Journal of Business Accounting and Finance Perspectives

J_Bus_Account_Financ_Perspect 2020, 3(1), 2; doi:10.26870/jbafp.2020.03.002

Article

Career concerns and earnings management

Elizabeth Demers ¹ and Chong Wang ²^,*

School of Accounting and Finance, University of Waterloo, 200 University Ave West, Waterloo, ON, Canada

Graduate School of Defense Management, Naval Postgraduate School, Monterey, CA 93943, USA

Corresponding author: cwang@nps.edu

How to cite: Elizabeth Demers, Chong Wang. Career concerns and earnings management. J. Bus. Account. Financ. Perspect., 2020, 3(1): 2; doi:10.26870/jbafp.2020.03.002.

Received: 27 August 2019 / Accepted: 6 February 2020 / Published: 10 February 2020

Abstract

Motivated by the disconnect between survey evidence documenting that executives prioritize implicit contracting (i.e., labor market-based career concerns) when making earnings management decisions (Graham et al., 2005) and the extant literature’s focus on explicit contracting to explain earnings manipulation, we analytically examine the role of managerial career concerns in earnings management. Building on Holmstrom (1982, 1999), we present a career concerns-based earnings management model that incorporates the unique reversing nature of earnings management. A key insight derived from the model is that whether the predictions of a traditional career concerns model prevail, which is to say that managers engage in more income-increasing behavior in their early years, critically depends upon the reversal characteristics of the earnings management vehicle chosen.

Keywords:

earnings management; career concerns; accruals and real activities management; accruals reversal; reversal of earnings management

JEL Classification:

M40; M41

1. Introduction

Top executives of US publicly traded firms often face intense incentives to manage earnings.1 Most of the extant academic literature links the incentives for earnings management, directly or indirectly, to explicit contracts such as executive cash bonus maximization contracts (Healy, 1985) or the avoidance of debt covenant violations (Defond and Jiambalvo, 1994).2 While such explicit contracting-based incentives undoubtedly play an important role, the literature has been almost silent about the effects of implicit contracts and incentives over earnings management.3 This is surprising, especially given the survey evidence of Graham et al. (2005), which documents that more than three quarters of responding executives consider upward mobility in the labor market (i.e., an implicit career incentive) to be more important than short-run current compensation benefits in influencing their earnings management decisions.4 Our study addresses this apparent disconnect between the importance placed upon implicit contracts by executives and the extant literature’s focus on explicit contracts by developing an analytical model of earnings management that builds upon the seminal career concerns work of Holmstrom (1982, 1999). The extensions introduced by our model, incorporating the particular features of alternative forms of earnings management, lead to some notable results that deviate from the predictions of conventional career concerns models. Whereas in the traditional career concerns setting managers are expected to exert themselves relatively more in their early years in order to influence the labor market’s assessment of their quality, our model predicts that managers will engage in less income-increasing earnings management in their early career stage (relative to their later stage) under certain conditions related to the value-destructive properties of particular types of earnings management.

Broadly speaking, executives can manage reported earnings in one of two ways. First, managers may use “accruals management”, which refers to the discretion available within Generally Accepted Accounting Principles (GAAP) (Dechow and Skinner, 2000) to intertemporally shift earnings from one period to another. Alternatively, executives may use real activities manipulations to alter the timing and amount of expenditures in order to affect the current period’s bottom line (e.g., cutting R&D or advertising expenditures). Unlike accrual-based earnings management, which simply shifts expenses or revenues into different reporting periods without directly affecting the underlying economics of the firm, earnings management based on real activities manipulations “deviate[s] from normal business practices” (Roychowdhury, 2006). Empirical evidence supports the notion that, as deviations from optimal behavior, real earnings manipulations have first-order cash flow consequences and are associated with longer-term value destruction to the firm (see, e.g., Li, 2010 and Vorst, 2015). Our model incorporates these characteristic differences between accrual-based and real earnings management, leading to predictions that differ from the standard career concerns model depending upon the earnings management vehicle chosen. Specifically, our model suggests that managers will undertake fewer real activities manipulations in the early stage of their careers relative to their later years, when this form of earnings management exceeds a certain threshold of value destruction.

Our paper contributes to the literature on earnings management and career concerns across several dimensions. First, to the best of our knowledge, ours is the first study to analytically explain earnings management behavior from the perspective of implicit as opposed to explicit incentives. Whereas in an explicit contracting setting the manager’s compensation is paid in accordance with some agreement at the end of the period based upon realized output, the wages of the manager in our model are determined at the beginning of the period based upon expected output. In other words, the managers in our model are rewarded for future potential rather than by contractually defined compensation based upon realized performance. Second, implicit incentives provide a different perspective relative to explicit contract-based incentives with which to view the pervasiveness of earnings management. Specifically, incentives from explicit contracts typically describe earnings management behaviors as “event-driven” and hence tend to occur on a less frequent basis.5 By contrast, incentives from implicit contracts (such as a manager’s value in the labor market) are perennially present rather than driven by a particular event, and hence they are likely to be more profound and persistent. Last and perhaps most importantly, although our analysis is anchored on the Holmstrom (1982, 1999) framework, the particularities of accruals reversal and value destruction of alternative earnings management tools lead to important predictions that are in direct contrast to Holmstrom in some cases.

The remainder of our paper is organized as follows. In Section 2 we set up, analytically solve, and discuss the predictions generated by our model, while Section 3 concludes the paper.

2. Model

2.1. Model setup and assumptions

We develop a model of earnings management based upon managerial career concerns by building on the seminal work of Holmstrom (1982, 1999). As interpreted by Autrey et al. (2007), “Career concerns are implicit incentives that arise because a manager expects future wages to be affected by the labor market’s use of publicly-observable performance measures to assess the manager’s ability.” A distinctive characteristic of career concerns models is that managers are incentivized by future, rather than current, compensation.

To illustrate the effects of career concerns, we adopt a simple three-period setting in which the manager is young, established, and then retired in each respective period. In Periods 1 and 2, the executive is responsible for managing the firm, while in Period 3 the manager may serve as a board member during retirement.6 Consistent with competitive labor markets being a key premise underlying the general career concerns framework, we also assume that the manager is paid, at the beginning of each period, the expected true output that she will deliver in the current period given her history of observed outputs.

Let ῆ denote the manager’s unobservable productivity, with the following prior distribution:

ῆ ~ N (m0, 1/h0)

(1)

where N (·,·) is the standard normal distribution function, m0 is the mean of the prior, and h0 is the precision of the prior (i.e., the inverse of its variance).

Let ỹt denote the true output, which can be written as follows:

ỹt = ῆ +εt,

(2)

where εt is a shock that is uncorrelated with the manager’s productivity ῆ. Shocks in each period are assumed to be normally distributed with precision hε, such that

εt ~ N (0, 1/hε).

(3)

We further assume that ῆ, ε1, ε2, and ε3 are jointly independent.

In Period 1, when the manager is young, she can “impress” the labor market in the short term by boosting reported profits via earnings management. Let

\tilde{x}

t denote the manager’s reported (or observed) output for period t, which may differ from the true output, ỹt, because the manager has the option of manipulating earnings during periods t = 1, 2. Thus, the observed output,

\tilde{x}

t, is the true output plus the current period’s manipulations (plus any consequences of the prior period’s manipulation).7

Period 1 reported earnings are therefore given by:

\tilde{x} 1 = \tilde{y} 1 + α 1 = \tilde{η} + α 1 + ε 1,

(4)

where α1 is the earnings management in Period 1.

A critical feature of our model, which we now introduce, is that the earnings managed in Period 1 will reverse, at least partially, or lead to value destruction, in Period 2. This assumption states that earnings cannot be indefinitely managed upwards, either via accruals or through real activities manipulations. Even though accruals management decisions are assumed to be value-neutral in that they only involve using the discretion allowed within GAAP to intertemporally shift reported earnings, while not impacting real activities or value creation, the manager nevertheless needs to balance the books via accruals reversals “eventually”. Attempts to persistently manage earnings via accruals will eventually trigger a forced earnings restatement or fraud investigation. More severely, real activities manipulations such as reductions in R&D spending that boost earnings in the short term are assumed to have an adverse real impact on future firm performance (i.e., they are value-destroying). Furthermore, the adverse future impact of real manipulations is usually much greater than the short-term earnings gains that they generate.

In Period 2, when the manager is at the established stage of her career, the Period 1 accrual-based earnings manipulations begin to reverse, or the value-destroying effects of real activities manipulations begin to manifest, at the rate of λ, where 0 < λ ≤ 1 if the channel of earnings management is accrual-based, and λ > 1 if earnings management is accomplished via real activities manipulations. The rationale for these values of λ is that accrual-channeled earnings management merely shifts earnings across periods on a dollar-for-dollar basis, whereas real activities manipulations have a greater long-term value-destroying effect (i.e., the future value destruction exceeds the short-term reported earnings gained from the manipulations). This assumption, that accruals reversal is at most dollar-for-dollar while the reversal of real activities is more than dollar-for-dollar, will prove to have important implications for the predictions generated by our model.

Once again in Period 2, the manager has the option to engage in a second round of earnings manipulations. Let α2 denote the amount of “new” earnings manipulations in Period 2. The reported output in Period 2 is given by:

\tilde{x} 2 = \tilde{y} 2 + α 2 - λ α 1 = \tilde{η} + α 2 - λ α 1 + ε 2 .

(5)

Period 2 earnings manipulations also reverse in the subsequent period, after the manager is retired, and the market includes this reversal in the determination of the manager’s Period 3 output. The Period 3 output,

\tilde{x}

3, which can be interpreted as the manager’s contribution to the firm’s value via board service during retirement, is given by:

\tilde{x} 3 = \tilde{y} 3 - λ α 2 = \tilde{η} - λ α 2 + ε 3 .

(6)

Let wt denote the executive’s wage in period t, paid at the beginning of period t. The assumptions of competitive labor markets and no explicit performance-based pay lead to the following:

w1 = E(ῆ|prior) ≡ m0,

(7)

w2(x1) = E(ῆ|x1) ≡ m1,

(8)

w3(x1, x2) = E(ῆ|x1, x2) ≡ m2.

(9)

Note that Equations (8) and (9) specify that later period wages are based on the labor markets’ projections of the manager’s future true output given her prior observed history of outputs. For example, good performance in the first period will lead to higher pay in the second period, as captured by w2 in the model. Note also that we use the notation {m0, m1, m2} to define the expected true outputs that are conditional upon prior realized outputs. Since {m0, m1, m2} are equivalent to the wages, managers do care about m1 and m2 and will try to manage them. m0 is predetermined and is therefore not subject to the manager’s influence.

It is also worth mentioning that since the manager’s wages are assumed to be determined at the beginning of each period based upon expected output, the model effectively assumes away the role of explicit compensation contracts. Alternatively stated, the manager in our model is rewarded for future potential and not for past performance, whereas in an explicit contracting setting the manager’s compensation is paid at the end of the period based upon realized output. Accordingly, all of our model’s predictions derive solely from the manager’s career concerns, which are characterized by implicit incentives and implicit contracts.

2.2. The derivation of optimal earnings management levels

2.2.1. Defining the information structure, timeline, and equilibrium

We begin by assuming the existence of a competitive labor market under which many firms compete for managerial services. Ours is thus an equilibrium model under which managers exhibit career concerns because they care about their value in the managerial labor market.

The information structure can be summarized as follows:

(a): The firm, the manager, and the market share the same prior belief about ῆ. In other words, there is no asymmetric information or adverse selection at the outset;
(b): The firm, the manager, and the market all see xt each period; and
(c): Only the manager sees αt. In other words, this is a hidden action.

The sequence of events proceeds as follows:

Period 1:
❼
The wage, w1, is paid at the beginning of the period;
❼
α1 is chosen by the manager; and
❼
$\tilde{x}$ 1 = ῆ + α1 + ε1 is realized at value x1.
Period 2:
❼
The wage, w2(x1), is paid at the beginning of the period;
❼
α2 is chosen by the manager; and
❼
$\tilde{x}$ 2 = ῆ + α2 − λα1 + ε2 is realized at value x2.
Period 3:
❼
The wage, w3(x1, x2), is paid at the beginning of the period; and
❼
$\tilde{x}$ 3 = ῆ − λα2 + ε3 is realized at value x3.

Although the manager’s earnings manipulations in each period, αt, constitute a hidden action, in equilibrium both the firm and the labor market correctly anticipate αt. In other words, information is symmetric along the equilibrium path. Moreover, the equilibrium follows the notion of a “Perfect Bayesian Equilibrium” under which sequential rationality applies.

2.2.2. Solving the model

The manager’s objective is to choose the optimal αt (t = 1,2) so as to maximize her utility. The manager’s utility function is represented as the sum of the undiscounted value of her lifetime compensation (i.e., wages), net of the cost of earnings management, as follows:

U = E [w1 − c (α1) + (w2 − c (α2)) + w3],

(10)

where c (αt) denotes the cost associated with earnings manipulations in period t, t∈{1, 2}, with the standard properties of c’ (αt) > 0, c’’ (αt)> 0, and c’ (0) = 0.

Let

\bar{α_{1}}

and

\bar{α_{2}}

denote the labor market’s conjectures of α1 and α2, respectively, while z1 and z2 represent the market’s conjectures of unmanaged earnings, defined as:

z_{1} \equiv x_{1} - \bar{α_{1}} = \tilde{η} + ε_{1},

(11)

z_{2} \equiv x_{1} - \bar{α_{2}} + λ \bar{α_{1}} = \tilde{η} + ε_{2} .

(12)

We apply the standard belief updating formula8 to obtain the conditional mean of normally distributed ῆ , given z1 and z2 to be as follows:9

E (\tilde{η} | (z 1, z 2)) = m 0 + \frac{h ε}{h 0 + 2 h ε} (z 1 + z 2 - 2 m 0) .

(13)

Thus, the market begins with prior m0, and adjusts its beliefs about ῆ based upon the information conveyed by z1 and z2.

Applying Equation (13), we can express w3 as follows:

w 3 (x 1, x 2) = E (\tilde{η} | x 1, x 2) = m 2 = m 0 + \frac{h ε}{h 0 + 2 h ε} (x 1 + x 2 - α 2 + (λ - 1) α 1 - 2 m 0) .

(14)

Moreover, w2 and w1 can be written as:

w 2 (x 1) = E (\tilde{η} | x 1) = m 1 = \frac{h_{0} m_{0} + h_{ε} (x_{1} - \bar{α_{1})}}{h_{0} + h_{ε}},

(15)

w1 = E(ῆ|prior) = m0,

(16)

where we rely on the standard belief updating formula for the substitution of E(ῆ|x1) =

m_{0} + \frac{h_{ε}}{h_{0} + h_{ε}} (z_{1} - m_{0})

into Equation (15).

The sequence {m0, m1, m2} represents the market’s updated beliefs about the manager’s ability over time. In particular, m1 and m2 can be rewritten as follows:

$m_{1} = \frac{h_{0} m_{0} + h_{ε} (x_{1} - {\bar{α}}_{1})}{h_{0} + h_{ε}} = \frac{h_{0}}{h_{0} + h_{ε}} m_{0} + \frac{h_{ε}}{h_{0} + h_{ε}} (x_{1} - \bar{α_{1}})$ and
$m_{2} = m_{0} + \frac{h_{ε}}{h_{0} + 2 h_{ε}} (x_{1} + x_{2} - \bar{α_{2}} + (λ - 1) {\bar{α}}_{1} - 2 m_{0}) = \frac{h_{0} + h_{ε}}{h_{0} + 2 h_{ε}} m_{1} + \frac{h_{ε}}{h_{0} + 2 h_{ε}} (x_{2} - {\bar{α}}_{2} + λ {\bar{α}}_{1})$ .

This clearly shows that, at each period, the market uses a weighted average of the last period’s belief, mt−1, and the new information revealed during the current period (i.e., the unmanaged earnings) in order to form its new belief, mt. Consequently, establishing a good reputation early on will have a long-lasting effect on the manager’s wages (i.e., utility). A similar insight is ged from Equations (14) and (15), wherein the Period 1 earnings management, α1, increases both w2 and w3 via x1, whereas the Period 2 earnings management, α2, only increases w3 (via x2). In other words, early-stage manipulations pay off more than later-stage manipulations. Thus, similar to the classic career concerns model, up to this point in our model managers have incentives to engage in more earnings manipulations at the early stages of their careers.

In contrast to Holmstrom (1982, 1999), however, α1 in our model reverses in Period 2, thereby decreasing Period 2 earnings (x2), which in turn lowers Period 3 wages (w3). This subsequent period reversal effect thereby becomes an offsetting factor to the traditional career concerns incentives, and the overall effect of the two offsetting forces generates different predictions relative to Holmstrom (1982, 1999) under various assumptions regarding λ (i.e., alternatively the rate of accruals reversal or the intensity of the value destruction of the real activities manipulations).

To solve the model, we use the “sequential rationality” implied by a Perfect Bayesian Equilibrium. At the beginning of Period 2, after w2 is paid, the manager chooses α2 to maximize E{−c(α2) + w3}. Using (14), we have the following first-order condition:

- c^{'} (α 2) + \frac{h ε}{h 0 + 2 h ε} = 0,

(17)

which in turn implies the closed-form solution:

c^{'} (α 2^{*}) = \frac{h ε}{h 0 + 2 h ε} .

(18)

The first-order condition (FOC) with respect to α1 is:

\frac{\partial U}{\partial α_{1}} = - c^{'} (α_{1}) + \frac{h_{ε}}{h_{0} + h_{ε}} + \frac{h_{ε}}{h_{0} + 2 h_{ε}} (1 - λ),

(19)

which yields the closed-form solution:

c^{'} (α_{1}^{*}) = \frac{h_{ε}}{h_{0} + h_{ε}} + \frac{h_{ε}}{h_{0} + 2 h_{ε}} (1 - λ) .

(20)

Equations (18) and (20) guarantee a maximum value for the objective function given that the second-order conditions are satisfied. Furthermore, we note that Equation (20) is just a marginal cost equals marginal benefit condition, with the left-hand side term c’(α1) representing the marginal cost (i.e., disutility) of α1 due to the convex cost of earnings management. The first item on the right-hand side,

\frac{h_{ε}}{h_{0} + h_{ε}}

, represents the marginal utility of α1 through its impact on w2, while the second term,

\frac{h_{ε}}{h_{0} + 2 h_{ε}} (1 - λ)

, represents the marginal utility of α1 through its impact on w3. It is worth mentioning that while the first term,

\frac{h_{ε}}{h_{0} + h_{ε}}

, is always positive, the sign and the magnitude of the second term critically depend upon the rate of reversal of the accruals or the intensity of the value destruction of the real activities manipulations, λ.

2.3. Interpretation of optimal earnings management levels

We now discuss and interpret the optimal levels of earnings management in Periods 1 and 2,

α_{1}^{*}

and

α_{2}^{*}

, respectively, for each of the two cases of earnings management based on accruals and real activities manipulation.

Case 1—Accruals: λ ≤ 1

The case of λ ≤ 1 implies a partial or full reversal of Period 1 earnings management activities in Period 2, and is thus applicable to accrual-channeled earnings management. If λ ≤ 1, then from (18) and (20), we have: c’(

α_{1}^{*}

) > c’(

α_{2}^{*}

), which immediately implies

α_{1}^{*}

α_{2}^{*}

. Under this scenario, where the reversal effect is at most dollar-for-dollar, our model yields predictions that are similar to those of the classic Holmstrom career concerns model: managers will manipulate earnings more when they are young in order to enhance the labor market’s perceptions of their abilities.

Case 2—Real Activities: λ > 1

In contrast to accrual earnings management, earnings management based on real activity manipulations involves managers taking actions to adjust the timing and/or scale of the firm’s underlying business activities away from their optimal level. Alternatively stated, these activities are value-destroying. The negative impact in Period 2 is therefore a multiple of the Period 1 earnings management based on real activities manipulation.

Sub-case 2.1: 1 < λ < 1 +

\frac{h_{ε}}{h_{0} + h_{ε}}

Under this scenario, the second term on the right-hand side of Equation (20) becomes negative, thereby dampening the magnitude of

α_{1}^{*}

. However, since the intensity of value destruction, λ, is moderate, the negative impact of the first period’s earnings management, α1, on the third period’s wages, w3 (i.e., the second term on the right-hand side of Equation (20)), is insufficient to dominate the positive impact of α1 on the second period’s wages, w2 (i.e., the first term on the right-hand side of Equation (20)). So, c’(

α_{1}^{*}

) > c’(

α_{2}^{*}

) still holds and therefore

α_{1}^{*}

α_{2}^{*}

. This suggests that, provided that the real activities manipulations are not overly value-destroying, younger managers’ desire to manage earnings in the early stages of their careers in order to impress the labor market will remain dominant. In other words, managers’ concerns about the value-destroying effects of the first period’s real activities manipulations will be insufficient to overturn the standard career concerns result of higher engagement in earnings management during the early career stage.

Sub-case 2.2: 1 < λ = 1 +

\frac{h_{ε}}{h_{0} + h_{ε}}

Under this scenario, c’(

α_{1}^{*}

) = c’(

α_{2}^{*}

), implying

α_{1}^{*}

α_{2}^{*} .

Thus, when the value-destroying effect of a particular real earnings management activity reaches a specific threshold, a younger manager exerts neither more nor less earnings management in the earlier versus later career stages. This is the tipping point at which the two offsetting forces of early gains and subsequent value destruction balance out. Beyond this point, a result opposite to that of the traditional Holmstrom model’s prediction will prevail, as shown below.

Sub-case 2.3: 1 +

\frac{h_{ε}}{h_{0} + h_{ε}}

< λ < 1 +

\frac{h_{0} + 2 h_{ε}}{h_{0} + h_{ε}}

Under this scenario, 0 < c’(

α_{1}^{*}

) < c’(

α_{2}^{*}

), leading to 0 <

α_{1}^{*}

α_{2}^{*}

. In other words, when earnings management based on real activities manipulations is more intensely value-destroying, the younger manager’s desire to impress the labor market yields to her concerns regarding the anticipated value destruction in the second stage of her career. These results therefore run contrary to the standard career concerns model in that, when the value destruction of real activities manipulations is sufficiently intensive, managers will engage in fewer real activities manipulations when they are young.

Sub-case 2.4: 1 +

\frac{h_{0} + 2 h_{ε}}{h_{0} + h_{ε}}

≤ λ

In this case, c’(

α_{1}^{*}

) ≤ 0, leading to the corner solution of

α_{1}^{*}

= 0. In this situation, where earnings management based on real activities manipulations is extremely value-destroying, younger managers will refrain from engaging in any such activity because even low levels of this kind of earnings management will lead to a loss of utility.

2.4. Discussion

The key insight from our model is that the extent to which the predictions of the traditional career concerns model prevail in an earnings management setting (i.e., whether the manager undertakes more earnings management in their early years relative to their later years) is critically dependent upon λ, the rate at which the Period 1 earnings management reverses or leads to value destruction. Figure 1 provides a graphical depiction of this relation between λ and the equilibrium levels of earnings management at different career stages (i.e.,

α_{1}^{*}

and

α_{2}^{*}

As shown, for any value of λ less than λ* = 1 +

\frac{h_{ε}}{h_{0} + h_{ε}}

, the traditional career concerns effect dominates, meaning that managers engage in more income-increasing accruals or real activities manipulations when they are young. The opposite prediction holds, however, once the value of λ exceeds the tipping point, λ*. Beyond this point, the manager’s concerns over the negative impact of value destruction on her future wage dominate her desire to signal her quality to the labor market in the early years such that the younger manager will engage in less of this type of earnings management. Once λ reaches

1 + (h_{0} + 2 h_{ε}) / (h_{0} + h_{ε})

, these concerns become so prohibitive that the manager does not engage in any real activities manipulations in her early years.

We make two additional observations regarding the model. First, while the optimal amount of earnings management in Period 1,

α_{1}^{*}

, monotonically decreases with λ, the optimal amount of earnings management in Period 2,

α_{2}^{*}

, is a constant that is independent of λ.

Second, it is interesting to note that the magnitude of the tipping point, λ* = 1 +

\frac{h_{ε}}{h_{0} + h_{ε}},

is negatively related to the ratio of the precision of prior information to the precision of new information (

\frac{h_{0}}{h_{ε}}

). Intuitively, greater values of h0 relative to hε imply that priors are relatively more informative than new information, and thus new information does not weigh heavily in the market’s assessment of the manager’s quality. Accordingly, when the market has more (less) precise prior information and less (more) precise new information, there is less (more) propensity for the manager to engage in earnings management activities in order to influence the labor market’s perceptions. Accordingly, a smaller (bigger) value of λ* is needed to reach the tipping point.

3. Conclusions

We analytically investigated the impact of managers’ career concerns on their earnings management decisions. Building upon the classic work of Holmstrom (1982, 1999), we incorporated the reversing feature of accruals and the value-destroying properties of earnings management based on real activities manipulations to generate important predictions. The key result deriving from our model is that not all earnings management activities are created equal. Settings in which the magnitude of a particular earnings management reversal or the intensity of its ensuing value destruction is smaller yield outcomes that are similar to the traditional career concerns result wherein the manager will engage in more earnings manipulations in the early stage of her career. By contrast, settings in which earnings management based on real activities manipulations is relatively more value-destroying lead to opposite predictions compared to those of the classic Holmstrom model, as our model suggests that managers will engage in less real activities manipulations in their early career stage. Our findings are novel to the literature and help to bridge the disconnect between survey evidence documenting that executives prioritize implicit contracting (i.e., labor market-based career concerns) when making earnings management decisions (Graham et al., 2005) and the extant literature’s focus on explicit contracting as an explanation for earnings management.

Acknowledgements

We thank Ana Albuquerque, Bill Baber, K.J. Euske, Jennifer Francis, Denis Gromb, Felix Höppe, Joe San Miguel, DJ Nanda, Marco Trombetta, Moqi Xu, Neng Wang, as well as seminar participants at Bocconi, Rotterdam School of Management at Erasmus University, the 6th Accounting Research Workshop in Bern, the Stockholm School of Economics’ Summer Accounting Research Conference, the AAA meetings in NYC, and the AAA-MAS conference in Seattle for helpful comments on earlier versions of this paper.

References

Aharony, J.; Wang, J.; Yuan, H. Tunneling as an incentive for earnings management during the IPO process in China. Journal of Accounting and Public Policy 2010, 29, 1–26. [Google Scholar] [CrossRef]
Autrey, R. L.; Dikolli, S. S.; Newman, P. Career concerns and mandated dis-closure. Journal of Accounting and Public Policy 2007, 26, 527–554. [Google Scholar] [CrossRef]
Bohren, O.; Haug, J. Managing earnings with intercorporate investments. Journal of Business Finance & Accounting 2006, 33(5 & 6), 671–695. [Google Scholar]
Bowen, R. M.; DuCharme, L.; Shores, D. Stakeholders’ implicit claims and accounting method choice. Journal of Accounting and Economics 1995, 20, 255–295. [Google Scholar] [CrossRef]
Cheng, Q.; Warfield, T. D. Equity incentives and earnings management. Accounting Review 2005, 80, 441–476. [Google Scholar] [CrossRef]
Dechow, P.; Skinner, D. Earnings management: Reconciling the views of accounting academics, practitioners, and regulators. Accounting Horizons 2000, 14(2), 235–250. [Google Scholar]
DeFond, M.; Jiambalvo, J. Debt covenant violation and manipulation of ac-cruals: Accounting choice in troubled companies. Journal of Accounting and Economics 1994, 17, 145–176. [Google Scholar] [CrossRef]
Gillan, S.; Hartzell, J.; Parrino, R. Explicit versus implicit contracts: Evidence from CEO employment agreements. Journal of Finance 2009, 64(4), 1629–1655. [Google Scholar] [CrossRef]
Graham, J. R.; Harvey, C. R.; Rajgopal, S. The economic implications of corporate financial reporting. Journal of Accounting and Economics 2005, 40, 3–73. [Google Scholar] [CrossRef]
Greene, W. H. Econometric analysis, 3rd ed.; Prentice-Hall: Upper Saddle River, 1997. [Google Scholar]
Healy, P. The effect of bonus schemes on accounting decisions. Journal of Accounting and Economics 1985, 7(1–3), 85–107. [Google Scholar] [CrossRef]
Higgins, H. Do stock-for-stock merger acquirers manage earnings? Evidence from Japan. Journal of Accounting and Public Policy 2013, 32, 44–70. [Google Scholar] [CrossRef]
Holmstrom, B. Managerial incentive schemes: A dynamic perspective. In Essays in economics and management in honour of Lars Wahlbeck; Swenska Handelshogkolan: Helsinki, 1982. [Google Scholar]
Holmstrom, B. Managerial incentive problems: A dynamic perspective. Review of Economic Studies 1999, 66, 169–182. [Google Scholar] [CrossRef]
Key, K. G. Political cost incentives for earnings management in the cable television industry. Journal of Accounting and Economics 1997, 23(3), 309–338. [Google Scholar] [CrossRef]
Levitt, A. The Numbers Game; (Unpublished remarks); 1998. Retrieved from http://www.sec.gov/news/speech/speecharchive/1998/spch220.txt.
Levitt, A. Reclaiming the profession’s heritage. Paper presented at American Partners Meeting of KPMG LLP, Orlando, FL, USA, 20 November 2003; 2003. [Google Scholar]
Li, X. Real earnings management and subsequent stock returns; (unpublished working paper); HKUST: Hongkong, 2010. [Google Scholar]
Patten, D.; Trompeter, G. Corporate responses to political costs: An examination of the relation between environmental disclosure and earnings management. Journal of Accounting and Public Policy 2003, 22, 83–94. [Google Scholar] [CrossRef]
Ronen, J.; Yaari, V. Earnings management; (Springer Series in Accounting Scholarship); Springer Sciences + Business Media, LLC: Berlin/Heidelberg, 2008. [Google Scholar]
Roychowdhury, S. Earnings management through real activities manipula-tion. Journal of Accounting and Economics 2006, 42, 335–370. [Google Scholar]
Teoh, S. H.; Welch, I.; Wong, T. J. Earnings management and the long-run market performance of initial public offerings. Journal of Finance 1998a, 53(6), 1935–1974. [Google Scholar] [CrossRef]
Teoh, S. H.; Welch, I.; Wong, T. J. Earnings management and the underperfor-mance of seasoned equity offerings. Journal of Financial Economics 1998b, 50(1), 63–99. [Google Scholar] [CrossRef]
Vorst, P. Real earnings management and long-term operating performance: The role of reversals in discretionary investment cuts. The Accounting Review 2015. forthcoming. [Google Scholar] [CrossRef]

1	For an excellent and extensive academic summary of the earnings management literature, we refer the reader to Ronen and Yaari (2008). Prominent examples of non-academic sources of attention to the issue of earnings management pressures include the famous speeches by Levitt (1998, 2003).
2	Additional contracting-based motives for earnings management that have been examined include political cost considerations (Key, 1997; Patten and Trompeter, 2003), the establishment of more favorable equity and bond pricing terms (Teoh et al., 1998a, 1998b; Aharony et al., 2010; Higgins, 2013), the reduction of debt renegotiation costs (Bohren and Haug, 2006), and executive equity compensation (Cheng and Warfield, 2005).
3	One exception is the study by Bowen et al. (1995). These authors consider the impact of various stakeholders’ implicit claims, notably excluding managerial career concerns, by accounting for method choices (i.e., not accruals or real activities management per se).
4	Furthermore, Gillan et al. (2009) document that fewer than half of the CEOs of S&P 500 firms have comprehensive explicit employment agreements.
5	A representative example of an “event-driven” explicit incentive to manage earnings upwards is the case where an executive chooses to engage in income-increasing manipulations in order to avoid her firm’s violation of a debt covenant.
6	We assume that the manager’s stature remains static throughout her career. In other words, the model abstracts away from other incentives such as “promotion.”
7	Alternatively stated, in our setting one can interpret ỹt to be “unmanaged earnings” and $\tilde{x}$ t to be “reported earnings,” except in period 3 when ỹ3 and $\tilde{x}$ 3 are interpreted instead as the value added from the manager’s post-retirement board service.
8	See, e.g., Greene’s (1997) Theorem 3.6 on the marginal and conditional normal distributions.
9	For the sake of parsimony, we do not define the variance here because it is not relevant to our model.

Figure 1. The earnings management levels at equilibrium (α) versus the magnitude of accruals management reversal or intensity of value destruction of earnings management based on real activities manipulations (λ).