اکتشاف رابطه بین انتخاب زبان در نامه مدیر عامل شرکت به سهامداران و اعتبار شرکت ها

This paper proposes a taxonomy to assist in more clearly locating research on aspects of the association between corporate reputation and corporate accountability reporting. We illustrate how our proposed taxonomy can be applied by using it to frame our exploration of the relationship between measures of reputation and characteristics of the language choices made in CEO letters to shareholders. Using DICTION 5.0 software we analyse the content of the CEO letters of 23 high reputation US firms and 23 low reputation US firms. Our results suggest that company size and visibility each have a positive influence on the extent to which corporate reputation is associated with the language choices made in CEO letters. These results, which are anomalous when compared with those of Geppert and Lawrence (2008), highlight the need for caution when assessing claims about the effects on corporate reputation arising from the language choice in narratives in corporate annual reports.

We respond to the challenge by Adams (2008) to refine knowledge of the extent to which corporate reporting influences corporate reputation. We begin by proposing a taxonomy of prior studies that have analyzed the association between corporate reputation in a corporate accountability reporting context. Influenced by Schwaiger's (2004) contention that any identification of factors influencing corporate reputation should be based on reliable empirical evidence, we then draw upon our proposed taxonomy for two purposes. First, to demonstrate how our research enquiries can be located and understood; and second, to contribute empirical evidence that will help to develop a better understanding of the relationship between the word choice in annual report CEO letters to shareholders and corporate reputation. A particular feature of this paper is the juxtaposing of our results with those reported in a study in which Geppert and Lawrence (2008) identified some language factors likely to influence the level of corporate reputation. Thus, theoretically and empirically, we explore the extent to which language choice in a corporate reporting and accountability narrative influences corporate reputation. This paper builds on the work of Geppert and Lawrence (2008) who measure narrative disclosures in corporate reports as a proxy for corporate reputation. First, we contextualize the Geppert and Lawrence (2008) study, and other similar papers, in the form of an overarching taxonomy to apply to research linking corporate narrative disclosures and corporate reputation. Second, we replicate the Geppert and Lawrence (2008) study to the extent possible, and extend that study, where feasible, given our contradictory findings. We were concerned about the findings they reported because extant research into the written discourse of CEOs prompted doubts about the reliability of those results. In particular, their findings that high reputation firms use less complex (shorter), less varied and more concrete words appeared simplistic. We believe that the association between CEO text and corporate reputation is much more subtle and nuanced than advanced by Geppert and Lawrence (2008); and that such association is dependent heavily on company context. Organizations are “discourse constructions” (Fairhurst & Putnam, 2004). As such, it seems appropriate to study (as we do here) how the language used in corporate reports might construct corporate reputation. In particular, we are intrigued by the implications of results reported by Geppert and Lawrence (2008): that is, that a high reputation might accrue to a company simply from adopting a communications (or accountability) strategy that involves a more relaxed writing style, more desirable words (such as those associated with honesty, virtue and self-sacrifice), more present tense verbs, and a focus on immediate concerns of everyday life. As part of our enquiries, we replicate the Geppert and Lawrence study, to the extent possible, using (as they did) data sourced in Fortune magazine's annual America's Most Admired Companies survey, and in CEO letters to shareholders in annual reports of major US companies. However, unlike Geppert and Lawrence, we found no statistically significant associations between chosen language-related variables and corporate reputation. The important influence that corporate communications can exert on corporate reputation has been highlighted in argument that “communications was one of the ‘six key drivers’ of corporate reputation” (the others were competitive effectiveness, marketing leadership, customer focus, familiarity/favorability, and corporate culture) (Greyser, 1999, p. 179). In an increasingly competitive global economy, companies are keen to identify the drivers of corporate reputation so that they can acquire a sustainable competitive advantage (Schwaiger, 2004). Thus, it is not surprising that corporate reputation, and how it can be sustained and improved, has become an important element in the strategic communications of companies (Dolphin, 2004). Corporate annual reports (and the CEO shareholder letters they contain) are used not only for accountability purposes but also to create corporate reputation, corporate image and corporate credibility. However, corporate reputation is not synonymous with either corporate image (Dutton & Dukerich, 1991, p. 537) or corporate legitimacy (Bebbington, Larrinaga-González, & Moneva-Abadía, 2008b). Whereas image describes attributes members of a firm believe outsiders use to distinguish a firm, reputation describes the actual attributes outsiders ascribe to a firm. Following Schwaiger (2004, p. 48), a reputation is formed based on information about a firm's relative positioning in an organizational field, on marketing and accounting signals concerning performance, on institutional signals concerning social norms, on signals concerning firm strategy, and on attributes based on a firm's past actions. Because reputation is based on information and signals about a firm, corporate reporting is an ideal medium for communicating that information and those signals. Gotsi and Wilson (2006) tease out the differences between the dynamic concepts of corporate reputation and corporate image. They conclude that there is a bilateral relationship between corporate reputation and corporate image. Corporate reputations are largely dependent on everyday impressions people form of companies, based on firm behavior, communication and symbolism. Different stakeholders are likely to assess corporate reputation differently, depending on their economic, social and personal background. Gotsi and Wilson (2006, p. 29) conclude with a definition of corporate reputation as: …a stakeholders’ overall evaluation of a company over time. This evaluation is based on the stakeholder's direct experiences with the company, any other form of communication and symbolism that provides information about the firm's actions and/or a comparison with the actions of other leading rivals. Analysis and debate in accounting and accountability reporting literature about matters of corporate reputation have usually been conducted within a framework of legitimacy theory – and the need to build and maintain organizational legitimacy. Deephouse and Carter (2005) argue that organizational legitimacy focuses on a social acceptance that is derived from conforming to social norms and expectations – whereas organizational reputation involves social comparison. He and Baruch (2010) consider organizational identity and organizational legitimacy simultaneously. They observe that organizational identity is not fixed but depends on the social context in which organizational identity is narrated. They maintain that legitimacy can be created, maintained and repaired by choice of environment and social referents; and, importantly in the context of the present study, they identify external communications (such as impression management tactics, verbal accounts and stakeholder information disclosure) as the means of achieving this (He & Baruch, 2010, p. 44). Table 1 compares the three concepts of reputation, image and legitimacy. In the present context, it is important to acknowledge that managers use their annual reports “to construct and maintain desired images … and thus to shape the interactional frames from which reputations emerge” (White & Hanson, 2002, p. 291) – this is what Abrahamsson, Englund, and Gerdin (2011, p. 347) refer to as managers’ “desired future image.” Table 1. Organizational reputation, image, and legitimacy. Concepts Key aspects Time dimension Definition Organizational reputation • Whole organization • Short-term A general, temporally stable, shared evaluative judgment about a firm • Quality • Stable • Evaluation Organizational image • Aspect of the organization (e.g., investment image) • Short-term A dynamic perception of a specific area of organizational distinction • Quality • Dynamic • Evaluation Organizational legitimacy • Whole organization or industry • Long-term A shared general judgment about normative appropriateness • Social norms and rules • Stable • Appropriateness Reproduced from Brennan and Merkl-Davies (in press) and adapted from Table 1 of Highhouse, Brooks, and Gregarus (2009, p. 1487). Table options In the context of CEO letters to shareholders, we should also be alert to the possibility that a halo effect operates: that the reputation of the CEO extends to exert a positive reputation effect on the company as well. That possibility is consistent with the findings of an eight country study by Kitchen and Lawrence (2003) that CEO reputation is the third most important factor influencing corporate reputation; and with the view that CEO reputation and corporate reputation are intertwined increasingly, particularly as the CEO is the chief communicator for the corporation.

The empirical study we have conducted is important because of its capacity to inform corporate communications and narrative corporate reporting practice by inferring credibility (or otherwise) to prior reported research-based generalizations (Eden, 2002). Our study is important in helping to preserve the integrity of the cumulative empirical foundation of corporate communications knowledge (Hubbard & Vetter, 1991). The importance of studies such as this was emphasized in an editorial in the Academy of Management Journal which stressed that “the value of empirical management research is profoundly augmented … by encompassing a large number of high-quality replication studies” ( Eden, 2002, p. 841). Although our sample size is larger than Geppert and Lawrence (2008), we nonetheless accept that it is based on a relatively small sample size. We acknowledge Bamber, Christensen, and Gaver (2000) and express caution in generalizing the study results. The results we report raise doubts about the reliability of the findings reported by Geppert and Lawrence (2008). One explanation for the difference is a “maturation effect.” The Geppert and Lawrence study used data for 2001/2002 whereas data used in the present study are for 2006/2007. Mäkelä and Laine (2011), in their longitudinal study of one company, find evidence of change in reporting style over time, suggesting that a replication for a different time period can be expected to lead to differing results. Indeed, conceivably, the American business environment following the implosion of Enron in December 2001, and the enactment of the Sarbanes-Oxley Act 2002, was one of greater respect for corporate governance and ethical propriety. This possibly rendered corporate reputation less susceptible to corporate communication strategies that rely on the words CEOs use in letters to shareholders. Another explanation is that Geppert and Lawrence (2008) have studied firms that are more likely to be at the extremes of a high reputation/low reputation ranking continuum. In contrast, the present study classifies those in the upper quartile of a reputation ranking as high reputation, and those in the lower quartile as low reputation, thereby potentially diluting the results observed. A further explanation is that the present study is focused tightly on the top 100 Fortune companies, whereas the Geppert and Lawrence (2008) sample is drawn from companies spread across the top 500 Fortune companies. Although Geppert and Lawrence (2008) provide only partial sample selection details, 10 companies in their high reputation sample group can be identified: these range from Home Depot (ranked 18 in the Fortune 500 listing in 2002) to Harley Davidson (ranked 466 in the Fortune 500 list for 2002). Five of the 10 identifiable companies are ranked between 100 and 500 in the Fortune 500 listing. A similar profile is evident in the seven companies identifiable in the low reputation sample group of Geppert and Lawrence (2008). The difference in results seems explainable by Geppert and Lawrence (2008) having analyzed US companies that are smaller and less prominent than those analyzed here. Thus, the results suggest that the extent to which corporate reputation is associated with the textual characteristics of CEO letters may be influenced by company size and visibility. Whether there is a relationship between the words in CEO letters and corporate reputation is an important matter for corporate communicators – and one that deserves fuller analysis based on longitudinal data, a sounder theoretical base, stronger statistical design, and a commitment to analyze text in specific corporate context. However, further research should not rely solely on software-based quantitative analysis of words in CEO letters, but should embrace qualitative techniques (such as close reading) as well. A combination of quantitative and qualitative techniques will be conducive to the provision of valuable, comprehensive and potentially mutually reinforcing insights to, and understandings of, corporate reputation. Explanations for discretionary narrative disclosures as a proxy for corporate narratives also deserve more attention. Are disclosures for the purpose of providing shareholders and stakeholders with incremental useful information, or are they motivated by managers creating impressions that influence stakeholders’ perceptions of firm reputation? It is likely that the association between CEO text and corporate reputation is much more subtle and nuanced than Geppert and Lawrence (2008) would lead us to believe, and dependent heavily on company context. Several avenues for further research arise. First, it would be beneficial to explore the extent of the relationship, if any, between reputation and the motive for discretionary narrative disclosures. There would seem to be strong intuitive grounds to expect the relationship to be stronger where the overriding objective is to engage in impression management, rather than to provide incremental information. Second, our results highlight the need for further inquiry into the extent to which firm size is a mediating variable between corporate communications and corporate reputation. Generally, the effect of language choice on reputation and on the effectiveness of accountability in corporate annual reports warrants on-going enquiry. To that end, the taxonomy we propose should be facilitative.