3. A Critical Realist Perspective on Earnings Management

For example, DeAngelo (1986) strongly rejects the hypothesis that managers who propose to take a public corporation private understate the corporation’s earnings prior to the management buyout, and Perry and Williams (1994) indicate that managers manipulate discretionary accruals downward during the year preceding a public announcement of management’s intention to go private.

For example, an influential article by Teoh et al. (1998b) indicates that during the year that a corporation goes public, initial public offering firms (IPOs) engage in more income-increasing (depreciation) methods. However, contrary to popular belief, Ball and Shivakumar (2008) report that IPOs do not inflate earnings and tend to report more conservatively “in order to meet the market demand for higher quality financials from public firms, and in response to public-firm regulation” (p. 346).

DeFond (2010) specifies, “abnormal accruals models suffer from the inherent limitation that we are unable to validate the accuracy of their predictions. For example, we are unable to verify whether our estimates of discretionary accruals are the result of management’s opportunistic accounting choices or just an artifact of the particular model we are using” (p. 404).

David Hume (1711–1776) was a Scottish philosopher, economist and historian during the Age of Enlightenment.

Collier (2005) provides the following example: “A dog barking may cause a squirrel to run up a tree, but it is not the case that every time a dog barks, a squirrel runs up a tree; there may be no squirrels around, or the squirrel may be blasé about dogs. There is nothing that always follows a dog barking or always precedes a squirrel running up a tree” (p. 328). Collier (2005) suggests that we should express natural laws “not as constant conjunctions—whenever a dog barks, a squirrels runs up a tree-but as tendencies: dogs tend to bark at squirrels, and squirrels tend to run up trees if dogs bark at them” (p. 329).

Bhaskar (2008a) states, “the real basis of causal laws are provided by the generative mechanisms of nature” (p. 3).

Bhaskar (2008a) proposed a theory where “statements of laws are tendency statements. Tendencies may be possessed unexercised, exercised unrealised, and realized unperceived (or undetected) by men; they may also be trans-formed” (p. 7).

As noted by Lawson (1997), “a statement of a tendency, according to its primary usage here, is not about long-run, ‘normal’, usual, or average outcomes at the level of events. Nor is it reducible to a counterfactual claim about events or states of affairs that would occur if the world were different. Indeed, it is not a claim about anything at the level of the actual course of events at all. Rather it is a transfactual statement about the typically non-empirical activity of a structured thing or agent; here transfactuals are not counter-factuals but take us to the level at which things are going on irrespective of the actual outcome. A statement of a tendency, in other words, is […] an unconditional statement about something non-actual and non-empirical. […] a statement of natural necessity without qualifications attached. […] about a power that is being exercised whatever events ensue” (p. 23).

Bhaskar (2008a) clarifies “that generative mechanisms must exist and sometimes act independently of men and that they must be irreducible to the patterns of events they generate is presupposed by the intelligibility of experimental activity” (p. 42).

Bhaskar (2008a) explains, “To ascribe a power is to make a statement about possibilities that may not be actualized and that are possessed by the thing whether or not they are known by men; so powers cannot be reduced to their exercise or our ignorance. Now, if powers are possessed by things that act in open systems, their exercise must be normically qualified and they must be seen as tendencies₁” (p. 223).

Bhaskar (2008a) explains, “the cause of a failure of a car to move when the gear is in neutral is not something distinct from and extraneous to the mechanism responsible for its normal motion” (p. 225).

Hartwig (2007) observe, “tendencies₁ are normically qualified powers. Such powers may be possessed unexercised, such that all people have the power to steal, but most refrain from exercising it. Tendency₂ is stronger notion of tendency, conveyed in the example: ‘Kleptomaniacs have a tendency to steal’. The second case implies an enduring predisposition towards a certain type of effect or behaviour. Bhaskar argues that in a closed system there is no difference between these two types of tendency: once their intrinsic conditions are satisfied they are enabled, and they then only require some triggering stimulus to be released or put in motion. The crucial difference between the two types of tendency is revealed in open systems. Here the enduring predisposition is related to the pre-existing satisfaction of intrinsic enabling conditions […]. Thus in an open system a tendency₁ will need to be stimulated. In other words, it only occurs under a restricted range of conditions” (p. 459).

“In order to apply any tendency₁ or normic statement we must know when the antecedent or stimulus conditions for the mechanism it designates are satisfied. But this does not warrant the prediction of the tendency’s₁ fulfilment, i.e. the consequent’s realization, which depends upon the system being closed, and in particular upon the non-intervention of countervailing causes” (Bhaskar 2008a, p. 223).

Fleetwood (2011) clarified: “of any thing, there is a set of intrinsic properties or intrinsic enabling conditions—denoted ic_w and ic_x—that must be satisfied if it is to have an exercised tendency₁ to do ø; […]. Of anything, there may be a further set of intrinsic properties or intrinsic enabling conditions—denoted ic_y and ic_z—that must be satisfied if it is to have an actualised tendency₂ to do ø” (p. 11).

Bhaskar (2008a) observed, “Offsetting causes are often assumed to be always extrinsic. But the cause of a failure of a car to move when the gear is in neutral is not something distinct from and extraneous to the mechanism responsible for its normal motion…. Now intrinsic offsetting causes may or may not directly interfere with the operation of the mechanism responsible for the satisfaction of the intrinsic enabling conditions. If they do, then we must say that the tendency₂ is no longer possessed.” (p. 225).

Bhaskar refers to tendency₆ using different subscripts in different places. Indeed, Bhaskar (2008b) refers to tendency₆ as “tendency_e” (p. 72) and Bhaskar (2010) refers to tendency₆ as “tendency_f” (Bhaskar 2010, p. 62).

Bhaskar (2000) argues that this tendency is stronger than most of the concepts of tendencies discussed herein; more specifically, from tendency₁ to tendency₆. In addition, he refers to tendency₈ as tendency_e*.

However, Fleetwood (2011) observes that the difference between tendency₆, tendency₇, and tendency₈ “is due to the nature of the systems the tendencies operate in” (p. 13) since all enabling conditions and all stimulating or releasing conditions are satisfied.

Sayer (2000) uses the following example to demonstrate that the same mechanism can produce different outcomes: the causal mechanism “economic competition can prompt firms to restructure and innovate or to close” (p. 15).

To demonstrate that different mechanisms can produce the same outcomes, Sayer (2000) explains that one can lose a job for a variety of reasons (p. 15) or to clarify, one can lose a job because of a variety of causal mechanisms.

For a discussion of the primary causes, see Chap. 2, Sect. 2.​4.

Here, I argue against the causal law of a constant conjunction model (whenever A occurs, earnings management occurs), although I cannot a priori exclude either the opposite (that earnings management causes something), the presence of other generative mechanisms that may cause something, or the absence of any causal law of a constant conjunction model type.

In the extant literature, discretionary accruals and earnings management are frequently used as synonyms (Kothari 2001). In alignment with conventional practice (Peasnell et al. 2000), I use the terms “managed accruals,” “discretionary accruals,” “unexpected,” and “abnormal accruals” interchangeably. Similarly, the terms “unmanaged accruals,” “non-discretionary accruals,” “expected,” and “normal accruals” are used interchangeably.

The Healy (1985), the DeAngelo (1986), the Jones (1991), the Dechow et al. (1995), and the Kothari et al. (2005) models are presented in Chap. 2, Appendix.

For an overview of the two approaches to calculate total accruals, see Chap. 2, Sect. 2.​3.​1. Precise information regarding cash flows and accruals has become available only after specific accounting principles became effective (Kothari 2001). For example, for US GAAPs, the Statement of Financial Accounting Standards No. 95 (FAS No. 95) issued in 1987 became effective for the annual financial statements of fiscal years ending after July 15, 1988. In 1992, the International Accounting Standards Board issued International Accounting Standard No. 7 (IAS No. 7), which became effective only in 1994 and requires firms to provide cash flow statements. However, local generally accepted accounting principles often do not require a statement of cash flows. Therefore, when investigating, for example, non-listed companies, it will be necessary to extrapolate cash flows from other statements.

For example, Jones (1991) assumes that nondiscretionary accruals depend on a change in revenues and the level of property, plants, and equipment (Bernard and Skinner 1996, p. 315). The model requires parameters to be estimated during a given period (i.e., the estimation period). Jones (1991) adopted a time-series approach to obtain separate firm-specific estimated parameters for each firm. Jones (1991) consistently excluded firms with less than 14 years of observations and regressed using the longest available time-series of data immediately prior to the “event” year (i.e., the year that earnings management is suspected). However, the Jones (1991) model and other models, such as the modified Jones model employed by Dechow et al. (1995), assumes that during the estimation period, no systematic earnings management is hypothesized. More examples can easily be provided. Healy (1985) assumes that non-discretionary accruals for each firm equal the mean accrual for all other firm-years in his specific portfolio sample. The adapted version of Healy (1985) by Dechow et al. (1995) assumes that non-discretionary accruals for each firm equal the mean of past years’ total accruals. Furthermore, DeAngelo (1986) uses the total accrual of the immediately prior period as a benchmark for the current accrual, absent income manipulation; i.e., the non-manipulated normal accruals value for the current period equals the total accrual value of the prior period.

Bhaskar (2008a) defines this mistake as the epistemic fallacy, “This consists in the view that statements about being can be reduced to or analysed in terms of statements about knowledge; i.e., that ontological questions can always be transposed into epistemological terms” (p. 26). To clarify, “this consists in confusing the ontological order with the epistemic order, priority in being with priority in deciding claims to being, the question of what has relatively underived (or independent) existence with the question of what entitles us to regard some kinds of statements as grounds for other kinds of statements, etc. In particular, the question of what is capable of independent existence must be distinguished from the question of what must be the case for us to know that something is capable of independent existence” (p. 242).

For the purpose of this study, however, it is not necessary to extend the analysis beyond tendency₅.

Certain widely accepted definitions were presented in Chap. 2, Sect. 2.​2. Specifically, earnings management can be defined as “a purposeful intervention in the external financial reporting process, with the intent of obtaining some private gain” (Schipper 1989, p. 92); alternatively, earnings management “occurs when managers use judgment in financial reporting and in structuring transactions to alter financial reports to either mislead some stakeholders about the underlying economic performance of the company or to influence contractual outcomes that depend on reported accounting numbers” (Healy and Wahlen 1999).

More specifically, financial statements are the responsibility of both managers and those responsible for corporate governance.

Being a manager involves holding a position or performing function within the organization, along with exercising managerial intent, discretion, etc. By virtue of being part of the organization, “management is vested with a set of causal powers that defines its nature” (Tsoukas 1994, p. 297).