slack is not shared. In light of the results of Experiment 1, we design a second experiment to investigate how firms can alleviate the unwanted consequences of shared interest on honesty. Drawing on elastic justification theory (Hsee, 1995 and Hsee, 1996), we predict that managers will be less able to rely on shared interest to self-justify misreporting if other employees have a higher-order preference for truthful reporting. In Experiment 2, we elicit the assistant’s non-binding preference and communicate it to the manager (i.e., reporting honestly versus inflating the budget to maximize wealth). We also include a baseline condition in which no preference is communicated. Consistent with our prediction, manager–participants who know that the assistant prefers truthful reporting report significantly more honestly than managers who know that the assistant prefers wealth-maximization or managers who do not know the assistant’s preference. Our findings have several implications for management accounting research and practice. Our study identifies ways in which control systems may create positive or negative externalities on one another. From a positive externality perspective, our findings suggest that controls limiting the discretion managers have over expenditures or practices that would allow them to share slack with subordinates reduce the need for controls over the budgeting process. From a negative externality perspective, our findings suggest that group-based performance incentives may decrease honesty in budgeting, thereby undermining the effectiveness of the budgeting process. This finding is especially important given that group-based incentive plans are widely used in organizations (DeMatteo et al., 1998 and Fisher et al., 2003) and are believed to have a positive effect on organizational outcomes (Hollensbe & Guthrie, 2000). If group-based incentives lower the effectiveness of budgeting, this effect should be included in management’s cost-benefit analyses as part of their endeavor to maximize the total effectiveness of management control systems. Our study also contributes to the stream of research that investigates how non-pecuniary preferences have a bearing on the effectiveness of management control (e.g., Christ, 2010, Coletti et al., 2005 and Hannan, 2005). Conventional economic theory predicts that, in our setting, whether the benefit of slack is shared should have no impact on the manager’s reporting behavior because the wealth-maximizing level of slack is unaffected by any shared interest. Our study extends existing research by showing that, to the extent that shared interest allows managers to readily self-justify misreporting, honesty in managerial reporting is likely to be undermined, with resultant negative effects on the firm’s decision quality and operating efficiency. As such, our study contributes to the growing stream of research documenting that firms can benefit by considering a broader range of preferences than assumed by conventional economic theory. Our study also provides insights into how firms can play an active role in mitigating the negative side effect resulting from shared interest in slack creation. In light of prior findings that individuals tend to comply with social norms (e.g., Cialdini, Kallgren, & Reno, 1991),4 accounting research suggests that firms can benefit from incorporating positive, ethical norms into management control systems (Merchant and Van der Stede, 2007 and Noreen, 1988). Our study highlights another incremental benefit of cultivating healthy, pro-social norms within the firm: it helps eliminate the potential moral buffer that managers may exploit to self-justify misbehavior. Otherwise, if the prevalent norm is materialistic or ambiguous, it could increase the justifiability of opportunistic behavior and, consequently, exacerbate such behavior. The remainder of the paper is organized as follows. The next section presents the theoretical framework, hypotheses, method, results, and supplemental analysis for Experiment 1. The third section does the same for Experiment 2. The fourth section discusses the findings and concludes the paper.

We conducted two experiments to investigate the effect of shared interest in slack creation on honesty in budget reporting. Drawing on moral disengagement theory, we argue that managers can self-justify their opportunistic misreporting when the benefits resulting from slack are shared, which in turn leads to less honesty in budget reporting. In our first experiment, we find that, consistent with our prediction, managers report less honestly when the benefit of misreporting is shared with another employee than when the benefit accrues solely to the manager. The result holds irrespective of whether the other employee observes the manager’s report (i.e., whether the other employee has direct knowledge of misreporting). Supplemental data analyses are consistent with the manager using shared interest to self-justify misreporting. This effect of shared interest on budgetary slack is detrimental to organizations because, in our setting, such slack lowers the efficiency of resource allocation and undermines organizational effectiveness. We design a second experiment to investigate whether the other employee’s preference (for honesty or wealth-maximization) affects the manager’s behavior in the presence of shared interest. We contend that when the other employee has a known preference for honest reporting, the manager cannot easily self-justify misreporting. In our second experiment, we measure the other employee’s preference and communicate that preference to the manager. We also include a control level in which no preference is communicated. We find that managers report more honestly when the other employee has a known preference for honesty. By comparison, we do not find any difference in honesty when the other employee has a known preference for wealth-maximization or an unknown preference. Hence, the manager’s default assumption appears to be that the other employee prefers wealth-maximization, at least to a greater extent than honesty. Our research findings are subject to several potential limitations. In our experiment, we use a manager–assistant hierarchical arrangement, in which the manager has complete authority for budget reporting, to preclude possible confounds such as diffusion of responsibility. Yet, this arrangement puts the assistant–participant in a disadvantageous position (i.e., receives a lower base salary). Although our supplemental data provide evidence that managers were unwilling to give up some of their own payoff to make the assistant better off, we cannot rule out the possibility that managers’ reports were influenced by their concerns for the disadvantaged assistant. Further research is needed to investigate whether our results generalize to settings in which slack is shared with employees who are not disadvantaged relative to the reporting manager. Providing the manager with complete authority also abstracts away from group settings where the firm may benefit by allowing subordinate input into the decision process or by group-based incentive plans. Future research could investigate how the potential costs indentified in our study interact with these types of benefits from group settings. In addition, our failure to find support for H2a could be due to weak manipulation of the assistant’s awareness of misreporting (i.e., the manager and assistant do not know each other and the manager’s decision is anonymous). It would be interesting to investigate whether the results would differ if the familiarity or intimacy between the manager and assistant were increased. Finally, to eliminate fairness concerns, we do not have a participant act as the headquarters in the experiment. Such concerns, however, may be present in naturally occurring organizational settings. Future research can explore whether and how the manager’s general honesty level is influenced by fairness concerns. We note that, because this paper focuses on incremental effects observed by comparing between experimental conditions, our results are not likely to be systematically affected by the absence of a headquarters–participant. Despite these limitations, our findings have important implications for organizations regarding how to induce managers to truthfully communicate private information. If the organization can build an ethical organizational environment in which most employees live up to pro-social moral principles, it would be difficult for managers to justify potential opportunistic behavior and, thereby, would likely curb or mitigate such behavior. This type of preemptive control approach may be more effective than the traditional, compliance approach that ex post disciplines misbehavior ( Booth and Schultz, 2004, Murphy, 1993 and Paine, 1994). Otherwise, to the extent that employees’ moral values are generally low or ambivalent, the quality of information transmitted within the organization may be undermined.