Book Reviews

Abstract

Ewert and Wagenhofer's (hereafter, EW) monograph presents information economics-based research on earnings management, earnings quality, and conservatism. The monograph focuses on rational expectations models, in which stock price is a key factor in the manager's reporting decisions. EW largely ignore the related literatures on real effects and contracting. Notwithstanding the exclusion of these other perspectives, the authors are left with a large body of research to present and discuss.The monograph would make an excellent reading for a Ph.D. class, particularly when supplemented by other readings on contracting and/or real effects that discuss the same topics. (EW recommend Kanodia [2007].) The EW monograph not only presents a large body of important research, it also conveys a sense of how to conduct impactful research. The authors demonstrate an appreciation for the views of a wide variety of accounting researchers (including researchers who might identify themselves as theorists or empiricists) and accounting standard setters and regulators. The connections EW make between these views are particularly insightful.EW use the Fischer and Verrecchia (2000) model of earnings management as their starting point, focusing on linear equilibria, although noting that nonlinear equilibria have been used to explain observed patterns of earnings management (e.g., Guttman, Kadan, and Kandel 2006). Earnings management is personally costly to the manager. When that cost is common knowledge, the market perfectly undoes the manager's earnings management in the linear equilibrium. However, earnings management cannot be avoided—the manager engages in costly manipulation to credibly communicate true earnings. Under a nonlinear equilibrium, pooling is possible (two or more different true earnings amounts are reported as the same), preventing the market from perfectly undoing the earnings management. When the manager's cost of earnings management varies across managerial types and the manager alone knows his type, the market can no longer perfectly undo earnings management, even under a linear equilibrium.EW present many variations of the basic linear equilibrium model, including their own model of accounting and real earnings management (Ewert and Wagenhofer 2005). In their model, earnings management is personally costly to the manager, while real management is costly to the firm's shareholders. Tightening accounting standards can lead to less accounting earnings management, as one might expect, but can also lead to increased earnings management. Key determinants are the exogenous observable compensation weights the manager's compensation places on first- versus second-period earnings and the manager's weight on first-period stock price, which is not observed by shareholders. It is when these exogenous parameters foster relatively limited accounting earnings management that tightening accounting standards (increasing the manager's cost of earnings management) can result in increased accounting earnings management. An important idea is that tightening accounting standards increases the earnings response coefficient, which always increases the manager's incentive to engage in real earnings management but has an ambiguous effect on accounting earnings management.EW move on to discuss models in which the precision of accounting measurements is uncertain, which can help explain observed features of earnings management. In these models, earnings management is motivated by the desire to both mislead investors into thinking earnings are better than they are and the inference investors will draw about earnings precision from the report (e.g., Kirschenheiter and Melumad 2002). When compared to similar research in finance or economics on other topics, the paucity of research that formally tests predictions from models is a striking absence from the accounting literature on earnings management. Li (2010), which builds on the work of Kirschenheiter and Melumad (2002), is an exception.The monograph then discusses potential benefits to earnings management that can arise when earnings management is used to convey other private information managers have. A related contracting story that I particularly like is Demski (1998), in which earnings smoothing is used to convey a manager's acquired expertise. Before concluding the section on earnings management, EW discuss models that relax the assumption of investor rationality.The institutional justification for some of the assumptions employed in this broad class of earnings management models is unclear to me. For example, what is the justification for the particular information asymmetries assumed about compensation and/or the manager's shareholdings? Also, in the base model, the manager's role is limited to reporting. Arguably, the stewardship role of accounting is absent. (In fairness, even contracting models in which stewardship is central express a narrow view of stewardship when compared to Ijiri [1975].) Moreover, the distinction between disclosure and recognition is limited. In earlier models of disclosure, managers were not allowed to lie. In the later but closely related models of recognition, the main modeling difference seems to be that managers can lie at a personal cost. From our experience with employee stock option accounting, is the main difference between disclosure and recognition that disclosed amounts are more reliable?EW's inclusion of real earnings management and accrual reversals is a welcome dose of stewardship and accounting structure. However, I was left wondering if real and accounting earnings management are modeled as being too similar. Both types of earnings management reverse. Accounting earnings management reverses because of accrual reversals. Real earnings management reverses because manipulation to increase first-period earnings (e.g., skimping on R&D) is met with a more-than-offsetting decrease in second-period performance. In the model, the main difference is that the cost of earnings management is borne by the manager, while the cost of real management is borne by the firm's shareholders. In practice, are CEOs the ones who bear the costs of accounting earnings management? When earnings management is uncovered and penalized by the SEC or in private litigation, it seems instead to be the current shareholders who end up paying the vast majority of the fines and other costs.Before concluding their discussion on earnings management, EW provide a brief discussion of the related contracting literature. Since this section discusses Arya, Glover, and Sunder (1998), let me note a couple of minor points of disagreement. First, bankruptcy constraints on payments can be thought of as a restriction on the outcome set, which is allowed for by the Revelation Principle's assumptions (as long as the constraints do not depend on the agents' private information). Second, our paper is only partly a literature review, using the Revelation Principle in a way other than it was intended (as a solution technique) by placing the principal (rather than a disinterested mediator) in the position of mechanism designer and using the Revelation Principle's assumptions to classify models of earnings management. We also present two new models. In one, allowing for earnings management can reduce the cost of hiring a manager by relaxing his individual rationality constraint, because he faces less short-run scrutiny. Would you want to live in a glass house (unless it is built by Mies van der Rohe or Philip Johnson)? While giving the manager short-term cover, the accounting structure (accrual reversals) preserves information about long-run performance. In the second model, allowing for earnings management can provide better managerial incentives, because it limits the principal's (e.g., the board's) role in bailing out the manager if things go awry. Installing an accounting system that allows for short-term earnings management serves as a commitment device by the principal not to intervene ex post. The lack of intervention is anticipated by the manager and motivates him to do a better job himself.The second part of EW's monograph begins with a discussion of the standard setters' conceptual frameworks, noting that the IASB's interpretation of relevance is so broad that it is not useful in guiding standard setting. The section focuses on EW's recent research on earnings quality. The main point seems to be “that the information content of reported earnings cannot be derived from simply observing the mechanics of an accounting standard, but it depends on subtle interactions between accounting standards, earnings management, and rational inferences of investors” (p. 50). EW define earnings quality as a reduction in the market's uncertainty about cash flows due to reported earnings and then study how changes in other characteristics of the environment affect earnings quality, producing a mix of relatively straightforward and subtler results. The manager is assumed to value market prices, reported earnings, and smooth earnings and to find earnings management personally costly. Instead of describing such models as non-contracting models, one might describe them as reduced-form contracting models. Their model, which is also replete with accounting structure such as working capital accruals, is particularly useful in thinking through various definitions of earnings quality adopted in empirical studies (Subsections 3.5 and 3.6), although, as EW note, a disadvantage associated with any particular formalization is that “it is difficult to draw more general conclusions for variations in the model setting” (p. 73).In the last section of their monograph, the authors discuss recent research on accounting conservatism, including their own. EW distinguish between two approaches to modeling conservatism: (1) conservatism as a change in probability distributions, or bias, which has been widely adopted in many models, and (2) conservatism as a restricted message space. EW discuss implied P-E relationships, the contracting and non-contracting usefulness of conservatism, and interactions between conservatism and earnings management.Kwon, Newman, and Suh (2001) is an early managerial contracting model of conservatism. They show that, with a risk-neutral manager, conservatism reduces the expected cost of providing incentives, because the likelihood ratios are improved (the probability of obtaining a good outcome is reduced under both high and low effort). Glover and Lin (2013) show that, once accounting reversals are taken into account, the incentive effects of conservatism are neutralized unless the contracting relationship facilitates an “incentive spillback”—in that case, conservatism is again optimal. Gigler, Kanodia, Sapra, and Venugopalan (2009) study debt contracting and show that conservatism either does not matter or produces inefficiencies, because a conservative bias makes low reports less informative, challenging the popular view that conservatism protects debt-holders.EW note that practices conventionally described as examples of accounting conservatism such as asset write-downs are inconsistent with the definition of accounting conservatism as bias. The recent paper of Beyer (2012) is an important exception to the bias formulation of conservatism, using practice as the starting point for modeling conservatism. Gao (2013) is also noteworthy in its two-step representation of accounting measurement, which seems more descriptive of practice than the formulations employed in other models of conservatism. He develops the standard argument that conservatism arises in response to managerial opportunism.It seems to me that accounting conservatism is inherently a multi-period concept. We “don't count our chickens until they've hatched” (delayed recognition of good news in order to subject it to additional verification) and are quick to diagnose and report anticipated failures and the reason for the failure (early recognition of bad news accompanied by expanded disclosure). In ongoing work with coauthors, we show that such a definition of accounting conservatism can facilitate efficient liquidation and other decisions (for example, those normally built into or triggered by debt covenant violations), because of the early diagnostic information that conservatism provides about potential bad outcomes. At the same time, conservatism provides efficient stewardship incentives, because the slow recognition of good news facilitates an incentive spillback from the second period to the first. In a capital market setting (which is outside the scope of our work), one can imagine the delay of good news preventing an over-heated response that might provide managers and others with incentives to “take the money and run.”Fleischman, Funnell, and Walker's new book, Critical Histories of Accounting: Sinister Inscriptions in the Modern Era, is a welcome addition to the “Routledge New Works in Accounting History” series. The book opens with Funnell and Walker's concise and helpful introduction that situates the volume's papers within the wider accounting history and social science literatures. The authors also outline the four themes—annihilation, subjugation, exploitation, and exclusion—used to organize “eleven previously published, significant articles, authored by prominent critical accounting historians” (Introduction, p. 1). In outlining these themes, the introduction also draws out commonalities across the papers, commonalities that serve to explain the significance of the book's subtitle. Although drawing upon diverse historical episodes that span over two centuries and across several continents, each paper illustrates how a particular group used accounting practices (or professional accounting associations) to help accomplish objectives or plans that disadvantaged, exploited, subjugated or even killed others. These illustrations are further elaborated by Fleischman (Chapter 1) in a previously unpublished chapter. This chapter provides an overview of some theoretical debates that have occurred within the critical accounting history literature and offers a glimpse of the wide-ranging set of issues that have been addressed by critical accountants. The Introduction and first chapter successfully establish the aims of the volume as well as provide an entrée into the critical accounting literature for those who wish to explore it further.Funnell and Walker describe the volume's first theme as consisting of efforts to exterminate another group so thoroughly as to leave no trace. Given its frequent characterization as a neutral, calculative practice, accounting seems unlikely to be associated with such efforts. However, two chapters illustrate the possibly “essential contributions of accounting to governments and individuals who are determined to carry out policies of annihilation” (Introduction, 2). Funnell (Chapter 2) investigates the role of accounting in the bureaucratic apparatus established to enact the Nazi Final solution. Here, accounting served the important function of allowing administrators and bureaucrats to separate themselves from the consequences of their work by reducing individual, unique humans to numbers—the number killed or remaining to be killed, the transportation bills for shipments of persons to concentration camps, etc. Importantly, accounting reports were also employed as a technique to maintain the accountability of the SS personnel who sorted property confiscated from exterminated Jews. Funnell argues that the detailed inventories of such confiscated assets were seen to sanctify this activity by demonstrating that the confiscation was not motivated by “selfish greed but by the need to purify the German race” (p. 67).Lippman and Wilson (Chapter 3) also draw upon accounting practices used during the Holocaust to support their argument that the choices of accountants about what information to present as well as its form of presentation has important effects upon the decisions made by organizations and other entities. They demonstrate how reports presenting the revenues and expenses associated with the use of slave labor served to depict such slaves as productive inputs that could be used for the financial advantage of private German firms. Such reports failed to recognize the horrifying conditions under which the Jewish slaves labored even as the reports employed estimates of the useful life of slave laborers. As another example, the authors describe the cost/benefit analysis employed to reach the “rational” decision to burn children alive rather than to incur the additional costs of gassing them prior to incineration.Three chapters explore the volume's second theme. In contrast to policies of annihilation, policies of subjugation seek to subdue a population and often involve the efforts of an imperial power to gain access to, and control over, the property of others. Funnell and Walker (Introduction, 4) describe these chapters as showing the uses of accounting information to define the issues and problems associated with consolidating and extending imperial power as well as to “create political and social discourses … which would then inform and legitimize political debate and strategies essential to the preferred imperial political goals and beliefs.”The section opens with Graham and Neu's (Chapter 4, p. 85) study of the “mundane technologies of accounting” employed by the Canadian government in its efforts to convert the indigenous population to economic citizenship. The authors describe the different methods employed by local agents to convert observations about indigenous life into the reports submitted to government centers. Such reports were then used by politicians and others to assess the accomplishment of control over indigenous populations as well as the need for further actions. The policies enacted by the Canadian government were underpinned by a paternalism that presumed First Nation's peoples should be treated as “wards or children of the state” (p. 91) and stripped of their agency or subjugated. Graham and Neu show how the administrative reports of the era reflected back these beliefs by representing this governmentally imposed lack of agency as evidence of the continuing need for paternalism.Walker (Chapter 5) explores the social implications of accounting publicity in the administration of British poor laws. Although he notes that systematic accounting was used as more commonly thought—to control expenditures and prevent mismanagement of public funds—he also stresses its importance for controlling the pauper. Printed lists of paupers provided various details about such persons, including the ascribed reasons for relief (e.g., aged, bastard, cripple). In so doing, the lists often placed “an additional stigmatizing label on the recipient of relief” (p. 105). The lists also permitted surveillance of the poor by facilitating the capacity of community members to identify the “undeserving poor.” This identification of paupers also worked to distance the poor from their neighbors. The potentially stigmatizing effects of ascriptions, coupled with this distancing from neighbors, served a further purpose in discouraging some potential claimants from applying for aid and thereby reducing public expenditures for the poor.O'Regan (Chapter 6) illustrates how British government responses to the Irish Famine of the mid-1840s relied upon a “vast architecture of accounting.” This architecture included detailed lists to identify those capable and “deserving” of work; weekly reports were sent to Dublin, detailing the name of laborers, nature and quantity of work performed, and rate of payment (p. 118), and various accounting and internal control practices were introduced to ensure that payment was made only for work performed. O'Regan argues that these controls, reports, and numbers were employed to reform the perceived moral lapses of the Irish by replacing “indolence with industry, passivity with endeavour” (p. 116) as well as to discipline the behavior of those authorized to provide relief.Three chapters explore the connections between accounting and exploitation. Funnell and Walker (Introduction, pp. 6–7) describe exploitation as a moral concept, one that denotes taking advantage of another person in a way that degrades her/him. Hopper (Chapter 7) focuses on the exploitation of labor and outlines how four epochs of capitalism have shaped, and have been shaped by, management accounting and efforts to control labor processes. During the first epoch, detailed records of wages and the introduction of new systems for determining prime cost marked “the beginning of a transfer of financial knowledge from the worker to the factory owner … [which] helped make visible where returns might best be creamed off and labour be intensified and disciplined” (132). In the second epoch, Taylorism and the development of standard costs increased control over labor processes, as product jobs were redesigned and simplified to allow for the replacement of craftsmen with semi-skilled labor. In the third epoch, Fordism, new accounting techniques such as transfer pricing, flexible budgeting, and ROI measures accompanied the exploitation of economies of scale through increased firm concentration and vertical integration. The fourth epoch saw decreases in manufacturing jobs and union membership and rising numbers of low-wage workers, as flexible manufacturing, subcontracting of low-value-added activities, and other practices came to characterize workplaces.Oldroyd, Fleischman, and Tyson (Chapter 8) explore the role of accounting in the U.S. and in the British West Indies (BWI) slave economies. They maintain that their work shows accounting as “an instrument of both social control and social change” (p. 155). The paper argues that accounting helped to perpetuate the myth that slaves were less than human by classifying slaves as inventory alongside livestock such as cattle and mules. The authors also contend that accounting may have served a more protective role by highlighting the value of slaves and thereby contributing to incentives to maintain their health and well-being. For me, this argument falls flat, as similar statements might be made to argue that accounting contributed to the better treatment of other property like cattle. So, while accounting may have provided plantation agents with incentives to feed and house slaves sufficiently to ensure their continuing capacity to work, accounting would not have given incentives to provide slaves with the conditions to prosper as human beings (access to freedom, justice, etc.). However, the authors' illustrations of accounting's use as an emancipative force in the BWI provides clearer indications of the positive potentialities for accounting.Hooper and Kearins (Chapter 9) explore another case in which accounting and other expertises assisted in implementing policies to exploit and disadvantage an indigenous group. They document how such expertises were used to exploit the Maori of New Zealand and gain control over Maori lands while allowing politicians to “maintain a façade of sympathy” (p. 160) toward the Maori. Accounting facilitated this hypocrisy by distancing the actions associated with dispossession from the realm of active political decision-making. Dispossession was accomplished through the activities of experts who established the rents received and commissions charged for Maori lands held in public trust as well as determined the balances paid to, or withheld from, the Maori. Accounting statements of income and distributions captured the results of this activity by experts and subsequently were used to show that the lands held by the Maori generated low returns. This “evidence” was then used to justify the sale of these lands. The authors argue that “while accounts may convey a sense of facticity and neutrality, both these attributes are suspect. … Accounting acts as a key technology in achieving a redistribution of assets, particularly where accounts can be shown to reveal a poor return” (p. 167).In this final section, attention shifts away from accounting practices, reports, and techniques and toward the ways that professional accounting groups have employed practices of exclusion. Three chapters examine the various devices used to exclude groups such as women, African-Americans, and Afro-Caribbeans from membership in professional groups and from the practice of accounting. In their contribution to this section, Kirkham and Loft (Chapter 10) argue that the exclusion of women was fundamental to the professionalization of accounting in England and Wales in the period 1871–1931. The authors document a significant increase in the numbers of women employed as clerks and bookkeepers during this period even as women were excluded from professional accounting groups. They then convincingly argue that these parallel developments were associated with a discursive construction of bookkeeping as feminine and of accounting as masculine, a construction that facilitated the exclusion of women from the ranks of association members even as it served to establish accounting as a profession.Exploring the experiences of African-Americans attempting to gain entry into the U.S. profession during the years 1914–1965, Hammond and Streeter (Chapter 11) provide another vivid account of the exclusionary practices of the accounting profession. In order to qualify for the CPA exam (and licensure), most states required the completion of a one- to three-year apprenticeship with a practicing CPA. This requirement posed a significant obstacle for otherwise qualified African-Americans, as their employment opportunities were negligible at white firms. Indeed, by 1945, only nine African-Americans had earned their CPA designation. The authors focus on the experiences of individual African-Americans in order to illustrate the combination of “unique characteristics and individual effort” (p. 194) that characterized the success of the earliest African-American CPAs. Interview excerpts provide us with an opportunity to listen to these individuals describe their experiences and to provide a sense of the obstacles they overcame. For example, Bernadine Gines poignantly recounts her experience in searching for employment after completing an M.B.A. at New York University. During her first interview, she described how “One of the partners advised that he would not give me a job, but that his wife was looking for a maid if I knew anyone who wanted that job” (p. 195). Despite this unpromising beginning, Ms. Gines persisted, found employment, and eventually earned her CPA.Annisette (Chapter 12, p. 200) explores “the multiple ways in which race has been deployed to bring about majority under-representation and exclusion” in Trinidad and Tobago (T&T). In colonial T&T, accountancy was associated with white “Britishness.” Indeed in 1938, all six professionally qualified accountants were British nationals. With the expansion of the corporate sector during the colonial period, the race-based division of labor was further refined with the importation of British accountants to perform the more “expert” work, with the Coloured/Mixed category performing bookkeeping tasks and with African and East Indian segments of the T&T population excluded from accounting. British professional societies such as the ICAEW and ACCA contributed to this situation with requirements such as the completion of apprenticeships in Britain or the completion of “approved accountancy work.” These requirements effectively barred Afro- and Indo-Trinidadians from the accounting profession, given their limited opportunities for obtaining such employment. Following independence in 1962, the newly formed T&T professional association was eventually dominated by white chartered, practicing firms. Annisette details how various policies enacted by this association effectively continued earlier practices of racial exclusion, albeit using more subtle mechanisms. She argues that these actions illustrate how the “establishment of alternative routes to the profession can facilitate new patterns of exclusion as more accessible routes themselves become racialized” (p. 214).Although editors can always be taken to task for selecting article x rather than article y for inclusion in an edited volume, the reviewer's task is to evaluate whether the collection as a whole works to accomplish the volume's aims. Although the brief summaries provided here do not capture the complexity and nuances found in these articles, the summaries do provide ample evidence that these articles effectively elucidate the organizing themes of annihilation, subjugation, exploitation, and exclusion. Furthermore, the collection seems likely to achieve the purpose stated in the Introduction—an effort to interest a new generation of scholars in the “intellectual possibilities and rewards of doing research in the area” (Introduction, 1). The chapters provide a solid overview of the wide-ranging sets of questions that may be, and have been, addressed within accounting history research. The chapters also illustrate the variety of materials that may be employed as evidence in developing answers to such questions. Finally, this rich variety of questions and materials coupled with the quality of writing across the chapters also suggests creative challenges and opportunities in conducting accounting history research. As such, the editors have succeeded in producing an introduction to the possibilities for and rewards of accounting history research.Although introducing a new generation of scholars to such possibilities and rewards is an admirable aim, the contribution of the volume is not limited to this purpose and deserves a wider readership. The volume might also be usefully employed in accounting seminars at