Earnings management: Who do managers consider and what is the relative importance of ethics?

Abstract

Using responses from Australian CFOs and CEOs to a case-based survey and interviews, we provide insights into managers’ earnings management (EM) decisions. Ethics has the greatest explanatory power for our participants’ EM assessments. Ethical concerns about EM deter EM, but surprisingly, ethical concerns about not managing earnings and missing market expectations motivate EM. The primary economic motivation for EM is shielding current shareholders from the short-term costs of missing market expectations. We find considerable heterogeneity regarding the extent to which CFOs and CEOs believe EM is lying. Finally, CFOs and CEOs believe that each are significantly involved in initiating EM.

JEL Classification:

M40, M41, M48

Keywords

Earnings management ethics real earnings management

1. Introduction

It is well documented that earnings management (EM) is pervasive and continues to occur (Burgstahler and Dichev, 1997; Chu et al., 2019; Graham et al., 2005; Hope and Wang, 2018; Jones, 1991; Leuz et al., 2003).¹ While the existing research has provided interesting and valuable insights into EM, the insights are limited because the research treats managers’ EM decision processes as a ‘black box’. Fundamental to identifying and designing effective and efficient solutions for reducing EM is understanding factors that underlie managers’ EM decisions and the relative importance of those factors.

EM is a decision undertaken by managers and as such our goal in this article is to better understand the decision and to begin to open the black box. We designed a case-based survey and interviews to address three broad issues that have not been adequately addressed in prior EM research. The first is the relative importance of managers’ perceptions of the following three factors on both their decision whether to manage earnings and their decision regarding the use of accounting versus operational adjustments: (1) costs and benefit of EM to themselves; (2) costs and benefits of EM to individual stakeholder groups (i.e. current shareholders, future shareholders, debtholders and other stakeholders such as employees and suppliers); and (3) ethics of EM. We label these three factors: manager self-interest, stakeholders and ethics. The second issue is the extent to which managers believe that EM is lying and whether these beliefs affect their EM decisions. Related to this, we also examine the extent to which managers’ beliefs that EM is lying explains the explanatory power of ethics on their EM decisions. The third issue is the relative importance of CEOs versus CFOs in initiating EM.

We find that a significant majority of managers would be prepared to manage earnings where there are incentives to meet market expectations as presented in our case-based study. When we examine which of the three factors explains the decision we find each of the following are individually associated with the EM decision: manager self-interest, current shareholders, future shareholders, debtholders, other stakeholders and ethics. However, when we reflect the natural EM setting where managers likely consider multiple factors simultaneously – and thus include all the three factors simultaneously in our explanatory model – only ethics and current shareholders continue to have significant explanatory power. Ethics explains 17% of the EM assessment and current shareholders 5%.

Our interview results are consistent with this. The majority of the interviewees believe that CFOs and CEOs are fully aware that they will bear costs if they miss market expectations, but – consistent with our empirical results – they do not believe that such costs ultimately drive most EM decisions, and instead, believe that concerns related to current shareholders drive most EM decisions.

We also find that the CFOs’ and CEOs’ EM decisions reflect a consideration of inter-temporal wealth transfers. A significant ongoing debate is the extent to which managers myopically focus on the short-term at the expense of the long-term (Elliot et al., 2011; Fuller and Jensen, 2010; Gigler et al., 2014). While we find that the EM decision is associated positively with the net benefit of EM to current shareholders, it is also associated negatively with the cost of EM to future shareholders. Thus, when deciding whether to manage earnings, CFOs and CEOs appear to trade-off the net benefits of EM to current shareholders against the cost of EM to future shareholders, with the net benefit of EM to current shareholders having approximately twice the explanatory power of the net cost of EM to future shareholders.

Finally, in the choice between using accounting adjustments versus operational adjustments for EM, we find that CFOs’ and CEOs’ perceptions about the ethics of the two approaches is substantially more important than are the relative economic costs of the two approaches. The stronger CFOs’ and CEOs’ beliefs that the use of accounting (real operational) adjustments is ethically questionable, the more likely their firms would use real operational (accounting) adjustments to manage earnings. From the interviews conducted, all believed that accounting adjustments are ethically questionable while operational adjustments are not. Additional analyses reveal that CFOs and CEOs’ primary concern with operational adjustments is negative performance effects, which includes (in decreasing order of concern) operating performance, employee morale and stock price. For accounting adjustments, concerns are more diverse (in decreasing order of concern): loss of reputation/trust (primarily their own, not the firm’s), reduced accounting flexibility, negative performance effects (with slightly greater concern about stock price than operating performance) and monitoring costs.

The results discussed above treat ethics as a uni-dimensional construct when it actually has many dimensions. This suggests that understanding CFOs’ and CEOs’ EM decisions requires an understanding of why ethical considerations matter. The secondary purpose of this article is to consider one possible reason: the extent to which the explanatory power of ethics on CFO’s and CEOs’ EM decisions is driven by their perceptions that EM is lying. We focus on lying for two reasons. First, in their definition of EM, Healy and Wahlen (1999: 368) assert that the intention underlying EM is to ‘mislead’, which means ‘to cause someone to believe that something is not true’ (Merriam-Webster, 2019). This suggests that EM could be perceived to be lying. Second, there is a recent stream of research in economics that suggests that individuals may perceive the act of lying to be immoral – and thus unethical – regardless of the consequences to themselves or others (Cai and Wang, 2006; Charness and Dufwenberg, 2006, 2010; Ellingsen and Johannesson, 2004; Erat and Gneezy, 2012; Gneezy, 2005; Hurkens and Kartik, 2009; Jiang, 2013; Lundquist et al., 2009; Sutter, 2009; Vanberg, 2008).²

We find that a substantial portion of the explanatory power of ethics in the CEOs’ and CFOs’ EM decision is attributable to them perceiving EM to be lying. Perceptions that EM is lying accounts for approximately 31% of the explanatory power of ethics in the CEOs’ and CFOs’ overall EM decision and approximately 71% of the explanatory power of ethics in their decision about whether to use operational or accounting adjustments to manage earnings. We also find that CFOs and CEOs without an accounting qualification are significantly more likely to perceive accounting adjustments as lying. Another question that is motivated by mixed archival research is the relative importance of CFOs versus CEOs in initiating EM (e.g. Chava and Purnanandam, 2010; Erickson et al., 2006; Feng et al., 2011; Friedman, 2014; Graham and Harvey, 2001; Jiang et al., 2010). We find that the majority of both groups perceive the EM decision as a ‘joint decision’ in both the survey responses and the insights from the interviews.

Our study makes a number of contributions to the literature. First, we contribute to the EM literature by providing insight into the black box of CFOs’ and CEOs’ EM decision process. Rather than simply identifying a situation where EM is expected to occur – based on assumptions about factors CFOs and CEOs will consider – and then observing whether EM occurs, we document factors that actually affect CFOs’ and CEOs’ EM decisions and the relative importance of those factors. Contrary to conventional wisdom, we find that manager self-interest is not an important determinant of EM. Instead, we find that perceptions about the extent to which EM is ethically questionable, and the economic consequences to stakeholders drive CFOs’ and CEOs’ EM decisions. Our results will be of interests to EM researchers as our insight into the ‘black-box’ of why managers engage in EM provides some context to address the significant concerns that the EM archival models are misspecified due to correlated omitted variables giving rise to spurious findings (Ball, 2013; Jackson, 2018; Kothari et al., 2005; McNichols, 2000).

Second, we contribute to the emerging literature on the role of ethical attributes in economic decision-making in general and the EM decision in particular. Prior research provides evidence that ethical considerations can mitigate EM in settings where managers have incentives to act opportunistically (e.g. Cohen et al., 2007; Greenfield et al., 2008; Johnson et al., 2012). Our results extend these prior studies by providing evidence about how CFOs and CEOs trade-off ethical considerations against their own self-interests and the interests of other stakeholders.

Third, our results provide insight into why ethics is important to the EM decision. We find that a significant portion of the ethical consideration for both EM decisions is due to perceptions that EM (in particular via accounting adjustments) is lying. To the best of our knowledge, our study is the first to examine both whether EM is perceived as lying and whether this perception affects the EM decision.³ Our results regarding lying are consistent with the recent growing body of research in economics that finds that individuals are less likely to choose an economic action if it involves a lie (Ellingsen and Johannesson, 2004; Gneezy, 2005; Hurkens and Kartik, 2009; Sanjiv and Gneezy, 2012). Our results suggest that to understand EM, we need a better understanding of when and why people view EM as lying.⁴ Our ethics and lying results also provide empirical evidence for Graham et al.’s (2005) speculation for their ‘surprising’ finding that CFOs would prefer to manage earnings by real operational EM rather than accounting EM. They speculate – but do not test – that their finding could be due to accounting adjustments being less preferable to operational adjustments because CFOs view the latter as more ethical (Bruns and Merchant, 1990). We find that ethics explains approximately 43% of CFOs’ and CEOs’ decision about whether to manage earnings via accounting adjustments or operational adjustments, and that approximately 71% of the explanatory power of ethics in this decision is attributable to CFOs’ and CEOs’ perceptions that EM is lying. This finding should be of interest to regulators and those charged with governance within organizations. It shows that attempts to reduce EM really requires some reframing of the discussion and also more education on the fact that EM is lying and/or unethical so that it is not rationalized as normal business practice.

Overall, these results provide evidence to support the fact that characteristics of CFOs’ and CEOs’ are relevant in whether companies will engage in EM, which is consistent with the expectations from the review by Libby et al. (2015) and also emerging archival research that explores this type of association (e.g. Buchholz et al., 2020; Ham et al., 2017).

The remainder of this article is organized as follows. In the next section, we present the background and research questions. In Section 3, we discuss our methodology; in Section 4, we present our results. Section 5 concludes the article.

2. Background and research questions

2.1. Prior archival earnings management research

There has been a large volume of archival research that has examined EM through the use of discretionary accruals models (Dechow et al., 2010; Jones, 1991) and to a lesser extent real earning management (Roychowdhury, 2006). However, this stream of research has two limitations. First, it has provided no insight into the ‘black-box’ of why managers engage in EM. Second, significant concerns have been raised over the reliability of the measures of EM and it has been suggested the models are misspecified due for example to correlated omitted variables giving rise to spurious findings (Ball, 2013; Jackson, 2018; Kothari et al., 2005; McNichols, 2000). To address both limitations, we survey CFOs and CEOs of listed Australian companies using a confidential case-based survey to measure both if they engage in EM and why.

2.2. Managers’ earnings management decisions

There is a large volume of archival, economics-based EM research (see, e.g. Dechow and Skinner, 2000; Fields et al., 2001; Healy and Wahlen, 1999). This research has tended to focus on two types of factors as motivating EM: (1) manager economic self-interest (e.g. incentive compensation) and (2) firm-related economic factors (e.g. stock price motivations and debt covenants) (Graham et al., 2005). The latter factors, however, can affect multiple parties, so it is not clear from prior research who managers consider when deciding whether to manage earnings.⁵

Consistent with Evans et al. (2015), as well as positive accounting theory (Watts and Zimmerman, 1986), we assume that managers’ EM decisions reflect a cost-benefit trade-off, with EM occurring when managers perceive that the benefits of EM exceed the costs of EM.⁶ Within this framework, we consider the role of three general factors: (1) managers’ expected costs and benefits of EM to themselves (i.e. manager self-interest), (2) managers’ expected costs and benefits of EM to the firms’ various stakeholder groups and (3) managers’ perceptions about the ethics of EM.

A central tenant of agency theory is that the objective function that governs managers’ decisions and actions reflects only the economic costs and benefits to themselves; that is, managers ‘act so as to maximize their own welfare’ (Watts and Zimmerman, 1986: 3). Consistent with this, the extent economic-based EM research typically assumes that EM is motivated by manager self-interest (e.g. Cheng and Warfield, 2005; Healy, 1985; Holthausen et al., 1995). This self-interest could affect managers’ decision to engage in EM, as well as their decision about the approach to use to manage earnings.

However, agency theory also views the firm as a ‘nexus of contracts’ (Watts and Zimmerman, 1986: 196), which suggests that EM has the potential to result in wealth transfers between different stakeholder groups and/or between the manager and different stakeholder groups. Could the welfare implications of EM on the various stakeholder groups influence managers’ EM decisions? If managers are not motivated solely by economic self-interest, the expected costs and benefits of EM to the firm’s stakeholders has the potential to affect managers’ EM decisions. Managers could make trade-offs across stakeholder groups (e.g. shareholders versus debtholders), as well as inter-temporal trade-offs (e.g. existing shareholders versus future shareholders).⁷ Consistent with Jensen (2001), we consider the following stakeholder groups: current shareholders, future shareholders, debtholders and other stakeholders (e.g. employees, customers and suppliers).⁸

The final factor we consider is ethics. We do so based on a stream of EM research in accounting that provides evidence that managers do not perceive all EM practices and motivations to be ethically equivalent (e.g. Bruns and Merchant, 1990; Fischer and Rosenzweig, 1995; Johnson et al., 2012; Kaplan, 2001; Merchant and Rockness, 1994). In addition, a number of experiments provide evidence that ethics mitigate EM behaviour in contexts where managers have self-interested economic incentives to act opportunistically. Using student participants, Greenfield et al. (2008) find that an individual’s ethical orientation mitigates the effect of personal benefits on EM behaviour. Similarly, in an experiment with managers, Cohen et al. (2007) find that concern for fairness mitigates self-interested behaviour in a managerial accounting EM situation. Beaudoin et al. (2015) find that perceptions of the ethicalness of EM affect the EM decision. Supporting the findings of these experiments, in an archival study, Dikolli et al. (2020) find a positive association between the managerial trait of integrity and accruals quality.

With respect to the decision whether to use real operational adjustments versus accounting adjustments to manage earnings, only two studies have examined how managers trade-off the two EM approaches (Cohen and Zarowin, 2010; Zang, 2012), but neither considered the role of ethics. However, the two EM approaches are quite different, suggesting that the nature and magnitude of their respective costs and benefits are also likely to differ, implying that managers may not perceive the two approaches as ethically equivalent. Consistent with this, Graham et al. (2005) speculate – but do not test – that the CFOs in their survey prefer using operational adjustments to manage earnings because the CFOs perceive operational adjustments to be more ethical than accounting adjustments.

This discussion suggests that each of the following factors has the potential to influence both a manager’s overall EM decision and his or her decision about the approach for managing earnings: (1) the perceived costs and benefits of EM to the manager (the manager); (2) the perceived costs and benefits of EM to current shareholders, future shareholders, debtholders and/or other stakeholders (the stakeholders); and (3) the extent to which the manager perceives EM to be ethically questionable (ethics). However, it is an open question regarding which of these factors, if any, would influence managers’ EM decisions when the factors are considered simultaneously, as would occur in natural EM situations. This open question yields our first two research questions:

Research Question 1 (RQ1). What is the relative importance of managerial self-interest, stakeholders and ethics in managers’ decision to manage earnings?

Research Question 2 (RQ2). What is the relative importance of managerial self-interest, stakeholders and ethics in managers’ choice of approach in managing earnings?

2.3. Perceptions that EM is lying and the explanatory power of ethics

The discussion in the prior section suggests that ethical considerations may be an important factor in managers’ EM decision process. Ethics is a broad, multi-dimensional construct. Identifying the appropriate regulatory or corporate governance approach for addressing EM requires understanding the dimensions that underlie the explanatory power of ethics on managers’ EM decisions. Our second goal in this article is to delve deeper into why ethical considerations affect managers’ EM decisions. We consider one dimension: managers’ perceptions that EM is lying. Specifically, we examine the extent to which the explanatory power of ethics on managers’ EM decisions is driven by their perceptions that EM is lying. We again consider both the overall EM decision and the decision about using real operational adjustments versus accounting adjustments to manage earnings.

We focus on lying for two reasons. First, Healy and Wahlen (1999: 368)’s definition of EM indicates that one intention underlying EM is to ‘mislead’, which means, ‘to cause someone to believe that something is not true’ (Merriam-Webster, 2019). This suggests that EM could be perceived to be lying.

Second, prior research provides evidence consistent with individuals deriving disutility from lying, even when lying is in their economic self-interest. For example, numerous studies have documented that managers do not create maximum budgetary slack even though it is in their economic best interest to do so (e.g. Chow et al., 1988; Evans et al., 2001; Fisher et al., 2006; Stevens, 2002), suggesting that honesty is important to some managers (Cardinaels, 2016; Evans et al., 2001; Rankin et al., 2008). Similarly, a recent and growing body of research in economics explicitly examines lying per se and the inclination of some individuals to tell the truth despite lying being in their economic best interest (Cai and Wang, 2006; Charness and Dufwenberg, 2006, 2010; Ellingsen and Johannesson, 2004; Erat and Gneezy, 2012; Gneezy, 2005; Hurkens and Kartik, 2009; Jiang, 2013; Lundquist et al., 2009; Sutter, 2009; Vanberg, 2008). This latter stream of research finds that individuals who refrain from lying do so not because of the consequences of lying, but because they view lying per se as unacceptable behaviour. That is, they view lying per se as morally objectionable and accordingly derive disutility from engaging in lying.

This discussion suggests the following research question:

Research Question 3 (RQ3). To what extent do managers believe that EM is lying and how does this affect their decision-making on EM?

2.4. Who instigates EM?

The final issue that we explore relates to examining the relative importance of CFOs versus CEOs in initiating EM. This is motivated by recent archival research on this topic (e.g. Chava and Purnanandam, 2010; Erickson et al., 2006; Feng et al., 2011; Friedman, 2014; Graham and Harvey, 2001; Jiang et al., 2010). There are competing arguments for which has greater influence and the empirical results have been mixed, potentially due to inherent difficulties of addressing this issue archivally (Dechow et al., 2010). We circumvent this problem by gathering information from our participants about the extent to which they believe CEOs versus CFOs initiate EM. The research question is stated as follows:

Research Question 4 (RQ4). To what extent do CFOs versus CEOs initiate EM?

3. Method

3.1. Case-based survey

3.1.1. Overview

We mailed the case-based survey⁹ to 1044 CFOs and 1180 CEOs of the largest Australian public companies.¹⁰ We received 236 responses, of which 225 are complete and thus useable. Our useable response rate of 10.1% approximates CFO response rates in prior studies (e.g. Graham et al., 2005; Graham and Harvey, 2001; Trahan and Gitman, 1995).¹¹ Of the 225 respondents, 62% (38%) are CFOs (CEOs). The average work experience is 23 years, including 4.8 years in their current positions. Furthermore, 65% have a professional accounting qualification and 92% are male.

The case describes a hypothetical company whose earnings, with 45 days left before year-end, are tracking 4% below market expectations.^12,13 The company has two options: (1) take no special action and under-perform market expectations by 4% or (2) try to improve reported performance by 4% via accounting and/or real operational adjustments. The case then asks participants to respond to a number of questions, which form our dependent and independent variables. All the variables are summarized in Appendix 1.

3.1.2. Dependent variables from case-based survey

The CFOs and CEOs first answered three questions that serve as our dependent variables. They first assessed the likelihood their firm – if in a situation similar to the case – would attempt to improve reported performance via accounting and/or operational adjustments, using an 11-point scale from 0% (‘certain not to happen’) to 100% (‘certain to happen’). Next, they answered a binary choice question about whether their firm would (1) take no special action or (2) try to improve reported performance via accounting and/or operational adjustments. Participants who selected the latter then assessed the extent their firm would rely on accounting versus operational adjustments, using an 11-point scale from 0% (‘rely solely on accounting adjustments’) to 50% (‘rely 50% on accounting adjustments and 50% on operational adjustments’) to 100% (‘rely solely on operational adjustments’). We label participant i’s three assessments as: EM_LIKELIHOOD_i, EM_DECISION_i and EM_APPROACH_i, respectively.

3.1.3. Independent variables from case-based survey

The remaining questions address items that potentially explain the CFOs and CEOs’ three EM assessments. First, using an 11-point scale from 0% (‘certain not to happen’) to 100% (‘certain to happen’), participants assessed the likelihood their firm could improve reported earnings by at least 4% using (1) accounting adjustments only (SUCCESS_ACEM_i), (2) operational adjustments only (SUCCESS_OPEM_i) and (3) a combination of the two (SUCCESS_COMBINE_i).

Second, using an 11-point scale from 0 (‘absolutely no costs’) to 100 (‘very large costs’), the CFOs and CEOs assessed the magnitude of the costs that selected groups would bear if the firm (1) did not manage earnings and missed market expectations, (2) managed earnings via accounting adjustments, or (3) managed earnings via operational adjustments. They assessed costs for the following groups: (1) CFO and CEO, (2) current shareholders, (3) future shareholders, (4) debtholders and (5) other stakeholders.¹⁴ For the no EM situation, we label participant i’s cost assessments as: MANAGER_NOEMCOST_i, CURSHR_NOEMCOST_i, FUTSHR_NOEMCOST_i, DEBT_NOEMCOST_i and OTHER_NOEMCOST_i. For example, CURSHR_NOEMCOST_i is participant i’s assessment of the magnitude of the costs current shareholders would bear if the firm did not manage earnings and missed market expectations. For the accounting (operational) EM situation, we label participant i’s cost assessments as: MANAGER_ACCOST_i, CURSHR_ACCOST_i, FUTSHR_ACCOST_i, DEBT_ACCOST_i, OTHER_ACCOST_i (MANAGER_OPCOST_i, CURSHR_OPCOST_i, FUTSHR_OPCOST_i, DEBT_OPCOST_i and OTHER_OPCOST_i), respectively. The CFOs and CEOs were also asked to list the two most important costs for each of the following: (1) take no action, (2) use accounting adjustments and (3) use operational adjustments.

Third, using an 11-point scale from 0 (‘not questionable at all’) to 100 (‘extremely questionable’), the CFOs and CEOs assessed how ethically questionable it would be to (1) do nothing and miss market expectations, (2) use accounting adjustments to avoid missing market expectations and (3) use operational adjustments to avoid missing market expectations. These three assessments are ETHICS_NOEM_i, ETHICS_ACEM_i and ETHICS_OPEM_i, respectively.

Finally, the CFOs and CEOs assessed the degree to which improving reported performance using accounting adjustments is lying and then made the same assessment for operational adjustments. The 11-point scale ranged from 0% to 100%. These two assessments are LYING_ACEM_i and LYING_OPEM_i, respectively.

3.2. Interviews

Subsequent to the case-based survey, we conducted individual, in-person interviews with seven CFOs and three CEOs, using a semi-structured interview protocol based on the structure and insights from the case-based survey. The interviewees knew their and their firms’ identities would be confidential. Their average work experience is greater than 30 years, and all were CFOs and/or CEOs for at least 15 years. Four of the 10 are female and six are retired.^15,16

4. Results

4.1. Descriptive statistics for dependent variables

Figure 1 presents histograms of our three dependent variables. Per Panel A, there is considerable variation across our participants for EM_LIKELIHOOD. Overall, the CFOs and CEOs believe there is a 54% chance their firms would manage earnings if facing a situation similar to the case. However, when forced to make a choice (see Panel B), 68% of the CFOs and CEOs believe their firms actually would manage earnings in this situation. For the CFOs and CEOs who believe their firms would manage earnings (see Panel C), there is a clear preference for operational adjustments. Our mean response of 74% with a preference for operational adjustments is similar to Graham et al.’s (2005) finding of 78% for US CFOs.

Figure 1.

Histograms of dependent variables. Panel A: Managers likelihood of undertaking EM (EM_LIKELIHOOD) on a scale from 0% Not Likely to 100% Likely. Panel B: Managers decision to undertake EM (EM_DECISION). Panel C: Managers choice of EM approach (EM_APPROACH) between accounting and operational on a scale from 100% Accounting to 100% Operational.

While obviously these participants come from a range of listed companies, we are not able to examine firm-specific factors as part of our analysis. The reason is that as part of the requirements of ethics approval for this study we were required to guarantee not only confidentiality but also anonymity.

4.2 Managers’ EM decisions

4.2.1. What factors affect managers’ decision to manage earnings?

Our first research question addresses the relative importance of different factors in managers’ decision on whether to manage earnings. In the survey, we measure managers’ perceptions on (1) likelihood that EM will deliver the desired earnings; (2) ethics of EM; (3) costs and benefits of EM to themselves; and (4) costs and benefits of EM to specific stakeholder groups. We first address this question using univariate tests. The results for *_NOEMCOST in Panel B of Table 1 are consistent with economic issues creating pressure to manage earnings. That is, CFOs and CEOs who believe their firms would manage earnings believe the following groups will incur significantly larger costs if the firm does not manage earnings and misses market expectations (largest p-value < 0.05, two-tail): themselves, current shareholders, future shareholders and other stakeholders. Per Panel A, the CFOs and CEOs believe that if their firm does not manage earnings and misses market expectations, they and current shareholders will bear the largest (and equal) costs. Surprisingly, ethical considerations appear to create pressure to manage earnings. CFOs and CEOs who believe their firms would manage earnings have significantly stronger beliefs that not managing earnings in this situation is ethically questionable (Panel B, 33.49 versus 6.39, p-value < 0.01, two-tail).

Table 1.

Descriptive statistics and univariate tests of factors affecting EM decision.

	Panel A: Descriptive statistics			Panel B: Univariate tests
	M	Median	SD	EM-DECISION		t-statistic
				Yes	No
				M	M
SUCCESS
SUCCESS_ACEM	38.10	30	33.48	40.28	33.53	1.49
SUCCESS_OPEM	51.43	50	28.03	59.46	34.20	6.37***
SUCCESS_COMBINE	61.82	70	32.59	68.15	48.24	4.01***
MANAGER
MANAGER_NOEMCOST	43.20	50	28.23	46.93	25.38	2.68***
MANAGER_ACCOST	38.21	30	33.77	38.10	38.45	−0.36
MANAGER_OPCOST	29.16	20	25.66	28.17	31.25	−1.07
CURRENT SHAREHOLDERS
CURSHR_NOEMCOST	43.47	50	26.36	48.37	33.06	3.74***
CURSHR_ACCOST	26.22	20	25.94	27.45	23.61	0.79
CURSHR_OPCOST	23.64	20	21.38	23.46	24.03	−0.22
FUTURE SHAREHOLDERS
FUTSHR_NOEMCOST	20.09	10	22.31	22.55	14.79	2.51**
FUTSHR_ACCOST	33.29	20	28.83	33.14	33.61	−0.01
FUTSHR_OPCOST	29.16	20	25.03	27.39	32.92	−1.66*
DEBTHOLDERS
DEBT_NOEMCOST	17.27	10	21.96	18.54	14.58	1.26
DEBT_ACCOST	22.50	10	25.22	21.38	24.86	−1.04
DEBT_OPCOST	17.19	10	19.03	15.86	20.00	−1.37
OTHER STAKEHOLDERS
OTHER_NOEMCOST	20.40	15	20.03	23.16	14.58	2.86***
OTHER_ACCOST	22.31	20	22.56	21.76	23.47	−0.55
OTHER_OPCOST	24.62	20	23.30	25.49	22.78	0.50
ETHICS
ETHICS_NOEM	24.78	10	32.90	33.49	6.39	5.74***
ETHICS_ACEM	67.01	75	28.98	65.07	71.11	−1.61
ETHICS_OPEM	33.53	30	28.72	24.54	52.50	−7.72***

This table presents the descriptive statistics for our independent variables. The variables are responses by 225 CEOs and CFOs to individual questions in a case-based survey. The variables reflect responses to questions on an 11-point scale that ranges from 0 to 100. Definitions of the variables are provided in Appendix 1. ***, ** and * indicate a significant coefficient at the 1%, 5% and 10%, respectively, alpha level or better based on a two-tailed test.

The results for SUCCESS are consistent with opportunity being an important determinant of EM. CFOs and CEOs who believe their firm would manage earnings have significantly stronger beliefs that operational adjustments – whether alone or in combination with accounting adjustments – will deliver the desired earnings (both p-value < 0.01, two-tail).

With respect to rationalizing EM, only two items are significant. The first is ethical considerations. CFOs and CEOs who believe their firms would manage earnings have significantly weaker beliefs that using operational adjustments is ethically questionable (24.54 vs 52.50, p < 0.01, two-tail). Interestingly, both groups of CFOs and CEOs have strong beliefs that accounting adjustments are ethically questionable (65.07 vs 71.11). The second significant item is the cost to future shareholders of operational adjustments. CFOs and CEOs who believe their firms would manage earnings believe operational adjustments will impose significantly lower costs on future shareholders (27.39 vs 32.92, p-value < 0.10, two-tail).

The univariate results, however, do not reflect reality, where managers presumably simultaneously consider a variety of factors when deciding whether to manage earnings. This suggests that the only way to understand the actual factors that affect managers’ EM decisions – and the relative importance of those factors – is to include all the factors in the analysis and simultaneously assess their effects. We do this using the ordinary least squares (OLS) regression in Equation (1). The dependent variable is EM_LIKELIHOOD_i, which is CFO or CEO i’s assessment regarding the likelihood that his or her firm would manage earnings if facing a similar situation¹⁷

\begin{array}{l} E M_L I K E L I H O O D_{i} = α + β_{1} S U C C E S S_{i} + β_{2} M A N A G E R_{i} + β_{3} C U R S H R_{i} \\ + β_{4} F U T S H R_{i} + β_{5} D E B T_{i} + β_{6} O T H E R_{i} + β_{7} E T H I C S_{i} + ε_{i} \end{array}

(1)

For expositional parsimony, we use summary constructs to refer to the independent variables, consistent with how they are categorized in Table 1. SUCCESS_i refers to CFO or CEO i’s three assessments of the likelihood of achieving the desired earnings using accounting adjustments only, operational adjustments only, or a combination of the two. Similarly, MANAGER_i refers to CFO or CEO i’s three assessments of the costs to CFOs and CEOs of no earnings management, accounting adjustments, and operational adjustments. Thus, the reported results for the MANAGER construct are the sum of these individual estimates. CURSHR_i, FUTSHR_i, DEBT_i and OTHER_i are similar to MANAGER_i, except that CFO or CEO i’s cost assessments are for current shareholders, future shareholders, debtholders and other stakeholders, respectively. Finally, ETHICS_i refers to CFO or CEO i’s three assessments of how ethically questionable each of the following is (1) not managing earnings and missing market expectations, (2) accounting adjustments and (3) operational adjustments.

Table 2 reports the results from estimating Equation (1). Panel A reports results at the aggregate construct level. For example, the MANAGER construct is the sum of individual adjusted R²s of MANAGER_NOEMCOST, MANAGER_ACCOST and MANAGER_OPCOST.¹⁸ Panel B reports results at the disaggregated individual level.¹⁹

Table 2.

Likelihood assessment of earnings management.

$E M_L I K E L I H O O D_{i} = α + β_{1} S U C C E S S_{i} + β_{2} M A N A G E R_{i} + β_{3} C U R S H R_{i} + β_{4} F U T S H R_{i} + β_{5} D E B T_{i} + β_{6} O T H E R_{i} + β_{7} E T H I C S_{i} + ε_{i}$
	Panel A: Independent explanatory power aggregated into constructs.
	Baselinemodel		MANAGERmodel		CURSHRmodel		FUTSHRmodel		DEBTmodel		OTHERmodel		ETHICSmodel		Fullmodel
	R ²	F-statistic	R ²	F-statistic	R ²	F-statistic	R ²	F-statistic	R ²	F-statistic	R ²	F-statistic	R ²	F-statistic	R ²	F-statistic
SUCCESS	0.213	19.49***	0.164	14.88***	0.151	14.61***	0.206	18.04***	0.228	17.37***	0.171	13.90***	0.074	9.43***	0.060	5.43***
MANAGER			0.025	3.84***											0.003	1.38
STAKEHOLDER
CURSHR					0.079	8.50***									0.050	6.79***
FUTSHR							0.020	3.43***							0.004	1.43
DEBT									0.042	4.11***					0.018	1.42
OTHER											0.027	3.56***				1.01
ETHICS													0.208	14.48***	0.169	20.57***
Independent R² total	0.213		0.189		0.230		0.226		0.270		0.198		0.282		0.304
Correlated R²	0.000		0.048		0.062		0.007		−0.015		0.042		0.139		0.170
Total adjusted R²	0.213	19.49***	0.237	11.58***	0.292	15.09***	0.233	11.33***	0.255	12.56***	0.240	11.71***	0.421	25.67***	0.474	9.63***
N		225		225		225		225		225		225		225		225
			Panel B: Incremental effect of individual variables within constructs.
			MANAGERmodel		CURSHRmodel		FUTSHRmodel		DEBTmodel		OTHERmodel		ETHICSmodel		Fullmodel
		Coeff.	t-statistic	Coeff.	t-statistic	Coeff.	t-statistic	Coeff.	t-statistic	Coeff.	t-statistic	Coeff.	t-statistic	Coeff.		t-statistic
Intercept		20.770	3.66***	13.06	2.36**	22.971	4.11***	24.852	4.73***	22.330	4.12***	0.449	6.69***	41.657		5.92***
SUCCESS
SUCCESS_ACEM		−0.150	−1.78*	−0.122	−1.50	−0.178	2.09**	−0.162	1.91*	−0.137	1.61	−0.115	−1.53	−0.065		−0.88
SUCCESS_OPEM		0.310	3.43***	0.289	3.24***	0.293	3.12***	0.300	3.25***	0.288	3.16***	0.188	2.29**	0.189		2.27**
SUCCESS_COMBINE		0.317	3.23***	0.298	3.12***	0.383	3.73***	0.356	3.60***	0.314	3.17***	0.259	3.01***	0.154		1.70*
MANAGER
MANAGER_NOEMCOST		0.263	3.11***											−0.096		−1.12
MANAGER_ACCOSTS		−0.119	−1.65*											0.061		0.80
MANAGER_OPCOSTS		−0106	−1.08											−0.154		−1.56
CURRENT SHAREHOLDERS
CURSHR_NOEMCOST				0.437	5.00***									0.375		4.19***
CURSHR_ACCOSTS				−0.078	−0.89									0.102		1.02
CURSHR_OPCOSTS				−0.187	−1.72*									−0.264		−2.15*
FUTURE SHAREHOLDERS
FUTSHR_NOEMCOST						0.287	2.79***							0.014		0.13
FUTSHR_ACCOSTS						−0.145	−1.54							−0.194		−2.06*
FUTSHR_OPCOSTS						−0.045	−0.41							0.120		1.10
DEBTHOLDERS
DEBT_NOEMCOST								0.297	2.59**					−0.029		0.23
DEBT_ACCOSTS								−0.289	2.54**					−0.260		−1.91*
DEBT_OPCOSTS								−0.049	−0.33					0.212		1.32
OTHER STAKEHOLDERS
OTHER_NOEMCOST										0.349	2.81***			0.027		0.21
OTHER_ACCOSTS										−0.191	1.73*			0.080		0.57
OTHER_OPCOSTS										−0.001	−0.01			0.130		1.14
ETHICS
ETHICS_NOEM												0.302	5.20***	0.300		4.86***
ETHICS_ACEM												−0.113	−1.61	−0.100		−1.33
ETHICS_OPEM												−0.350	−4.75***	−0.358		−4.73***
Independent R² total			0.189		0.230		0.226		0.270		0.198		0.282			0.304
Correlated R²			0.048		0.062		0.007		−0.015		0.042		0.139			0.170
Total adjusted R²			0.237		0.292		0.233		0.255		0.240		0.421			0.474
F-statistic			11.58***		15.09***		11.33***		12.56***		11.71***		25.67***			9.63***
N			225		225		225		225		225		225			225

Definitions of the specific variables are provided in Appendix 1. ***, ** and * indicate a significant coefficient at the 1%, 5% and 10%, respectively, alpha level or better based on a two-tailed test. The ‘Independent R² total’ row reports the sum of the independent explanatory powers for all independent variables in the model while the ‘Correlated R²’ row reports the sum of the explanatory powers related to the correlations between all the independent variables in the model. The lines are just used to signify a sub-total.

In Panel A, the baseline model includes only SUCCESS because believing that EM will achieve the desired earnings is presumably a necessary condition for managing earnings. Consistent with this, the R² of SUCCESS is positive and significant (F-statistic = 19.49; p-value < 0.01, two-tail). Each of the next six models adds one construct to the baseline model in order to provide insight into the univariate importance of each group. The MANAGER, CURSHR, FUTSHR, DEBT, OTHER, and ETHICS constructs, along with SUCCESS, each have incremental power in explaining EM_LIKELIHOOD (smallest F-statistic = 3.43; all p-values < 0.01, two-tail). Thus, each construct – when considered individually – has significant incremental explanatory power for our participants’ EM likelihood assessment. These results are generally consistent with the univariate results in Table 1.²⁰

The last model in Panel A is the full model and includes all seven constructs. In this more realistic multivariate setting – where all the constructs are simultaneously allowed to explain the EM assessment – ethics, current shareholders, and likelihood of success are all significant (smallest F-statistic = 5.43; p-value < 0.01, two-tail), but manager self-interest, future shareholders, debtholders, and other stakeholders are not (largest F-statistic = 1.43; p-value = 0.24, two-tail).

Turning to the full model in Panel B, consistent with the results in Panel A, none of the individual MANAGER variables are significant (largest statistic = −1.56). These results reaffirm that when potential explanatory factors are considered simultaneously, manager self-interest is not an important determinant of the EM decision.

The full model results in Panel B also highlight two trade-offs in EM decisions. The first is ethical concerns about not managing earnings and missing market expectations versus ethical concerns about using operational adjustments to manage earnings. The former increases the likelihood of EM (ETHICS_NOEM; t-statistic = 4.86; p-value < 0.01, two-tail) while the latter decreases the likelihood (ETHICS_OPEM; t-statistic = −4.73; p-value < 0.01, two-tail).

Prior research has focused on ethics as a mechanism for reducing EM (e.g. Beaudoin et al., 2015; Cohen et al., 2007; Greenfield et al., 2008). Our results suggest it can also be an important motivator of EM, as not managing earnings is viewed as ethically questionable. Consistent with this, nine of the 10 CFOs and CEOs we interviewed believe it would be ethically questionable to not manage earnings (or at least not to explore ways to improve reported earnings) in a situation similar to the one in the case. For example, one interviewee stated,

I suppose the question is does the CFO have an ethical responsibility to ensure the financial performance is maximized. And I suppose I’d say yes to that question in the sense that in many ways their role is – as custodians for – on behalf of the shareholders. And if – I suppose if you work on the basis that the shareholders have an expectation that management’s role is to enhance the value of their company then in that sense it would be unethical for CFOs, management not to make decisions that resulted in the financial performance being as strong as it possibly could be.

The second trade-off is the costs to current shareholders if earnings are not managed and market expectations are missed (CURSHR_NOEMCOST) versus (1) costs to current shareholders from operational adjustments (CURSHR_ OPCOST) and (2) costs to future shareholders and debtholders from accounting adjustments (FUTSHR_ACCOST and DEBT_ACCOST). The cost to current shareholders of no EM increases the likelihood of EM (t-statistic = 4.19; p-value < 0.01, two-tail) while the other three costs each decrease the likelihood (smallest t-statistic = −1.91; p-value < 0.10, two-tail). Comparing the coefficients for CURSHR_NOEMCOST and CURSHR_OPCOST, costs to current shareholders from not managing earnings has approximately 50% more explanatory power for the likelihood of EM than do costs to current shareholders from operational adjustments. Comparing the sum of the absolute values of the coefficients for CURSHR_NOEMCOST and CURSHR_OPEM to the coefficient for FUTSHR_ACCOSTS or DEBT_ACCOSTS, costs to current shareholders have at least 2.5 times more explanatory power than do costs to either future shareholders or debtholders.

Prior research documents that managers myopically manage for short-term performance because current shareholders focus on short-term performance (Bhojraj and Libby, 2005; Elliot et al., 2011; Gigler et al., 2014; Graham et al., 2005; Healy and Palepu, 1993). However, it is an open question whether managers believe that focusing on short-term performance (1) does not impose costs on future shareholders or (2) imposes costs on future shareholders, but less than the benefits to current shareholders. Our finding of an inter-temporal trade-off between current shareholders and future shareholders provides support for the latter perspective.

The CFO and CEO interviewees confirm the importance of current shareholders, with all 10 interviewees mentioning that current shareholders are the primary group considered in EM decisions. The interviewees also acknowledge that while CFOs and CEOs are fully aware of the economic incentives they have to manage earnings to meet market expectations, these incentives are ‘in the back of your mind’ rather than the primary motivating factor. One CFO even stated that she had removed herself from the EM decision-making process when the EM would have personally benefitted her.

4.2.2. What factors affect the choice of approach in managing earnings?

Our second research question addresses factors that affect the extent to which managers use accounting versus operational adjustments to manage earnings. We test this using the OLS regression in Equation (2), with the 153 CFOs and CEOs in our case-based survey who believe their firms would manage earnings. The dependent variable is EM_APPROACH_i, which is measured on a scale from 0 (‘rely solely on accounting adjustments’) to 50 (‘rely 50% on accounting adjustments and 50% on operational adjustments’) to 100 (‘rely solely on operational adjustments’). The independent variables are the same as those in Equation (1)

\begin{array}{l} E M_A P P R O A C H_{i} = α + β_{1} S U C C E S S_{i} + β_{2} M A N A G E R_{i} + \\ β_{3} C U R S H R_{i} + β_{4} F U T S H R_{i} + β_{5} D E B T_{i} + β_{6} O T H E R_{i} + β_{7} E T H I C S_{i} + ε_{i} \end{array}

(2)

Table 3 reports the results for estimating Equation (2). The baseline model includes the individual variables for SUCCESS and MANAGER. Each subsequent model adds the individual variables for the added construct. Contrary to the conceptual foundation of the extant economic-based EM research (e.g. Dechow and Skinner, 2000; Fields et al., 2001; Healy and Wahlen, 1999), in the full model, none of the variables for MANAGER, nor for any stakeholder group, are significant (largest t-statistic = 1.57; p-value = 0.12, two-tail).²¹ These results suggest that once CFOs and CEOs decide to manage earnings, the choice of EM approach is not driven by the economic costs of the two approaches to either themselves or any individual stakeholder group.

Table 3.

Earnings management approach.

$E M_A P P R O A C H_{i} = α + β_{1} S U C C E S S_{i} + β_{2} M A N A G E R_{i} + β_{3} C U R S H R_{i} + β_{4} F U T S H R_{i} + β_{5} D E B T_{i} + β_{6} O T H E R_{i} + β_{7} E T H I C S_{i} + ε_{i} .$
	MANAGERmodel		CURSHRmodel		FUTSHRmodel		DEBTmodel		OTHERmodel		ETHICSmodel		Fullmodel
	Coeff.	t-statistic	Coeff.	t-statistic	Coeff.	t-statistic	Coeff.	t-statistic	Coeff.	t-statistic	Coeff.	t-statistic	Coeff.	t-statistic
Intercept	0.672	3.66***	66.418	11.27***	65.888	11.76***	65.099	12.22***	64.631	11.27***	0.492	8.66***	47.723	8.22***
SUCCESS
SUCCESS_ACEM	−0.137	−2.50	−0.129	−2.34	−0.133	−2.41	−0.168	−3.18***	−0.140	−2.52	−0.066	−1.36	−0.091	−1.90*
SUCCESS_OPEM	0.337	4.97***	0.323	4.75***	0.327	4.70***	0.333	5.08***	0.360	5.17***	0.254	4.28***	0.267	4.44***
SUCCESS_COMBINE	−0.054	−0.82	−0.040	−0.60	−0.048	−0.70	0.010	0.15	−0.034	−0.52	−0.031	−0.56	−0.014	−0.23
MANAGER
MANAGER_NOEMCOST	−0.070	−1.18	−0.057	−0.86	−0.079	−1.28	−0.079	−1.35	−0.038	−0.57	−0.045	−0.80	−0.047	−0.74
MANAGER_ACCOSTS	0.129	2.43	0.069	1.14	0.080	1.40	0.080	1.48	0.079	1.30	0.031	0.63	−0.010	−0.17
MANAGER_OPCOSTS	−0.185	−2.60*	−0.150	−1.91	−0.162	−2.21	−0.152	−2.04	−0.174	−2.30	−0.122	−1.96	−0.075	−1.05
CURRENT SHAREHOLDERS
CURSHR_NOEMCOST			−0.026	−0.35									0.029	.045
CURSHR_ACCOSTS			0.0148	1.91									−0.066	0.90
CURSHR_OPCOSTS			−0.074	−0.87									−0.104	−1.17
FUTURE SHAREHOLDERS
FUTSHR_NOEMCOST					−0.037	−0.51							0.031	0.43
FUTSHR_ACCOSTS					0.139	2.04							0.082	1.21
FUTSHR_OPCOSTS					−0.062	−0.34							0.094	1.23
DEBTHOLDERS
DEBT_NOEMCOST							−0.084	−1.07					−0.066	−0.74
DEBT_ACCOSTS							0.287	3.26***					0.168	1.57
DEBT_OPCOSTS							−0.231	−2.06					−0.037	−0.30
OTHER STAKEHOLDERS
OTHER_NOEMCOST									−0.119	−1.23			−0.098	−1.06
OTHER_ACCOSTS									0.123	1.37			−0.069	−0.69
OTHER_OPCOSTS									0.020	0.220.492			−0.057	−0.69
ETHICS
ETHICS_NOEM											0.033	0.84	0.064	1.56
ETHICS_ACEM											0.361	7.01***	0.379	6.85***
ETHICS_OPEM											−0.212	−3.37***	−0.052	−3.81***
Total adjusted R²		0.24		0.25		0.25		0.32		0.24		0.45		0.49
F-statistic		8.49***		6.18***		6.21***		8.16***		5.95***		13.47***		7.25***
N		153		153		153		153		153		153		153

Definitions of the specific variables are provided in the Appendix 1. ***, ** and * indicate a significant coefficient at the 1%, 5% and 10%, respectively, alpha level or better based on a two-tailed test.

Two items drive our CFOs and CEOs’ assessments. First, stronger beliefs that operational (accounting) adjustments are ethically questionable decrease (increase) reliance on operational adjustments (ETHICS_OPEM; t-statistic = −3.81; p-value < 0.01, two-tail; ETHICS_ACEM; t-statistic = 6.85; p-value < 0.01, two-tail). Comparing the absolute values of the coefficients for ETHICS_ACEM and ETHICS_OPEM suggests that beliefs about the ethics of accounting adjustments explain seven times more of the EM approach decision than do beliefs about the ethics of operational adjustments. Second, stronger beliefs that operational (accounting) adjustments will be successful increase (decrease) reliance on operational adjustments (SUCCESS_OPEM; t-statistic = 4.44; p-value < 0.01, two-tail; SUCCESS_ACEM; t-statistic = −1.90; p-value < 0.10, two-tail).

Overall, these results suggest that the decision about the EM approach is independent of the economic costs of the two approaches to managers and individual stakeholder groups. Instead, this decision is driven primarily by CFOs and CEOs’ beliefs about (1) which approach is more ethically questionable and (2) the likelihood that real operational adjustments will deliver the desired reported earnings. More of our CFOs and CEOs believe that operational adjustments are both less unethical and more likely to be successful, resulting in a greater preference for operational EM adjustments. In discussing why the CFOs in their survey prefer operational adjustments, Graham et al. (2005) speculate – but do not test – that the reason is because CFOs have ethical concerns about using accounting adjustments. Our results provide empirical support for Graham et al.’s speculation.

The interviewees provide additional support for Graham et al.’s speculation, as all 10 believe that accounting adjustments are ethically questionable while operational adjustments are not. One CFO described operational adjustments as ‘just managing the business’. However, the ethics of accounting adjustments is more nuanced. Several interviewees believe that accounting adjustments that simply ‘correct’ overly conservative initial positions (e.g. ‘whether all of the provisions were appropriate’) are ethical because such adjustments simply reflect the firm’s underlying economic reality. The size of the accounting adjustment is also a consideration, as one interviewee noted, ‘Well it depends how far you push it’.

4.2.3 Additional analyses

The results in Tables 2 and 3 provide evidence about how the magnitude of costs borne by various parties affects the overall EM decision and the EM approach decision, respectively. What is the nature of these costs? Our CFO and CEO survey participants were asked via open-ended questions to list the two most important costs in each of the following situations: (1) not manage earnings and miss market expectations, (2) EM via accounting adjustments, and (3) EM via operational adjustments.^22,23

Panel A of Table 4 reports the distributional frequency of the costs for the first situation. These costs create pressure for EM. The rank order of the costs is ‘loss of value/share price’ (44.1% of responses); ‘reputation/loss of confidence’ (20.3%); ‘future financing costs’ (19.0%); ‘compensation’ (11.2%); and ‘market reaction costs’ (3.1%).

Table 4.

Costs related to managing earnings.

Panel A: Nature of costs from not engaging in EM and missing the earnings benchmark.
	Number		Percent of total
Loss of value/share price
Effect on themselves	2		0.7
Effect on current shareholders	24		8.2
Effect on future shareholders	4		1.3
Effect on other stakeholders	2		0.7
Effect on unspecified individual	16
Subtotal	48
No additional detail provided	82		27.8
Total loss of value/share price		130		44.1
Reputation/confidence
Effect on themselves	26		8.8
Effect on company	4		1.4
Unspecified	30		10.2
Total reputation/confidence		60		20.3
Increase in future financing costs
Debt violations	9		3.1
Dividends	13		4.4
Raising capital	19		6.4
Unspecified	15		5.1
Total increase in future financing costs		56		19.0
Compensation
For themselves	18		6.1
For employees	3		1.0
Unspecified	12		4.1
Total compensation		33		11.2
Costs of addressing market reaction
Direct costs	7		2.4
Indirect costs	2		0.7
Total addressing market reaction costs		9		3.1
Other		7		2.3
Total		295		100.0

Panel B: Nature of costs from accounting adjustments and operational adjustments.
	Accounting adjustments						Operational adjustments
	Number			Percentage			Number			Percentage
Loss of reputation/trust
Reputation
Themselves	9			3.2			1			0.4
Company	5			1.8			3			1.1
Unspecified	15			5.4			5			1.9
Total reputation		29			10.4			9			3.4
Trust
Themselves	11			4.0			6			2.3
Company	1			0.4			1			0.4
Unspecified	24			8.6			6			2.3
Total trust		36			13.0			13			5.0
Total loss of reputation/trust			65			23.4			22			8.4
Negative effect on performance
Reduced share price
Short-term effect	10			3.6			7			2.7
Long-term effect	9			3.2			9			3.4
Unspecified	14			5.0			7			2.7
Total reduced share price		33			11.8			23			8.8
Reduced operating performance
Short-term effect	0			0.0			29			11.1
Long-term effect	4			1.4			22			18.4
Unspecified	19			6.8			62			23.7
Total reduced operating performance		23			8.3			113			43.1
Reduced employee morale		0			0.0			24			9.2
Other negative performance effects		0			0.0			39			14.9
Total negative effects on performance			56			20.1			199			76.0
Reduced accounting flexibility			56			20.1			8			3.1
Monitoring costs
Auditors		22			7.9			0			0.0
Board of directors		2			0.7			0			0.0
Regulators		7			2.5			0			0.0
Institutional investors		3			1.1			0			0.0
Other		3			1.1			0			0.0
Total monitoring costs			37			13.3			0			0.0
Reduced quality of earnings			20			7.2			0			0.0
Ethics/management integrity
Ethics		2			0.7			0			0.0
Management integrity		7			2.5			5			1.9
Total ethics/management integrity			9			3.2			5			1.9
Time			8			2.9			4			1.5
Other			27			9.7			24			9.2
Total			278			100.0			262			100.0

The lines are just used to signify a sub-total.

Across the 48 CFOs and CEOs who provided more detail about decreased value/share price, current shareholders are mentioned the most as bearing this cost (24 of the 48 responses). For the 30 (21) CFOs and CEOs who provided more detail about reputation/confidence (compensation) costs, 26 (18) mentioned themselves as the bearer of this cost. Of the 41 CFOs and CEOs who provided more detail about increased financing costs, 19 are concerned about the effect on raising capital in the future, 13 about the effect on dividends, and nine about debt covenant violations. Finally, the cost of managing the market reaction to missing expectations includes both (1) direct costs, such as time spent managing analysts and other stakeholders, and (2) indirect costs such as management being distracted from core operating issues and additional time needed to identify and implement plans to correct performance.

Panel B reports the distributional frequency of our CFOs and CEOs’ beliefs about the most important costs of the two EM approaches. The three most common concerns for both types of adjustments are (1) negative performance effects, (2) loss of reputation and (3) reduced accounting flexibility. However, these three costs are not of equal concern across the two EM approaches. Concerns about negative performance effects are almost four times greater for operational adjustments than for accounting adjustments (76.0% versus 20.1%). Negative performance concerns from accounting adjustments are primarily related to stock price performance (11.8%). In contrast, negative performance concerns from operational adjustments are primarily related to operating performance (43.1%), with reduced employee morale also a substantial concern (9.2%). The interviews confirm operational adjustments raise concerns about performance and employee morale. The most common operational adjustment is staff reductions, mentioned by over half the interviewees.²⁴ However, there is concern about the longer-term adverse operational effects of staff reductions; one interviewee stated that staff reductions ‘might be a bit short sighted, and I suppose that might be the ethical dilemma’ while another stated, ‘you lose all that expertise’. Overall, the interviewees generally believe that operational adjustments of any type – not just staff reductions – will adversely affect employee morale.

For reputation effects, approximately three times as many CFOs and CEOs mention negative reputation effects for accounting adjustments than for operational adjustments (23.4% vs 8.4%). Similarly, approximately seven times as many mention reduced accounting flexibility for accounting adjustments than for operational adjustments (20.1% vs 3.1%).

Interestingly, contrary to the conventional wisdom that accounting adjustments give rise to monitoring costs (e.g. Zang, 2012), they are only the fourth-most common concern (13.3%). The CFOs and CEOs perceive auditors as the primary source of monitoring costs (7.9%).²⁵ Consistent with this, seven of the 10 interviewees mention auditors as an important stakeholder in EM situations, with most indicating that the auditor must be ‘comfortable’ with what is being done.²⁶

Only two survey participants mention the Board as a source of monitoring costs, suggesting that CFOs and CEOs do not perceive their directors as a deterrent to using accounting adjustments for EM. The interviews provide some insight into this. Nine of the 10 interviewees mentioned the Board and/or audit committee is an important stakeholder in EM situations. A representative comment is, ‘So it comes down to, you put forward a number of scenarios to potentially the Board who have to make that decision’. This suggests that the CFO and CEO interviewees do not view the Board as monitoring EM, but rather, the Board having some involvement in the EM decision. This raises the issue of whether Boards are appropriately executing their governance role with respect to EM.

4.3. To what extent do managers believe that EM is lying?

Our third research question addresses whether CFOs and CEOs believe EM is lying. Related to this, we also explore how this belief affects their EM assessments and how it explains the effect of ethics on their EM decisions. Figure 2 presents histograms of LYING_ACEM and LYING_OPEM. There is considerable heterogeneity among CFOs and CEOs for both types of adjustment, as responses range from 0% to 100% for both. However, there is a clear difference in beliefs regarding the two adjustments. The CFOs and CEOs’ average belief that EM is lying is 2.2 times greater for accounting adjustments than for operational adjustments (59.77% vs 27.43%; t-statistic = 8.15; p-value < 0.01, two-tail).

Figure 2.

Histogram of perceptions of the extent to which earnings management is lying. Panel A: Accounting adjustments. Panel B: Operational adjustments.

Consistent with this, six of the 10 CFO and CEO interviewees generally consider accounting adjustments to be lying versus only two for operational adjustments. However, several interviewees mention that the size of the adjustment can affect whether accounting adjustments are lies. As one interviewee noted, ‘it begs the question then, so at what point does it become an issue. If it’s $100,000, oh, that’s okay, if it’s $500,000 or if it’s a million, gee, is that getting too much, but so it’s a bit like the question, when is it okay to tell a lie?’ In addition, one of the four who believe that using accounting adjustments is not lying stated that using the flexibility allowed within accounting standards is ‘not lying, it’s just complying with the rules’.

We next explore whether CFO and CEOs’ beliefs that EM is lying affect their two EM assessments. We first perform (untabulated) univariate tests of LYING_ACEM and LYING_OPEM_i for those CFOs and CEOs who believe their firms would manage earnings versus those who believe they would not. While the CFOs and CEOs in both groups have relatively strong beliefs that using accounting adjustments is lying, those who believe their firms would manage earnings have significantly weaker beliefs that using accounting adjustments is lying (LYING_ACEM = 56 versus 68; t-statistic = −3.07; p-value < 0.01, two-tail). These CFOs and CEOs also have significantly weaker beliefs that using operational adjustments is lying (LYING_OPEM = 20 versus 44; t-statistic = −6.73; p-value < 0.01, two-tail).

To more formally assess the extent to which managers’ beliefs that EM is lying affects their EM decisions, we use Equations (3) and (4). These two equations extend Equations (1) and (2), respectively, to include the two individual LYING variables

\begin{array}{l} E M_L I K E L I H O O D_{i} = α + β_{1} S U C C E S S_{i} + β_{2} M A N A G E R_{i} + β_{3} C U R S H R_{i} \\ + β_{4} F U T S H R_{i} + β_{5} D E B T_{i} + β_{6} O T H E R_{i} + β_{7} E T H I C S_{i} + β_{8} L Y I N G_{i} + ε_{i} \end{array}

(3)

\begin{array}{l} E M_A P P R O A C H_{i} = α + β_{1} S U C C E S S_{i} + β_{2} M A N A G E R_{i} + β_{3} C U R S H R_{i} \\ + β_{4} F U T S H R_{i} + β_{5} D E B T_{i} + β_{6} O T H E R_{i} + β_{7} E T H I C S_{i} + β_{8} L Y I N G_{i} + ε_{i} \end{array}

(4)

Panel A (B) of Table 5 reports the results from estimating Equation (3) (Equation (4)). For parsimony, we aggregate the explanatory power of the individual variables into their constructs. The baseline model contains all the constructs except for ETHICS and LYING. The lying model adds LYING to the baseline model, the ethics model adds ETHICS to the baseline model, and the full model adds both.

Table 5.

Explanatory power of lying for the ethics of earnings managements acts.

$E M_L I K E L I H O O D_{i} = α + β_{1} S U C C E S S_{i} + β_{2} M A N A G E R_{i} + β_{3} S T A K E H O L D E R S_{i} + β_{4} E T H I C S_{i} + β_{5} L Y I N G_{i} + ε_{i} .$
	Panel A: Incremental effect of individual ETHICS and LYING constructs over baseline model.
	Lying model		Ethics model		Full model
	R ²	F-statistic	R ²	F-statistic	R ²	F-statistic
Baseline model	0.172	19.77***	0.135	12.85***	0.103	13.60***
ETHICS			0.169	20.60***	0.117	14.50***
LYING	0.064	9.91***			0.012	4.63***
Independent R² total	0.236		0.304		0.232
Correlated R²	0.135		0.170		0.256
Total adjusted R²	0.371	6.81***	0.474	9.63***	0.488	9.15***
N	225		225		225
Panel B: Incremental effect of individual ETHICS and LYING constructs over baseline model.
$E M_A P P R O A C H_{i} = α + β_{1} S U C C E S S_{i} + β_{2} M A N A G E R_{i} + β_{3} S T A K E H O L D E R S_{i} + β_{4} E T H I C S_{i} + β_{5} L Y I N G_{i} + ε_{i}$
	Lying Model		Ethics Model		Full Model
	R ²	F-statistic	R ²	F-statistic	R ²	F-statistic
Baseline model	0.153	25.21***	0.117	16.28***	0.119	15.28***
ETHICS			0.208	17.10***	0.060	5.63***
LYING	0.153	17.44***			0.010	2.37**
Independent R² total	0.306		0.325		0.189
Correlated R²	0.129		0.166		0.368
Total adjusted R²	0.435	6.21***	0.491	7.25***	0.497	6.79***
N	153		153		153

In the full model, the R² for LYING is significant for both the overall EM assessment and the EM approach assessment (smallest F-statistic = 2.37; p-value < 0.05, two-tail). For the disaggregated full model results (untabulated), the coefficient for LYING_ACEM is negative and significant for both assessments (smallest t-statistic = −2.12; p-value < 0.05, two-tail); the coefficient for LYING_OPEM is not significant for either assessment. Overall, these results are consistent with the strength of CFOs and CEO’s beliefs that EM is lying – particularly their beliefs about accounting adjustments – being a significant determinant of their EM decisions.

Finally, we examine the extent to which CFO and CEOs’ beliefs that EM is lying explain the effect of ethics on their EM decisions. To test this, we compare the explanatory power of ETHICS in the full model versus in the ethics model. In Panel A (B), the incremental explanatory power of ETHICS decreases from 0.169 (0.208) in the ethics model to 0.117 (0.060) in the full model. These decreases imply that LYING accounts for approximately 31% of the explanatory power of ETHICS in the overall EM assessment and approximately 71% in the EM approach assessment. Responses from the CFO and CEO interviewees are consistent with a link between lying and ethics. For example, one interviewee stated, ‘If the decisions or judgements are going beyond what might be considered ethical, then I think that probably does [move] into the lying category’.

Overall, our results suggest that CFOs and CEOs’ beliefs about whether EM is lying affects both EM decisions through two paths. The first is through a direct effect on their EM decisions. The second is through their beliefs about the ethics of EM.

4.4. To what extent do CEOs versus CFOs initiate EM?

Our final research question addresses the extent to which CEOs versus CFOs initiate EM. Using an 11-point scale from 0% (‘CEO initiated’) to 100% (‘CFO initiated’), the CFO and CEO survey participants responded to the following statement: ‘Indicate your belief about the degree to which attempts to improve reported performance via accounting and/or operational adjustments are initiated by a company’s CEO versus CFO’.

Figure 3 presents a histogram of the responses. A quarter of the participants answered 50, and the overall mean response is 41. The mean responses for CFOs and CEOs are similar at 42 and 39, respectively.²⁷ Overall, these results suggest that both CFOs and CEOs believe that each is significantly involved in initiating EM, but that CEOs have slightly greater influence. These results are generally consistent with our interviewees, with nine of them believing EM is a joint decision between the CEO and CFO. For example, one CFO stated, ‘Oh it’s both absolutely. So the CFO’s got to work very closely with the CEO so that there’s alignment in terms of what needs to be done’. And a second stated, ‘Well to me it’s a partnership and it’s a joint decision process that you are working through with the CEO’.

Figure 3.

Histogram of perceptions of the role of the CEO versus the CFO in initiating earnings management.

To further our understanding of CFOs versus CEOs, Table 6 presents univariate tests of differences between our CFO and CEO survey participants. Only three items significantly differ. First, CEOs have significantly stronger beliefs that not managing earnings is ethically questionable (ETHICS_NOEM = 31.46 versus 20.92; t-statistic = 2.33; p-value < 0.05, two-tail). Second, CEOs are significantly more likely to prefer operational adjustments (EM_APPROACH = 0.77 versus 0.72; t-statistic = 1.67; p-value < 0.10, two-tail). Third, and potentially explaining their greater preference for operational adjustments, CEOs have significantly stronger beliefs that using accounting adjustments is lying (LYING_ACEM = 0.65 vs 0.57; t-statistic = 1.97; p-value < 0.05, two-tail).

Table 6.

Univariate tests of CEOs versus CFOs.

	CEO		CFO		t-statistic
	Mean	Median	Mean	Median
DECISIONS
EM_LIKELIHOOD	0.55	0.70	0.54	0.60	0.13
EM_DECISION	0.69	1.00	0.68	1.00	0.17
EM_APPROACH	0.77	0.80	0.72	0.80	1.67*
ETHICS
ETHICS_NOEM	31.46	15	20.92	0	2.33**
ETHICS_ACEM	67.81	80	66.55	70	0.31
ETHICS_OPEM	29.51	20	35.85	30	−1.60
LYING
LYING_ACEM	0.65		0.57		1.97**
LYING_OPEM	0.26		0.28		−0.52
SUCCESS
SUCCESS_ACEM	0.36	0.30	0.39	0.30	−0.56
SUCCESS_OPEM	0.54	0.60	0.50	0.50	0.79
SUCCESS_COMBINE	0.64	0.80	0.60	0.70	0.85
MANAGER
MANAGER_NOEMCOST	45.78	50	41.69	45	1.05
MANAGER_ACCOST	38.07	20	38.30	30	−0.15
MANAGER_OPCOST	30.60	20	28.31	20	0.65
CURRENT SHAREHOLDERS
CURSHR_NOEMCOST	41.21	40	44.79	50	−0.98
CURSHR_ACCOST	27.47	20	25.49	20	0.55
CURSHR_OPCOST	23.25	20	23.87	20	−0.21
FUTURE SHAREHOLDERS
FUTSHR_NOEMCOST	18.80	10	20.85	10	−0.67
FUTSHR_ACCOST	34.46	20	32.61	25	0.46
FUTSHR_OPCOST	26.75	20	30.56	25	−1.10
DEBTHOLDERS
DEBT_NOEMCOST	17.59	10	17.07	10	0.17
DEBT_ACCOST	24.46	10	21.35	10	0.89
DEBT_OPCOST	14.58	10	18.72	10	−1.58
OTHER STAKEHOLDERS
OTHER_NOEMCOST	20.24	10	20.49	20	−0.09
OTHER_ACCOST	23.25	20	21.76	20	0.48
OTHER_OPCOST	26.87	20	23.31	20	1.11

The difference in beliefs regarding accounting adjustments as lying potentially reflects that 90% of CFOs have a professional accounting qualification versus only 25% of the CEOs. Those with an accounting qualification may be more knowledgeable about the extent to which accounting numbers (even unmanaged numbers) depend upon estimates, assumptions and judgement, which in turn may cause them to view EM as a reasonable revision of underlying estimates, assumptions and judgements within the discretion allowed under accounting standards. Consistent with this, the CFOs and CEOs without an accounting qualification (untabulated) have significantly stronger beliefs that using accounting adjustments is lying (LYING_ACEM = 67 versus 56; t-statistic = 2.41; p-value < 0.05, two-tail).

5. Discussion and conclusion

In this article, we use information gathered from Australian CFOs and CEOs via a case-based survey and interviews to provide a richer understanding of managers’ EM decisions and behaviour. We show that when managers are presented with a scenario where there is an incentive for EM the vast majority report that their company would manage earnings.

From subsequent questions of managers in the survey, this enables us to address several important gaps in the extant EM research. First, unlike prior research, we focus on who CFOs and CEOs consider when faced with a potential EM situation, simultaneously including manager self-interest and individual stakeholder groups (i.e. current shareholders, future shareholders, debtholders and other stakeholders) as independent variables in our explanatory model, along with the CFOs and CEOs’ perceptions about the ethics of EM. We find that ethical concerns are the most important determinants of both our CFOs and CEOs’ overall EM assessments and their EM approach assessments. While we find – consistent with conventional wisdom – that ethical concerns discourage EM, interestingly, we also find that they can motivate EM. Specifically, concerns that not managing earnings (and thus missing market expectations) is ethically questionable are positively associated with the overall EM assessment. For the overall EM assessment, it appears that CFOs and CEOs also trade-off the costs to current shareholders from not managing earnings against the costs to current (future) shareholders from using operational (accounting) adjustments for EM. We find no evidence that manager self-interest is important for either the overall EM assessment or the EM approach assessment

Second, we examine the extent to which CFOs and CEOs perceive EM to be lying. We find that their beliefs are quite heterogeneous, both for accounting and operational adjustments. However, on average, there is a much stronger belief that using accounting adjustments is lying. Stronger beliefs that EM is lying are also negatively associated with the overall EM assessment and the EM approach assessment. In addition, a significant portion of the explanatory power of ethics on the EM assessments is attributable to perceptions that EM is lying.

Third, we examine who initiates EM. We find that both CFOs and CEOs believe that each are significantly involved in initiating EM, with the CEO having slightly greater influence. In addition, CEOs have stronger beliefs that using accounting adjustments is lying, and consistent with this, a greater preference for using operational adjustments for EM. CEOs also have stronger beliefs that not managing earnings and missing the earnings benchmark is unethical.

Our results suggest several potential approaches for discouraging EM by addressing one or more of the following factors: pressure for EM, opportunity to successfully manage earnings, and ability to rationalize EM (Cressey, 1950). Potential approaches include (1) shifting the focus of CFOs and CEOs from current shareholders to all stakeholders and from the short-term to the long-term, consistent with the US Business Roundtable report that redefines the purpose of a corporation (Benoit, 2019), as well as the trend of linking executive compensation to ESG metrics (Shumsky, 2019); and (2) changing the language that regulators, Board members, auditors and the business press use when referring to EM so that the language is clearer that EM is lying and/or unethical. These potential solutions will not reduce EM overnight and may appear to be idealistic. However, as Jensen (2003: 404) notes, ‘it has taken many years to weave lying and deceit into the fabric of our businesses; cleansing the fabric will take time as well’. Given that 70 years of regulations have not been successful, maybe it is time to try new approaches.

A potential limitation of our study is that CFOs and CEOs may not have answered truthfully because they did not want to admit to engaging in questionable behaviour. Due to such concerns, we did not ask participants whether they would engage in EM if in a situation similar to the case; instead, we asked them whether their company would (Fisher, 1993). However, approximately 68% of the CFOs and CEOs stated that their companies would manage earnings if in a situation similar to the one in the case, suggesting, as Graham et al. (2005) note, that unwillingness to admit to questionable behaviour is not a significant issue. Similarly, almost all our interviewee CFOs and CEOs thought that firms in a situation similar to the one in the case would manage earnings, and a number of the interviewees were quite open in discussing questionable behaviours. Another limitation is the fact that we are unable to identify the companies of the CFOs/CEOs who responded to our survey, which was part of the ethical requirements to actually conduct the survey. We also only provided three options (do nothing; manage earnings by accounting adjustments; or manage earnings by operational adjustments) for participants in how they might respond to the situation presented where the company’s earnings are not going to meet market expectations. We acknowledge that there is a fourth possible option which is to ‘walk down’ analysts’ expectations. Finally, there are also other limitations inherent in using a survey method as is also highlighted by Dichev et al. (2013: 2), however they then state that the benefits include ‘a potential way to address often intractable issues related to omitted variables and the inability to draw causal links that are endemic to large-sample archival work’ as well as asking ‘ask key decision makers directed questions about their behaviour as opposed to inferring intent from statistical associations between proxy variables surrogating for such intent’.

We believe the results from this survey provide numerous avenues for future research as it shows that managers’ decision-making on this topic is more complex that is often assumed by prior research. There are a number of findings that warrant further investigation by research but broadly the fact that ethics (and perceptions of lying) have been shown to have such an impact on managers’ decisions is worth further exploration. Also, the fact that some managers think it is unethical to not manage earnings is quite contrary to the expectations from the prior literature and is certainly another issue that warrants further research. Finally, EM is a topic that has been the subject of a significant amount of archival research and this approach of using a survey has helped to address some of the intractable issues from that prior research, which is also in the spirit of a couple of other studies (e.g. Dichev et al., 2013; Graham et al., 2005). We suggest that the number of new insights we have been able obtain in what is a very well researched topic by using a survey and interviews shows that there is value in future research on EM using different research methods beyond looking at the archival databases.

Supplemental Material

sj-docx-1-aum-10.1177_03128962221137235 – Supplemental material for Earnings management: Who do managers consider and what is the relative importance of ethics?

Supplemental material, sj-docx-1-aum-10.1177_03128962221137235 for Earnings management: Who do managers consider and what is the relative importance of ethics? by Paul Coram, James R. Frederickson and Matthew Pinnuck in Australian Journal of Management

Footnotes

Appendix

Appendix 1

Definition of variables from case-based survey.

Variable	Definition	Scale
DEPENDENT VARIABLES
EM_LIKELIHOOD	Likelihood that the CEO/CFO’s company would manage earnings in a situation similar to the one in the case	11-point 0% (‘Certain not to happen’) to 100% (‘Certain to happen’)
EM_DECISION	Whether the CEO/CFO’s company would manage earnings in a situation similar to the one in the case	‘Yes’ or ‘No’
EM_APPROACH	If choose ‘Yes’ for EM_DECISION the approach the CEO/CFO’s company would use to manage earnings	11-point 0% (‘Rely solely on accounting adjustments’) to 50% (‘Rely 50% on accounting adjustments and 50% on operational adjustments’) to 100% (‘Rely solely on operational adjustments’)
INDEPENDENT VARIABLES
SUCCESS CONSTRUCT
SUCCESS_ACEM	Likelihood that the CEO/CFO’s company would be able to increase reported earnings to meet market expectations using only accounting adjustments	11-point 0% (‘Certain not to happen’) to 100% (‘Certain to happen’)
SUCCESS_OPEM	Likelihood that the CEO/CFO’s company would be able to increase reported earnings to meet market expectations using only operational adjustments	11-point 0% (‘Certain not to happen’) to 100% (‘Certain to happen’)
SUCCESS_COMBINE	Likelihood that the CEO/CFO’s company would be able to increase reported earnings to meet market expectations using a combination of accounting and operational adjustments	11-point 0% (‘Certain not to happen’) to 100% (‘Certain to happen’)
MANAGER CONSTRUCT
MANAGER_NOEMCOST	Perceived cost to the CEO and CFO if the firm does not manage earnings and misses the earnings benchmark	11-point 0% (‘Absolutely no costs’) to 100% (‘Very large costs’)
MANAGER_ACCOST	Perceived cost to the CEO and CFO if the firm uses accounting adjustments to manage earnings	11-point 0% (‘Absolutely no costs’) to 100% (‘Very large costs’)
MANAGER_OPCOST	Perceived cost to the CEO and CFO if the firm uses operational adjustments to manage earnings	11-point 0% (‘Absolutely no costs’) to 100% (‘Very large costs’)
CURSHR CONSTRUCT
CURSHR_NOEMCOST	Perceived cost to current shareholders if the firm does not manage earnings and misses the earnings benchmark	11-point 0% (‘Absolutely no costs’) to 100% (‘Very large costs’)
CURSHR_ACCOST	Perceived cost to current shareholders if the firm uses accounting adjustments to manage earnings	11-point 0% (‘Absolutely no costs’) to 100% (‘Very large costs’)
CURSHR_OPCOST	Perceived cost to current shareholders if the firm uses operational adjustments to manage earnings	11-point 0% (‘Absolutely no costs’) to 100% (‘Very large costs’)
FUTSHR CONSTRUCT
FUTSHR_NOEMCOST	Perceived cost to future shareholders if the firm does not manage earnings and misses the earnings benchmark	11-point 0% (‘Absolutely no costs’) to 100% (‘Very large costs’)
FUTSHR_ACCOST	Perceived cost to future shareholders if the firm uses accounting adjustments to manage earnings	11-point 0% (‘Absolutely no costs’) to 100% (‘Very large costs’)
FUTSHR_OPCOST	Perceived cost to future shareholders if the firm uses operational adjustments to manage earnings	11-point 0% (‘Absolutely no costs’) to 100% (‘Very large costs’)
DEBT CONSTRUCT
DEBT_NOEMCOST	Perceived cost to debtholders if the firm does not manage earnings and misses the earnings benchmark	11-point 0% (‘Absolutely no costs’) to 100% (‘Very large costs’)
DEBT_ACCOST	Perceived cost to debtholders if the firm uses accounting adjustments to manage earnings	11-point 0% (‘Absolutely no costs’) to 100% (‘Very large costs’)
DEBT_OPCOST	Perceived cost to debtholders if the firm uses operational adjustments to manage earnings	11-point 0% (‘Absolutely no costs’) to 100% (‘Very large costs’)
OTHER STAKEHOLDER CONSTRUCT
OTHER_NOEMCOST	Perceived cost to other stakeholders if the firm does not manage earnings and misses the earnings benchmark	11-point 0% (‘Absolutely no costs’) to 100% (‘Very large costs’)
OTHER_ACCOST	Perceived cost to other stakeholders if the firm uses accounting adjustments to manage earnings	11-point 0% (‘Absolutely no costs’) to 100% (‘Very large costs’)
OTHER_OPCOST	Perceived cost to other stakeholders if the firm uses operational adjustments to manage earnings	11-point 0% (‘Absolutely no costs’) to 100% (‘Very large costs’)
ETHICS CONSTRUCT
ETHICS_NOEM	The extent to which doing nothing and missing the earnings benchmark is ethically questionable	11-point 0 (‘Not questionable at all’) to 100 (‘Extremely questionable’)
ETHICS_ACEM	The extent to which using accounting adjustments to manage earnings is ethically questionable	11-point 0 (‘Not questionable at all’) to 100 (‘Extremely questionable’)
ETHICS_OPEM	The extent to which using operational adjustments to manage earnings is ethically questionable	11-point 0 (‘Not questionable at all’) to 100 (‘Extremely questionable’)
LYING CONSTRUCT
LYING_ACEM	The extent to which the use accounting adjustments to manage earnings is lying	11-point 0% to 100%
LYING_OPEM	The extent to which the use of operational adjustments to manage earnings is lying	11-point 0% to 100%

Acknowledgements

We thank workshop participants at The University of New South Wales, The University of Melbourne, London School of Economics and Political Science, and Maastricht University for their comments and suggestions. We also thank the practitioners and academics that provided valuable feedback on our instrument decision and the chief financial officers and chief executive officers who participated in our study.

Final transcript accepted 28 September 2022 by Andrew Jackson

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: We acknowledge the financial support from the Faculty of Business and Economics and Melbourne Business School of The University of Melbourne.

ORCID iD

Paul Coram

Supplemental material

Supplemental material for this article is available online.

Notes

References

Badkar

Gray

(2019) Molson Coors forced to restate results. Financial Times, 13 February, p. 12.

Ball

(2013) Accounting informs investors and earnings management is rife: Two questionable beliefs. Accounting Horizons 27: 847–853.

Beaudoin

Cianci

Tsakumis

(2015) The impact of CFOs’ incentives and earnings management ethics on their financial reporting decisions: The mediating role of moral disengagement. Journal of Business Ethics 128: 505–518.

Benoit

(2019) Move over, shareholders: Top CEOs say companies have obligations to society. The Wall Street Journal, 19 August. Available at: https://www.wsj.com/articles/business-roundtable-steps-back-from-milton-friedman-theory-11566205200

Bhojraj

Libby

(2005) Capital market pressure, disclosure frequency-induced earnings/cash flow conflict, and managerial myopia. The Accounting Review 80: 1–20.

Bruns

Merchant

(1990) The dangerous morality of managing earnings. Management Accounting 72: 22–25.

Buchholz

Lopatta

Maas

(2020) The deliberate engagement of narcissistic CEOs in earnings management. Journal of Business Ethics 167: 663–686.

Burgstahler

Dichev

(1997) Earnings management to avoid earnings decreases and losses. Journal of Accounting and Economics 24: 99–126.

Cai

Wang

JT-Y

(2006) Overcommunication in strategic information transmission games. Games and Economic Behavior 56: 7–36.

10.

Cardinaels

(2016) Earnings benchmarks, information systems, and their impact on the degree of honesty in managerial reporting. Accounting, Organizations and Society 52: 50–62.

11.

Charness

Dufwenberg

(2006) Promises and partnership. Econometrica 74: 1579–1601.

12.

Charness

Dufwenberg

(2010) Bare promises: An experiment. Economics Letters 107: 281–283.

13.

Chava

Purnanandam

(2010) CEOs versus CFOs: Incentives and corporate policies. Journal of Financial Economics 97: 263–278.

14.

Cheng

Warfield

(2005) Equity incentives and earnings management. The Accounting Review 80: 441–476.

15.

Chow

Cooper

Waller

(1988) Participative budgeting: Effects of a truth-inducing pay scheme and information asymmetry on slack and performance. The Accounting Review 63: 111–122.

16.

Chu

Dechow

Hui

, et al. (2019) Maintaining a reputation for consistently beating earnings expectations and the slippery slope to earnings manipulation. Contemporary Accounting Research 36: 1966–1998.

17.

Cohen

Zarowin

(2010) Accrual-based and real earnings management activities around seasoned equity offerings. Journal of Accounting and Economics 50: 2–19.

18.

Cohen

Holder-Webb

Sharp

, et al. (2007) The effects of perceived fairness on opportunistic behavior. Contemporary Accounting Research 24: 1119–1138.

19.

Cressey

(1950) The criminal violation of financial trust. American Sociological Review 15: 738–743.

20.

Dechow

Skinner

(2000) Earnings management: Reconciling the views of accounting academics, practitioners, and regulators. Accounting Horizons 14: 235–250.

21.

Dechow

Schrand

(2010) Understanding earnings quality: A review of the proxies, their determinants and their consequences. Journal of Accounting and Economics 50: 344–401.

22.

Dichev

Graham

Harvey

, et al. (2013) Earnings quality: Evidence from the field. Journal of Accounting and Economics 56: 1–33.

23.

Dikolli

Keusch

Mayew

, et al. (2020) CEO behavioral integrity, auditor responses, and firm outcomes. The Accounting Review 95: 61–88.

24.

Dreber

Johannesson

(2008) Gender differences in deception. Economics Letters 99: 197–199.

25.

Ellingsen

Johannesson

(2004) Promises, threats, and fairness. The Economic Journal 114: 397–420.

26.

Elliot

Hobson

Jackson

(2011) Disaggregating management forecasts to reduce investors’ susceptibility to earnings fixation. The Accounting Review 86: 185–208.

27.

Erat

Gneezy

(2012) White lies. Management Science 58: 723–733.

28.

Erickson

Hanlon

Maydew

(2006) Is there a link between executive equity incentives and accounting fraud? Journal of Accounting Research 44: 113–143.

29.

Evans

Hannan

Krishnan

, et al. (2001) Honesty in managerial reporting. The Accounting Review 76: 537–559.

30.

Evans

Houston

Peters

, et al. (2015) Reporting regulatory environments and earnings management: U.S. and Non-U.S. firms using U.S. GAAP or IFRS. The Accounting Review 90: 1969–1994.

31.

Feng

Luo

, et al. (2011) Why do CFOs become involved in material accounting manipulations? Journal of Accounting and Economics 51: 21–36.

32.

Fields

Lys

Vincent

(2001) Empirical research on accounting choice. Journal of Accounting and Economics 31: 255–307.

33.

Fischer

Rosenzweig

(1995) Attitudes of students and accounting practitioners concerning the ethical acceptability of earnings management. Journal of Business Ethics 14: 433–444.

34.

Fisher

Frederickson

Peffer

(2006) Budget negotiations in multi-period settings. Accounting, Organizations, and Society 31: 511–582.

35.

Fisher

(1993) Social desirability bias and the validity of indirect questioning. Journal of Consumer Research 20: 303–315.

36.

Forsyth

(1980) A taxonomy of ethical ideologies. Journal of Personality and Social Psychology 39: 175–184.

37.

Friedman

(2014) Implications of power: When the CEO can pressure the CFO to bias reports. Journal of Accounting and Economics 58: 117–141.

38.

Fuller

Jensen

(2010) Just say no to Wall Street: Putting a stop to the earnings game. Journal of Applied Corporate Finance 22: 59–63.

39.

Gendron

Bédard

(2006) On the constitution of audit committee effectiveness. Accounting, Organizations and Society 31: 211–239.

40.

Gigler

Kanodia

Sapra

, et al. (2014) How frequent financial reporting causes managerial short-termism: An analysis of the costs and benefits of reporting frequency. Journal of Accounting Research 52: 357–387.

41.

Gneezy

(2005) Deception: The role of consequences. The American Economic Review 95: 384–394.

42.

Graham

Harvey

(2001) The theory and practice of corporate finance: Evidence from the field. Journal of Financial Economics 60: 187–243.

43.

Graham

Harvey

Rajgopal

(2005) The economic implications of corporate financial reporting. Journal of Accounting and Economics 40: 3–73.

44.

Greenfield

Norman

Wier

(2008) The effect of ethical orientation and professional commitment on earnings management behaviour. Journal of Business Ethics 83: 419–434.

45.

Gryta

(2019) GE confronts legacy of bad power deals. The Wall Street Journal, 14 February, p. A1.

46.

Gryta

Benoit

(2018) SEC digs into GE Insurance – Legacy business had accounting problems. The Wall Street Journal, 1 December, p. B.1.

47.

Haddon

(2019) Kraft Heinz to restate years of fiscal results. The Wall Street Journal, 7 May, p. A.1.

48.

Ham

Lang

Seybert

, et al. (2017) CFO narcissism and financial reporting quality. Journal of Accounting Research 55: 1089–1135.

49.

Hannan

Rankin

Towry

(2006) The effect of information systems on honesty in managerial reporting. Contemporary Accounting Research 23: 885–918.

50.

Healy

(1985) The impact of bonus schemes on the selection of accounting principles. Journal of Accounting and Economics 7: 85–107.

51.

Healy

Palepu

(1993) The effect of firms’ financial disclosure strategies on stock prices. Accounting Horizons 7: 1–11.

52.

Healy

Wahlen

(1999) A review of the earnings management literature and its implications for standard setting. Accounting Horizons 13: 365–383.

53.

Holthausen

Larcker

Sloan

(1995) Annual bonus schemes and the manipulation of earnings. Journal of Accounting and Economics 19: 29–74.

54.

Hope

Wang

(2018) Management deception, big-bath accounting, and information asymmetry: Evidence from linguistic analysis. Accounting, Organizations and Society 70: 33–51.

55.

Hurkens

Kartik

(2009) Would I lie to you? On social preferences and lying aversion. Experimental Economics 12: 180–192.

56.

Jackson

(2018) Discretionary accruals: Earnings management. . .or Not? Abacus 54: 136–153.

57.

Jensen

(2001) Value maximisation, stakeholder theory, and the corporate objective function. Journal of Applied Corporate Finance 14: 8–21.

58.

Jensen

(2003) Paying people to lie: The truth about the budgeting process. European Financial Management 9: 379–406.

59.

Jiang

Petroni

Wang

(2010) CFOs and CEOs: Who has the most influence on earnings management. Journal of Financial Economics 96: 513–526.

60.

Jiang

(2013) Cheating in mind games: The subtlety of rules matters. Journal of Economic Behavior & Organization 93: 328–336.

61.

Johnson

Fleischman

Valentine

, et al. (2012) Managers’ ethical evaluations of earnings management and its consequences. Contemporary Accounting Research 29: 910–927.

62.

Jones

(1991) Earnings management during import relief investigations. Journal of Accounting Research 29: 193–228.

63.

Kant

(1838) Critic of Pure Reason (Trans. F Haywood). London: William Pickering.

64.

Kaplan

(2001) Further evidence on the ethics of managing earnings: An examination of the ethically related judgments of shareholders and non-shareholders. Journal of Accounting and Public Policy 20: 27–44.

65.

Kothari

Leone

Wasley

(2005) Performance matched discretionary accrual measures. Journal of Accounting and Economics 39: 163–197.

66.

Leuz

Nanda

Wysocki

(2003) Earnings management and investor protection: An international comparison. Journal of Financial Economics 69: 505–527.

67.

Libby

Rennekamp

Seybert

(2015) Regulation and the interdependent roles of managers, auditors, and directors in earnings management and accounting choice. Accounting, Organizations and Society 47: 25–42.

68.

Lundquist

Ellingsen

Gribbe

, et al. (2009) The aversion to lying. Journal of Economic Behavior and Organization 70: 81–92.

69.

McNichols

(2000) Research design issues in earnings management studies. Journal of Accounting and Public Policy 19: 313–345.

70.

Merchant

Rockness

(1994) The ethics of managing earnings: An empirical investigation. Journal of Accounting and Public Policy 13: 79–94.

71.

Merriam-Webster (2019) Merriam-Webster’s Learner’s Dictionary. Available at: http://www.learnersdictionary.com/definition/mislead (accessed September 10).

72.

Neuendorf

(2002) The Content Analysis Guidebook. Thousand Oaks, CA: SAGE.

73.

Rankin

Schwartz

Young

(2008) The effect of honesty and superior authority on budget proposals. The Accounting Review 83: 1083–1099.

74.

Roychowdhury

(2006) Earnings management through real activities manipulation. Journal of Accounting and Economics 42: 335–370.

75.

Sanjiv

Gneezy

(2012) White lies. Management Science 58: 723–733.

76.

Shumsky

(2019) More companies link executive pay to sustainability targets. The Wall Street Journal, 24 June. Available at: https://www.wsj.com/articles/more-companies-link-executive-pay-to-sustainability-targets-11561379745

77.

Stevens

(2002) The effects of reputation and ethics on budgetary slack. Journal of Management Accounting Research 14: 153–171.

78.

Sutter

(2009) Deception through telling the truth! Experimental evidence from individuals and teams. The Economic Journal 119: 47–60.

79.

Trahan

Gitman

(1995) Bridging the theory-practice gap in corporate finance: A survey of chief financial officers. Quarterly Review of Economics and Finance 35: 73–87.

80.

Vanberg

(2008) Why do people keep their promises? An experimental test of two explanations. Econometrica 76: 1467–1480.

81.

Watts

Zimmerman

(1986) Positive Accounting Theory. London: Prentice-Hall.

82.

Wicks

Freeman

Werhane

, et al. (2010) Business Ethics: A Managerial Approach. Upper Saddle River, NJ: Prentice-Hall.

83.

Zang

(2012) Evidence on the trade-off between real activities manipulation and accrual-based earnings management. The Accounting Review 87: 675–703.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.08 MB