تبعیض نژادی ضمنی علیه آمریکاییهای آفریقایی تبار در ارزیابی عملکرد کارت امتیازی متوازن

A dominant theme in critical accounting theory concerns the relation between human identities and accounting discourse and practices. Though this theme has strong antecedents in Marxist-inspired critique of ideology, research into this theme has employed diverse approaches; among them, genealogical studies (e.g., Miller and O’Leary, 1987), deconstructive studies (e.g., Shearer and Arrington, 1993), psychoanalytic studies (Roberts, 1991 and Roberts, 2009) and critical-rational studies (e.g., Power and Laughlin, 1996). We offer a different approach grounded in social-cognitive concerns with how implicit attitudes about race influence evaluation of others. We report on the results of an empirical, lab-based study of balanced scorecard evaluations and bonus allocations where race is a treatment effect and where the well-established tenets of Implicit Association Testing (IAT) are used to reveal that there are, indeed, propensities to unwillingly let racial prejudice intervene into our accounting-based evaluations of others. That intervention influences identity in ways that are morally unacceptable, degrading to black workers, and loaded with potential for negative material consequences for workers (e.g., less compensation due to racially determined and irrational performance evaluations). The paper illustrates one area of research in which methodologies adapted from conventional empirical, statistical approaches can enhance the emancipatory potential of critical accounting research.

To engage in a serious discussion of race in America, we must begin not with the problems of black people but with the flaws of American society – flaws rooted in historic inequalities and longstanding cultural stereotypes. Cornel West, Race Matters, p. 6 This study examines whether or not white American MBA and MS in Accounting students demonstrate implicit racial prejudice and whether or not such prejudice is present in evaluating the performance of managers in a balanced scorecard setting. The importance of the study derives from both the fact that accounting plays a significant role in the evaluation of persons for the quality of their work and the norm that the well-being of and judgment about a human being should not be influenced by the morally atrocious force of racial prejudice. Accounting evaluation, unlike many other forms of evaluation, has economic “muscle” in that accounting evaluations are used to allocate resources, to influence the nature of work, and to motivate certain kinds of behaviors that have powerful implications for the way in which day-to-day life is lived. Much research has been oriented toward several aspects of accounting's place in performance evaluation activities, though these aspects tend to be narrow, constrained to rational theories, and grounded in the desirability of capitalistic norms (e.g., treating workers as self-interested functionaries in need of control, incentives, and motivations). Such research has explored areas like the following: (1) the normative conditions that would make some forms of accounting more desirable than others (Merchant, 2006); (2) the salience of different types of accounting information to the performance evaluation process (e.g., Baiman and Rajan, 1995, Feltham and Xie, 1994 and Moers, 2005); and, (3) how different types of contracts and financial reward systems mediate both performance evaluation and performance quality (e.g., Bonner and Sprinkle, 2002, Ittner and Larcker, 1998 and Scott and Tiessen, 1999). In this study, we address performance evaluation in a balanced scorecard setting with experimental focus on race and the manner in which explicit and implicit attitudes about race influence performance evaluation decisions, and we do so in a manner more consistent with critical-theoretic and ethical concerns rather than functionalistic and capitalistic concerns. This study builds upon existing research in critical accounting theory which focuses on the ways in which accounting practices and institutions often enhance rather than diminish the morally outrageous consequences of racial prejudice. We will mention five such studies as indicative of this research stream. In a quite impressive case study, Davie (2005) demonstrates convincingly how State processes like affirmative action sustain rather than diminish racial prejudices which are deeply embedded in the discursive practices which secure the interests of capital. Accounting is exemplary of such discursive practices. In an accounting context, Kim (2008) illustrates how oral histories can and do reinforce hegemonic processes occlusive of the sort of understanding of racial, ethnic, and gender differences than would be emancipatory. Duff (2011) analyzes the photographic representation of blacks in Big Four annual reviews to provide strong evidence of a “homo-social,” white-male semiotics of human representation in those reviews, reviews which are retrograde from the standpoint of emancipatory understandings of diversity. In a similar vein, James and Otsuka (2009) provide empirical evidence of racial and ethnic discrimination against Chinese graduates of Australian universities among Australian accounting firms. James (2010) engages in a case study of cultural politics grounded in punk rock to display the commonalities, tensions, and paradoxes associated with the conflation of Marxist presumptions with racial identity. Our study is grounded in methods which some critical theorists would erroneously view as committed to a “positivistic” epistemology. That is because we use experimental design and statistical analysis of data in forming our argument. This is a category mistake. The use of a particular method does not in any sense entail a commitment to the view that the validity of a knowledge claim (the proper subject matter of epistemology) requires a particular “method”, nor does it entail that the method one uses is somehow presumed “superior” on epistemological grounds. This is nonsense, and making such a claim hoists critical theorists on their own petard – they are doing precisely what they accuse “mainstream” researchers of doing – promoting a particular approach to research as necessarily “better” in an epistemological sense. Research and scholarly inquiry is about making arguments, and methods should be accepted or rejected based upon how well or poorly they produce the best argument (the concern of rhetoric). Enough said; we should be well beyond tired and stale debates like this (see Arrington and Schweiker, 1992). Our approach opens accounting research to what is perhaps the most formidable domain of contemporary social-cognitive psychological research, an area concerned with attitudes in general and implicit attitudes more particularly. Implicit racist prejudicial attitudes are “… automatically activated and occur without full awareness upon exposure to a Black person” (Dovidio et al., 2009, p. 166). The paper examines the effect of both explicit and implicit attitudes about race on performance evaluation decisions and bonus allocations. An explicit attitude is one about which its holder is fully aware. A racist who knows and embraces his or her racism holds an explicit racist attitude. Though more nuanced definitions are available, in general an implicit attitude is one about which the person who holds it is not consciously aware that he or she does in fact possess it (see Petty et al., 2009, pp. 3–9 for extensive discussion of the meaning of implicit attitudes). If attitudes – like other personality attributes – influence judgments and decisions, including decisions about other peoples’ performance; and, if attitudes are implicit, then the outcome of accounting-based performance evaluations may be driven in part by unintended effects, effects that are both prejudicial at an ethical level and detrimental to organizational performance at a pragmatic level.1 Implicit racial biases and implicit gender biases are two outcomes that have been observed in a large number of studies (see Dovidio et al., 2009, Nosek et al., 2007b and Petty et al., 2009 for comprehensive reviews). This paper brings inquiry into implicit attitudes about race to bear upon performance evaluation in a balanced scorecard setting. The importance of understanding whether implicit and unintended racial attitudes influence performance evaluation goes without saying – legal, ethical, and political issues are all involved. The ethical implications for judging evaluators are very different across explicit racial bias and implicit racial bias. An explicit attitude of racial bias is one for which an evaluator is and should be accountable – he or she is simply a racist. Implicit racial bias is different: to the extent that ethical responsibility presupposes awareness, then implicit racial bias has less to do with ethical responsibility and more to do with the tacit consequences of the manner in which an evaluator's own identity has been constructed from within a social ontology where racial issues have been historically salient. The implicitly biased evaluator is simply unaware of that bias (see Banaji et al., 2003 and Messick and Bazerman, 1996).2 As mentioned, we investigate the nature of racial bias as manifest in the performance evaluation and reward judgments of managers in an experiment based on the balanced scorecard (BSC hereafter). The BSC is recognized as an important mechanism to facilitate performance evaluations in organizations. Its advocates also recommend linking BSC-based evaluations to reward systems (Kaplan and Norton, 1996). We measure racial attitudes and beliefs both implicitly and explicitly to investigate whether such biases influence these decisions.3 In our experiment, the racial identities (black or white) of two unit managers are manipulated. Participants, acting in the role of senior managers, evaluate the unit managers’ performances based on a comparison of common actual to target performance measures presented in a BSC format. All of our participants are white, and thus the validity of our results extends only to the racial attitudes of white Americans.4 In addition to performance evaluations, participants also allocate a fixed bonus pool of $20,000 between the two unit managers. Key dependent measures are the difference in the quantitative scores for performance evaluation of the two managers and the bonus allocations across the two managers. Participants then complete tasks that measure explicit and implicit attitudes about racial identity. Participants were evening MBA and MS in Accounting students from a U.S. southeastern university. The average age of participants was 32, and they had on average about10 years’ work experience. Thirty were male and 21 were female. Results show that differences in performance evaluations arise depending on the race of managers being evaluated. Specifically, when a white manager outperformed a black manager the difference in the two evaluations was greater than it was for either two white or two black managers. However, in sharp contrast, when a black manager performed better than a white manager a smaller evaluation difference was found compared to other pairings. We also find biases in performance among the four dimensions of the BSC, with greater bias appearing in the customer and the learning and growth dimensions. This seems consistent with some literature which suggests that the level of formality and objectivity of data may mediate implicit racial prejudice (see Johnson et al., 2008). Participants self-reported a favorable explicit attitude toward blacks, but implicitly measured attitudes revealed a preference for whites over blacks. Empirically, the strongest finding in this study is that the implicit attitude measure predicted performance evaluations and bonus awards when the white manager outperformed the black manager but not when the black manager outperformed the white manager. Implications of the study are twofold: (1) empirical evidence of implicit and biased attitudes toward blacks is observed in the study and influences both performance evaluations and bonus allocations; and, (2) given that an implicit bias for whites over blacks may be present in actual performance evaluation settings, management must seek out methods to mitigate this bias in addition to any measures mitigating explicit bias (e.g., diversity, EEO, etc.) to avoid both potentially costly discriminatory behavior and unjust moral degradation of the quality of life for some employees. Approaches to dealing with explicit bias (e.g., training sessions) may be ineffective for implicit bias. It seems that these sorts of implicit prejudices are very deeply grounded in the history of cultural practices and are thus not easily overcome. The remainder of the paper proceeds as follows. In the next section we discuss relevant literature on implicit bias and the BSC and propose our hypotheses. We then describe the experiment. Experimental results are then presented, and we conclude with a discussion of implications, limitations, and trajectories for future research.

Overall, the results of this study indicate that accounting practices related to performance evaluation may be influenced by implicit attitudes that yield some degree of racial bias in decisions. At a practical level, neither racial bias nor the consequences of such bias are ever acceptable; prejudices like these are simply reprehensible. Yet, the fact that prejudicial attitudes are implicit or tacit phenomena suggests that self-awareness may not be adequate to bring such prejudices to the level of consciousness for those who possess them. The force of the implicit within the ontology of selfhood is one example of just how complex and multidimensional human subjectivity is – our decisions and actions are influenced by personality factors embedded in cultural history in ways that do not rise to the level of consciousness for us. We should be vigilant in our awareness of and our efforts to eliminate the social reproduction of racial prejudice, particularly since it seems to operate in a tacit manner. Accounting discourse and practices focused on performance evaluation are powerful forces over the construction of subjectivity, of human identity. Whether we are aware of any racial prejudices or not is less important than the fact that our evaluations do in fact construct identity – economic identity, moral identity, social identity, and organizational identity. The force of ethics is not located in our consciousness; it is instead interpretable and understandable in terms of what we are doing to others in the act of accounting for them. If racial prejudice is a component of that construction of identity, then we have strong grounds for moral pause in reflecting on what accounting does to persons. For the most part, accounting research into performance evaluation has assumed rational and cognitive information processing models which facilitate inquiry into the relation between pieces of information and their effects on decisions (Hopwood, 2008). This study follows broad streams of IAT research to suggest that affective comportments of identity, even comportments about which we may be unaware, in this case implicit attitudes about race, are forces at work in performance evaluations of persons. As others have suggested (see for example Kennedy, 1993 and Kennedy, 1995) finding a way to mitigate bias is not a simple and obvious task. It is however a task of paramount practical, functional, and ethical importance. This study raises questions of relevance to future research and practice. The fact that implicit racial bias seems conditioned by task specificity – whether one is simply evaluating someone or whether one is allocating bonus dollars – is both practically relevant and quite appealing from an experimental perspective. What is it about bonus allocation that causes it to seemingly render implicit racial bias more innocuous? Other interesting possibilities have to do with differences in outcomes as one moves across the four dimensions of the balanced scorecard; it seems that the force of implied racial attitude is more salient for the two “softer” BSC categories (customer related and learning and growth) than it is for the other two categories (financial and internal business processes). The IAT and the scope of existing research suggests hundreds of ways in which the force of implicit attitudes is operative across a range of psychological, social, political, cultural, and ethical aspects of experience; the expanse of this research literature is represented well in Petty et al. (2009). Gender, sexual orientation, class-identity, ethnic-identity, religious-identity, and any number of categories of personhood which are historically painted with prejudicial and degrading force are interesting domains for meaningful accounting inquiry, and the level of development of this area in social psychology provides plenty of guidance on how such accounting inquiry might proceed. Accounting has historically disclaimed any force other than “neutral” and “objective” measurement of others – it is time that that risible myth be exposed for what it is. IAT research can go a long way in yielding such exposure. At a minimum, this study opens experimental possibilities for study of a richer, more psychologically and morally well-specified understanding of the relation between evaluators, accounting systems and processes, and those being evaluated. In addition to attitudes toward race, attitudes toward gender (see Gold et al., 2009), sexism, class, particular employee functions, and particular kinds of work may mean that performance evaluations, and other accounting-based decisions, are conditioned by a wide range of affective and largely unconscious comportments of identity and personality. Readers are encouraged to examine the hundreds of ways in which social-cognitive psychologists have deployed instruments like the IAT to move us closer to a phenomenologically rich understanding of just how complex evaluative judgments of others are. The fact that so very many people are judged routinely through accounting makes accounting a profoundly important area for this sort of research. The fact that the IAT is so well developed and so readily accessible means that there are few barriers to entry into this novel stream of accounting research. It is a particularly appealing approach for those interested in experimental studies aligned with the emancipatory interest that motivates critical-theoretic approaches to accounting. It is also an area where empirical research merges nicely with those critical-theoretic concerns for emancipation. Beyond the experimental possibilities, the capacity of this area of research to enhance more philosophical and qualitative approaches is rich indeed. From the outset of this paper, we have motivated our interest in experimental study of implicit racial prejudice from the standpoint of the social ontology of subjectivity, of the construction of identity. Social ontological concerns are almost never addressed in quantitative research studies, though they are ubiquitous to critical work. In this paper, we have tried to produce one example of how quantitative research might accommodate social-ontological concerns. Nevertheless, because the most important consequence of racial prejudice is what it means for the socially constituted identity of its victims, critical, qualitative and humanistic research approaches are perhaps of more value in future research efforts than are the quantitative approaches that we have employed. But accounting researchers interested in social ontology and the construction of identity can deploy arguments drawn from this “third wave” of research into prejudices and biases to clearly link the discourses and practices of accounting to many aspects of identity – race, sex and sexual preference, class, ethnicity, types of labor, geography, etc. If, as many have held, Modernity is the study of the individual, then a major aspect of postmodern inquiry – including this area – has to do with understanding the force of cultural practices and ideologies on what we have come to mean by the term “the individual.” Exposing prejudicial ways in which Modernity has constituted identity is essential to overcoming the pretensions of modern accounting, pretensions like claims to fairness, neutrality, objectivity, progress, etc. We encourage critical and “qualitative” researchers to welcome the ways in which implied attitude testing can enhance this “overcoming” agenda for critical work. To do so, we must all overcome tendencies to approach scholarship as if it could be described with binaries like “qualitative” and “quantitative.” Racism and racial prejudice are not unitary concepts, nor is their relevance limited to individuals. Largely because of the modern focus on both clarity of terms and on individuals, “racism” has not been traditionally understood in a way that captures its complexity, nor has the racism inherent in institutions been examined enough.22 Research agendas which focus on the complexity of racism (for example its embeddedness in other forms of oppression) can, like implied attitude studies, expose the “nonobvious,” “noncognitive,” somewhat invisible modes in which prejudice operates. In a study which engages in a more complex sense of racism and prejudice in the workplace than we address in this study, Fearfull and Kamenou (2006) provide an excellent example of how prejudice is embedded in institutions. That study is a model for those interested in more ethnographic approaches to the study of contemporary manifestations of prejudice in the workplace. A strong limitation of this study has to do with the fact that we only examined the attitudes of white Americans; and, even within that category, a nonrepresentative sample of all white Americans. While much evidence indicates that groups other than white Americans demonstrate implicit bias against blacks (Nosek et al., 2007b), it is certainly wrong both practically and experimentally to make any inferences about other groups in the context of this study. Future research in this area should select samples from other populations. Albert Einstein has been quoted as saying “[I]t is harder to crack a prejudice than an atom.” And that makes it rather naïve to assume that prejudice in areas like performance evaluation can be addressed only through appeals to conscious awareness and thereby remedied. Our implied attitudes are deeply engrained in tacit dimensions of our being, and attention to “the third wave” of research into prejudice can help us understand that guarding against the force and consequences of prejudice requires vigilance, both cognitive and moral vigilance. This seems particularly important in areas like performance evaluation where the force of prejudice has real material consequences (e.g., promotions, compensation, dismissal) for those who are the objects of such evaluations. The ubiquity of accounting, its force over the lives of millions of workers, seems among the most meaningful domains of experience for critical inquiry into the full range of effects of prejudice – implicit, explicit, or both.