Why don't people lie? Negative affect intensity and preferences for honesty in budgetary reporting

https://doi.org/10.1016/j.mar.2018.05.001Get rights and content

Abstract

Budgets are instrumental in management control systems but are prone to gaming behavior that creates slack and limits the effectiveness of budgets. Research suggests, however, that subordinates have preferences for adhering to a social norm of honesty that limits slack in their budgetary reporting. As such, an increased understanding of subordinates’ preferences for honesty can improve participative budgeting systems. We develop and test theory that increases our understanding of the drivers of preferences for honesty. We test the theory that preferences for honesty originate from an individual’s desire to avoid negative affect from violating social norms. Further, individuals systematically differ in the intensity with which they experience their negative affective reactions. Those with higher levels of this intensity (negative affect intensity, NAI), experience more negative affect and disutility from violating a norm of honesty. Thus, NAI is predictive of subordinates’ preference for honesty. Experimental results support our theory. Budgetary slack is constrained by preferences for honesty and NAI increases preferences for honesty. As such, preferences for honesty are a stronger informal control for subordinates with higher NAI. We discuss the implications of our theory for contract design and job assignment.

Introduction

Participative budgeting can improve firms’ planning and control if subordinates convey some of their private information (Libby and Lindsay, 2010). However, subordinates benefit from misreporting. Research on misreporting in participative budgeting suggests that subordinate preferences for adhering to an honesty norm (“preferences for honesty”)1 reduce budgetary slack and firms can benefit from understanding these preferences for honesty (Demski and Feltham, 1978; Mittendorf, 2006; Rankin et al., 2008; Douthit and Stevens, 2015). Our theory increases our understanding of why preferences for honesty exist and predicts which subordinates have stronger preferences for honesty. Specifically, we test whether the trait intensity with which one experiences negative affective reactions predicts their preferences for honesty.

Subordinates trade off their preferences for honesty and wealth when reporting (Luft, 1997; Evans et al., 2001; Brüggen and Luft, 2011) and individuals differ in the strength of their preferences for honesty (Murphy, 1993). We argue that preferences for honesty arise from negative affect. Subordinates create slack by misreporting, but this diverges from a social norm of honesty. Diverging from a social norm triggers a negative affective reaction (Damasio, 2007). Individuals are negative affect-averse and anticipate affect when making decisions (Rivis et al., 2009). The desire to avoid expected negative affect from violating a norm of honesty causes subordinates to exhibit preferences for honesty. As such, the more intensely one experiences their negative affective reactions (negative affect intensity, “NAI”), the stronger their preferences for honesty. Importantly, individuals possess a stable trait level of NAI (Larsen, 2009).2 Individuals with higher NAI experience greater negative affect from diverging from a norm of honesty. In turn, this increases the utility from adhering to a norm of honesty, holding preferences for wealth constant, and increases preferences for honesty. Thus, we predict that NAI is a key driver of subordinate preferences for honesty.

Several factors could reduce the influence of NAI on preferences for honesty. First, our theory assumes that honesty is a social norm in participative budgeting. If not, there can be no affective reaction to diverging from a norm of honesty and, thus, no preferences for honesty from NAI. While some argue that creating slack is a behavioral norm (Jensen, 2001) and preferences for honesty may not be salient (Rankin et al., 2008), the extant literature argues that preferences for honesty are important (Brown et al., 2009). Second, our theory assumes that individuals can anticipate the intensity of their affective reaction from diverging from a norm of honesty. However, individuals sometimes can err when predicting their affect intensity (Gilbert and Wilson, 2007) and such affective forecasting errors would potentially mute the effect of NAI on preferences for honesty. Finally, negative affect can have many sources, including losses of wealth (Kermer et al., 2006). To the extent that individuals view foregoing slack as a loss of wealth, it is possible that not misreporting could also cause negative affect and limit NAI’s predictive ability. However, this argument ignores the implicit property rights in the participative budgeting setting (Douthit and Majerczyk, 2018).3

We test our theory with an incentivized experiment where we measure NAI using the Affect Intensity Measure psychometric inventory (Larsen et al., 1986) prior to a participative budgeting task. In the budgeting task, the subordinate privately learns the actual cost and reports a budget to the firm. The subordinate keeps any slack, creating strong incentives to misreport. When the budget is a cost report, preferences for honesty restrict how much slack the subordinate creates. However, other intrinsic preferences, such as for fairness, also affect misreporting (Salterio and Webb, 2006; Brown et al., 2009), making slack a noisy proxy for the effect of preferences for honesty. Thus, we capture the incremental effect of preferences for honesty by manipulating whether the budget communication requires a factual assertion of the cost or does not require a factual assertion (offers a share of profits). While residual motivations operate in both settings, slack creation requires an explicit false assertion of fact when the budget requires a factual assertion but not when it takes the form of a profit offer. Thus, as in prior research, the difference in slack between treatments captures the effect of preferences for honesty incremental to other concerns (Rankin et al., 2008; Douthit and Stevens, 2015).

Results support our theory. A subordinate’s NAI is predictive of their preferences for honesty. Relatedly, in the setting where preferences for honesty are relevant, a subordinate’s NAI increases their self-reported desire to be honest. Further, we use our theory of why preferences for honesty exist to examine the effectiveness of preferences for honesty as an informal control. For subordinates with high (low) NAI, preferences for honesty are an effective (ineffective) control that yields higher (lower) firm profit than that expected from a formal hurdle rate (Antle and Eppen, 1985). As a whole, our results suggest that avoiding negative affect is the genesis of preferences for honesty and NAI can therefore predict the strength of preferences for honesty. Our results suggest that understanding the relative NAI of a subordinate population is important to firms when deciding whether to rely on formal or informal controls and supports claims that understanding preferences for honesty can improve contracting (e.g., Mittendorf, 2006).

Our study makes several contributions. We peer into the “black box” of why subordinates have preferences for honesty by investigating one source of such preferences – to avoid negative affect from norm divergence – capitalizing on a strength of experiments (Luft, 2016). We provide a strong test of our theory by using an ex ante measure of NAI to predict preferences for honesty. We also extend trade-off models used to explain slack creation (e.g., Luft, 1997) by examining a factor that explains why preferences for honesty exist and predicts their magnitude. Further, we extend a pair of financial misreporting studies which argue that anticipated negative affect drives preferences for honesty (Murphy, 2012; Mayhew and Murphy, 2014). They find that participants experience “residual negative affect following an unethical act (p. 424),” which demonstrates that dishonesty can lead to negative affect.4 We extend this research by directly testing the causal mechanism theorized to drive preferences for honesty, answering a call in Mayhew and Murphy (p. 439). We measure individual preferences for honesty and test the theoretical link between anticipated negative affect and ethical behavior.

We extend theory by including behavioral factors that predict deviations from traditional theory. Our finding that profit is higher under a contract relying on preferences for honesty than under a formal hurdle for high NAI subordinates supports claims that firms can improve contracts by understanding and incorporating preferences for social norms (Mittendorf, 2006; Stevens and Thevaranjan, 2010). This result also suggests that management control systems that encourage honesty in lieu of control systems that restrict dishonesty may be more valuable for high NAI subordinates (Salterio and Webb, 2006). We also contribute to research on the interplay of monetary incentives and affect. Farrell et al. (2014) find that monetary incentives mitigate, but do not remove the impact of affect when affect leads to dysfunctional behavior. We find that affect can support desirable behavior, through negative affect’s ability to drive preferences for honesty, despite incentives for opportunism. Thus, we extend research on how incentives, cognition, and affect combine (Lee and Allen, 2002).

Our study is valuable to firms as it refines our understanding of the psychological basis of preferences for honesty. Firms prefer employees with strong preferences for honesty, particularly in jobs with reporting functions. However, subordinates also likely see value to being perceived as having strong preferences for honesty (cynically, this increases chances for opportunism). Thus, subordinates with weak preferences for honesty may game questionnaires to mimic those with strong preferences for honesty. This behavior is less likely when firms assess preferences for honesty with NAI as this link is less obvious to subordinates. Further, understanding that preferences for honesty are caused by negative affective reactions to diverging from a norm of honesty may allow firms to increase preferences for honesty by heightening the salience of violations of social norms such that affective reactions to misreporting are more likely to occur. For example, firms could highlight the damage done to the company by bad budgets. Additionally, our results suggest that firms may benefit more from control systems that encourage honesty (sanction lying) with high (low) NAI subordinates, consistent with the discussion in Salterio and Webb (2006).

It is possible that the benefit of NAI we highlight is not sufficient to outweigh all the potential concerns associated with high NAI subordinates. However, high NAI individuals exist and firms can benefit from understanding both the strengths and weaknesses of these employees. Our study highlights one aspect where high NAI can be beneficial – increasing preferences for honesty. Our results suggest that firms can rely on preferences for honesty as an informal control more when subordinates with high NAI are in participative budgeting settings. The heterogeneity in individuals’ levels of NAI, and their subsequent preferences for honesty, helps explain some of the heterogeneity of budgeting procedures in practice and suggests that firms with relatively high (low) NAI subordinates will be more likely to rely on less (more) formal budget controls.

Section snippets

Participative budgeting and preferences for honesty

Subordinates often possess private information due to their proximity to operations. Firms can improve budgets by bringing subordinates into the budgeting process if subordinates accurately convey some of their private information. However, the incentives of subordinates and superiors diverge. Subordinates want budgets with considerable slack, while superiors want accurate budgets. Although traditional theory suggests that subordinates will maximize slack, there is a large body of research on

Experimental design

We conducted a 1 × 2 experiment that manipulated the mode of budget communication (Factual Assertion vs. No Factual Assertion) between-subjects. Participants were finance majors from a single class at a U.S. university. There were four sets of sessions, ranging from 10 to 28 participants. Each set of sessions consisted of two separate sessions that took place one week apart and lasted 90 min total. Participants earned one point of extra credit for the first session and four more points if they

Attention check to affect intensity measure

The AIM is designed to measure how strongly individuals experience affective reactions, given that a reaction has occurred. While the AIM is widely used and validated (Larsen, 2009), we conduct a test to provide assurance that our participants attended to the AIM.10 If participants attended to the AIM, then those with higher AIM scores should react more strongly to the first session’s affect-manipulation

Conclusion

We develop and experimentally test a model of why subordinates have preferences for honesty in participative budgeting reporting. Subordinates feel negative affect when diverging from a social norm of honesty, which causes a disutility. The size of this disutility is increasing in the intensity with which the affective reactions are felt. The desire to avoid this negative affect drives the desire to adhere to a norm of honesty and yields preferences for honesty. We measure participants’

Acknowledgments

We would like to thank Theresa Libby (Editor) and two anonymous reviewers for their helpful feedback. We would also like to thank Eric Chan, Randy Dumm, Jens Grober, Lynn Hannan, Mark Isaac, John Lightle, Weiming Liu, Jordan Lowe, Victor Maas, Michael Majerczyk, Don Moser, Doug Norton, Chuck Nyce, Bernhard Reichert, John Barry Ryan, Ashley Sauciuc, Timothy Shields, Doug Stevens, Laura Wang, and participants at the Florida State University Experimental Economics Group, the Florida State

References (71)

  • R. Larsen et al.

    Affect intensity as an individual difference characteristic: a review

    J. Res. Pers.

    (1987)
  • T. Libby et al.

    Beyond budgeting or budgeting reconsidered? A survey of North-American budgeting practice

    Manag. Account. Res.

    (2010)
  • J. Luft

    Cooperation and competition among employees: experimental evidence on the role of management control systems

    Manag. Account. Res.

    (2016)
  • P. Murphy

    Attitude, Machiavellianism and the rationalization of misreporting

    Account. Organ. Soc.

    (2012)
  • A. Newman

    An Investigation of how the informal communication of firm preferences influences managerial honesty

    Account. Organ. Soc.

    (2014)
  • M. Simonsson-Sarnecki et al.

    Factor structure and validity of the affect intensity measure in a Swedish sample

    Pers. Indiv. Differ.

    (2000)
  • D. Stevens et al.

    A moral solution to the moral hazard problem

    Account. Organ. Soc.

    (2010)
  • D. Williams

    Neuroticism and extraversion in different factors of the affect intensity measure

    Pers. Indiv. Differ.

    (1989)
  • A. Antle et al.

    Capital rationing and organizational slack in capital budgeting

    Manag. Sci.

    (1985)
  • K. Aquino et al.

    Overt anger inresponse to victimization: attributional style and organizational norms asmoderators

    J. Occup. Health Psychol.

    (2004)
  • S. Baiman et al.

    An experiment testing the behavioral equivalence of strategically equivalent employment contracts

    J. Account. Res.

    (1989)
  • A. Bandura

    Social cognitive theory of moral thought and actions

  • S. Barsade et al.

    Why does affect matter in organizations?

    Acad. Manag. Perspect.

    (2007)
  • C. Bicchieri

    The Grammar of Society: The Nature and Dynamics of Social Norms

    (2006)
  • J. Brown et al.

    Agency theory and participative budgeting experiments

    J. Manag. Account. Res.

    (2009)
  • M. Brunner et al.

    Sabotage in capital budgeting: the effects of control and honesty on investment decisions

    Eur. Account. Rev.

    (2018)
  • F. Bryant et al.

    Principal-components analysis and confirmatory factor analysis

  • E. Cardinaels et al.

    Think twice before going for incentives: social norms and the principal’s decision on compensation contracts

    J. Account. Res.

    (2015)
  • R. Cattell

    The Scientific Use of Factor Analysis

    (1978)
  • A. Damasio

    Neuroscience and ethics: intersections

    Am. J. Bioeth.

    (2007)
  • J. Demski et al.

    Economic incentives in budgetary control systems

    Account. Rev.

    (1978)
  • J. Douthit et al.

    Subordinate Perceptions of the Superior and Agency Costs: Theory and Evidence

    (2018)
  • J. Douthit et al.

    The robustness of honesty effects on budget proposals when the superior has rejection authority

    Account. Rev.

    (2015)
  • J. Elster

    Social norms and economic theory

    J. Econ. Perspect.

    (1989)
  • J. Evans et al.

    Honesty in managerial reporting

    Account. Rev.

    (2001)
  • Cited by (17)

    • Horseshoes, hand grenades, and regulatory enforcement: Close experience with potential sanctions and fraud deterrence

      2021, Organizational Behavior and Human Decision Processes
      Citation Excerpt :

      Individuals lose utility from negative affect and try to avoid it when making decisions (Joffily et al., 2014; Levine, Lench, Kaplan, & Safer, 2012). The desire to avoid negative affect is important for our study as unethical behavior, such as fraud, causes negative affect (Blay, Douthit, & Fulmer, 2019; Damasio, 2007; Ghorbani, Liao, Çayköylü, & Chand, 2013; Mayhew & Murphy, 2014). Further, the punishment or sanction of self-interested behavior can cause negative affect (de Kwaadsteniet et al., 2013; Joffily et al., 2014; Kermer, Driver-Linn, Wilson, & Gilbert, 2006; Koch, 2014).

    • Bargaining power as moderator of the “delay costs effect” in supply chain negotiations

      2021, Management Accounting Research
      Citation Excerpt :

      This variable takes a value of 1 if the seller’s initial offer is 120 (the price that split the bargaining pie equally), and zero otherwise. Finally, given that experimental sessions occurred at diff ;erent times/days, and individuals self-selected into their sessions, our models include fixed session eff ;ects (i.e., Fréchette, 2012; Blay et al., 2019). Table 1 indicates that the average initial bargaining gap in our sample is 5.28 euros, with a range that goes from 0 to 36 euros.

    • Subordinate perceptions of the superior and agency costs: Theory and evidence

      2019, Accounting, Organizations and Society
      Citation Excerpt :

      Our result that property rights norms, from role legitimacy, affect behavior is novel compared to research on property rights in peer bargaining and highlights the importance of our examination of property rights in a hierarchical agency setting. The dependent variables of these peer bargaining studies focus exclusively on fairness, whereas honesty is a major factor in budgetary reporting (Blay, Douthit, & Fulmer, 2019; Douthit & Stevens, 2015; Rankin et al., 2008). It is not obvious that fairness-based results directly transfer into ethical settings (Hoffman & Spitzer, 1985, p. 262), especially as fairness and honesty norms can interact (Douthit & Stevens, 2015).

    View all citing articles on Scopus
    View full text