Top management team social interaction and conservative reporting decision: A language style matching approach
Introduction
“The words we use in daily life reflect who we are and the social relationship we are in. This is neither a new nor surprising insight. Language is the most common and reliable way for people to translate their internal thoughts and emotions into a form that others can understand.”
— Tausczik and Pennebaker [1]
The upper echelons theory in the management literature suggests that corporate outcomes, including strategic decision choices and organizational performance, are mainly the reflections of values and cognitions from a top management team (TMT) [2,3]. Due to the bounded rationality of individuals, TMT members with diverse characteristics and backgrounds may behave differently while making corporate decisions, resulting in different business strategies and organizational outcomes [2,4,5]. To understand why organizations act as they do, we note that the prior literature has shown a sustained interest in investigating how TMT characteristics, such as social interactions among TMT members, can affect organizational outcomes (e.g., [[6], [7], [8], [9], [10]]). However, measuring interpersonal interactions that reflect stylistic differences remains a challenge for researchers due to the lack of direct observability [2,11]. To assess how an individual interacts with others in the process of socialization, one must analyze words and language, which are essential aspects of psychology and communication channels. In fact, what an individual says reflects the way he or she thinks. The degree of synchronized language style behavior, defined as verbal mimicry, is a critical component of human social interactions that affects group dynamics and the decision-making process [12,13]. If people are matched in their language styles, this suggests that their cognitive and psychological states are in harmony. Thus, verbal mimicry helps to strengthen individuals' group identification and participation efforts, which in turn improves group decision quality and performance [13,14]. To shed light on how cognitive styles affect corporate decisions through interpersonal communications among TMT members, this study measures executives' linguistic styles and uses them to represent the degree of TMT social interactions. The study investigates whether firms will adopt a more conservative reporting practice, ceteris paribus, when TMT social interaction is greater.
Many managerial characteristics are not directly amenable to being observed by researchers. The most common methodology in previous studies to obtain data on TMT characteristics has been surveys, which vary due to (1) the use of inconsistent measures that are linked with different research questions, (2) relatively small sample sizes that raise concerns about the generalizability of the findings, and (3) the difficulties of replicating data that can be extended to future research [3,15,16]. With the prevalence of technology, many more executive speeches have become publicly available, and with the rapid advancement in natural language processing techniques, extracting valuable information from unstructured texts for data analysis has become feasible [17,18]. Researchers have started to collect publicly available executives' speech transcripts during earnings conference calls and to utilize computer-assisted psycholinguistic measures to investigate how language affects group decision-making and firm valuation [12,13,19].
We collect earnings conference call transcripts and estimate a text-based measure of managers' speech patterns to construct a language style matching (LSM) score between two executives who are accountable for financial reporting quality, namely the Chief Executive Officer (CEO) and the Chief Financial Officer (CFO). LSM measures verbal mimicry, and it has been used in prior literature to measure social interaction level [[12], [13], [14]]. We examine the impact of LSM on corporate accounting reporting, focusing on accounting conservatism. Accounting conservatism requires firms to use stricter standards for recognizing bad news as losses than for recognizing good news as gains. As a result, conservative accounting reporting means that firms will recognize bad news reflecting losses quickly but recognize good news that is associated with gains in a more prudent way [20,21]. TMT with similar language styles, suggesting a better communication style alignment, facilitates closer social interactions between members during the process of scanning, transmitting, analyzing, and acting on new information. A better communication style alignment helps to reduce communication costs as well as forming a stronger identity of members' values or cognitions as a decision-making group. Consequently, TMT members are likely to form a united attitude toward making corporate decisions efficiently, such as accounting reporting decisions for dealing with bad news versus good news.
Based on 10,531 unique firm-year observations from 2007 to 2014, the results show that managers' social interaction level, as measured by LSM scores, is positively associated with accounting conservatism. The findings support the notion that, with the alignment of linguistic style, better social interactions between TMT members facilitate the establishment of group cognitive style that plays a vital role in the process of group decision-making and leads to more conservative reporting behavior when a firm is responding to bad news that have negative effects on firm value. In addition, the positive association between managers' social interactions and accounting conservatism is more pronounced when a firm operates in a highly competitive business environment, there is severe information asymmetry during the stakeholders' decision-making process, and when it is under financial distress.
The study contributes to our understanding of managers' social interactions in the upper echelons of corporations based on linguistic style similarities that reflect the cognitive base of TMT members as a group for making corporate decisions. The findings of the study contribute to the literature in the following three ways. First, the study employs a novel psychological text-mining approach, LSM, to capture the cognitive aspect of TMT social interaction and measure its influence on managers' decision-making behavior [1,13,14]. LSM measure goes beyond the traditional ways of using either individuals' demographical characteristics or analyzing interview or questionnaire survey results to assess managers' personalities and psychological propensities. Publicly available earnings conference call transcripts provide rich data for researchers to conduct a large-scale linguistic style analysis and address the previous research challenge due to data availability, which limits the measurement of managerial cognition [8,9]. Second, the study echoes the call in the prior literature to take into consideration the effect of cognition in strategic decision making [11]. By investigating the effect of managerial cognition through language styles, a more substantive construct for social interaction, the study contributes to the literature in upper echelons theory by documenting the empirical evidence to supplement prior theoretical work on a deeper construct to measure the values and perceptions of TMT and how it affects group cognitions and behaviors [3]. Third, the prior literature in accounting provides explanations and justifications like contracting and strategic choices for the firm's conservative reporting behavior [20,[22], [23], [24]]. The study builds on the decision making, information systems (IS), and accounting literature by employing a computer-mediated, text-based linguistic style technique to examine the cognitive aspect of TMT social interaction in a formal business event setting, namely earnings conference calls. Specifically, the study analyzes how such interactions affect corporate accounting reporting decisions in processing good versus bad corporate news. The empirical findings of the study fill the gap on how executives' cognitive values collectively as a group, measured by the degree of language style similarity, affect a firm's conservative reporting decisions.
The remainder of the study is structured as follows. We first review the theoretical background on TMT social interaction and corporate accounting reporting, followed by hypotheses. We then introduce the methodology, research sample, and model specification. The results section presents our findings and discusses implications. Conclusions are provided in the last section.
Section snippets
TMT values and cognitions
The organization is a reflection of its top managers' visions and cognitive schema [3]. The upper echelons theory proposed by Hambrick and Mason [2] suggests that the psychological and observable characteristics of TMT members affect how they make decisions and in turn have links with organizational performance. Whenever internal or external stimulus occurs, TMT members will experience the critical decision-making process involving scanning and analyzing relevant information, forming
Hypotheses
Homogeneous teams make decisions more quickly than heterogeneous teams [2,3,25]. Language is a critical component of human social process [42]. Mimicry in language, namely verbal mimicry, leads to partiality and increases individuals' incentives to engage with others. Verbal mimicry shapes the social interaction as language is exchanged to form consensus and coordination between individuals [12]. As top managers are social beings, we argue that the degree of TMT social interactions, measured by
Language style matching
To measure the cognitive aspect of TMT social interaction, we conduct a text-based linguistic style analysis using an LSM approach to investigate the similarity level of language styles used by executives, specifically CEOs and CFOs, during earnings conference calls. We collect earnings conference call transcripts as our data source. The reasons are as follows. As an increasingly important disclosure channel used by firms, earnings conference calls provide the fundamental financial information
Empirical results and discussions
Table 3 presents the descriptive statistics. LSM score (LSM) and earnings (EPS) are all left-skewed, whereas annual stock return (R) is right-skewed. The median of the indicator variable for negative stock return (D) is 0, indicating that more than half of the stock return is positive in our sample.
The correlation matrix among variables of interest is presented in Table 4. For brevity, we focus on the Spearman rank-order correlations, as the two correlations are generally consistent with each
Conclusions
The study investigates the linkage between managers' social interactions and reporting decisions by employing a text-based LSM method. LSM measures the degree to which two people in conversation coordinate by matching their language styles, which reveals important cues to understand the cognitive aspect of social interactions between individuals. To the best of our knowledge, the study is the first to address managers' cognitive values as a group by analyzing linguistic styles in understanding
Acknowledgement
Ting Zhang acknowledges the supports from the National Natural Science Foundation of China (Grant No. 71772150) and Young Talent Recruiting Plans of Xi'an Jiaotong University (Grant No. GL1J009).
Ting Zhang is a Ph.D. student at the School of Management in Xi'an Jiaotong University. Her research focuses on revealing the value of unstructured data (e.g., financial reports, conference call transcripts, online reviews and social media) by combining machine learning and econometrics.
References (66)
- et al.
Upper echelons research revisited: antecedents, elements, and consequences of top management team composition
J. Manag.
(2004) - et al.
Exploring top managers’ innovative IT (IIT) championing behavior: integrating the personal and technical contexts
Inf. Manag.
(2014) - et al.
Does mutuality matter? Examining the bilateral nature and effects of CEO–CIO mutual understanding
The Journal of Strategic Information Systems
(2016) - et al.
A text analytics framework for understanding the relationships among host self-description, trust perception and purchase behavior on Airbnb
Decis. Support. Syst.
(2020) - et al.
Financial fraud detection using vocal, linguistic and financial cues
Decis. Support. Syst.
(2015) The conservatism principle and the asymmetric timeliness of earnings
J. Account. Econ.
(1997)The contracting benefits of accounting conservatism to lenders and borrowers
J. Account. Econ.
(2008)- et al.
Inherent limitations of demographic proxies in top management team heterogeneity research
J. Manag.
(1999) - et al.
Mining the voice of employees: a text mining approach to identifying and analyzing job satisfaction factors from online employee reviews
Decis. Support. Syst.
(2019) - et al.
A linguistic signaling model of social support exchange in online health communities
Decis. Support. Syst.
(2020)
Acquisition profitability and timely loss recognition
J. Account. Econ.
Corporate suppliers and customers and accounting conservatism
J. Account. Econ.
Institutional ownership and conservatism
J. Account. Econ.
Estimation and empirical properties of a firm-year measure of accounting conservatism
J. Account. Econ.
Knowledge, compensation, and firm value: an empirical analysis of firm communication
J. Account. Econ.
Do financial market developments influence accounting practices? Credit default swaps and borrowers’ reporting conservatism
J. Account. Econ.
Asymmetric timeliness of earnings, market-to-book and conservatism in financial reporting
J. Account. Econ.
The psychological meaning of words: LIWC and computerized text analysis methods
J. Lang. Soc. Psychol.
Upper echelons: the organization as a reflection of its top managers
Acad. Manag. Rev.
Internal determinants of corporate social performance: the role of top managers
Acad. Manag. J.
Those closest wield the sharpest knife: how ingratiation leads to resentment and social undermining of the CEO
Adm. Sci. Q.
Antecedents of IS strategic alignment: a nomological network
Inf. Syst. Res.
Top management team functional diversity and organizational innovation in China: the moderating effects of environment
Strateg. Manag. J.
Making use of difference: diversity, debate, and decision comprehensiveness in top management teams
Acad. Manag. J.
How information technology governance mechanisms and strategic alignment influence organizational performance: insights from a matched survey of business and IT managers
MIS Q.
An empirical test of cognitive style and strategic decision outcomes
J. Manag. Stud.
Language style matching as a predictor of social dynamics in small groups
Commun. Res.
Take their word for it: the symbolic role of linguistic style matches in user communities
MIS Q.
Linguistic style matching in social interaction
J. Lang. Soc. Psychol.
The effect of social capital of the relationship between the CIO and top management team on firm performance
J. Manag. Inf. Syst.
Textual analysis in accounting and finance: a survey
J. Account. Res.
Conservatism in accounting part I: explanations and implications
Account. Horiz.
The information role of conservatism
Account. Rev.
Cited by (6)
How customized managerial responses influence subsequent consumer ratings: The language style matching perspective
2024, Decision Support SystemsDoes the new entrant eat my pie or enlarge my pie? Market entry investigation in the online-to-offline on-demand context
2023, Decision Support SystemsThe Effects of Top Management Team External Social Capital on Firm's Openness: Exploring the Mediating Role of Strategic Knowledge Integration
2023, IEEE Transactions on Engineering ManagementCOGNITIVE DIVERSITY AT THE STRATEGIC APEX: ASSESSING EVIDENCE ON THE VALUE OF DIFFERENT PERSPECTIVES AND IDEAS AMONG SENIOR LEADERS
2022, Academy of Management AnnalsAn Unsupervised Algorithm for Qualitative Coding of Text Data: Artifact Design, Application, and Evaluation
2021, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Ting Zhang is a Ph.D. student at the School of Management in Xi'an Jiaotong University. Her research focuses on revealing the value of unstructured data (e.g., financial reports, conference call transcripts, online reviews and social media) by combining machine learning and econometrics.
Fang-Chun Liu is an Assistant Professor of Accounting at Muma College of Business, University of South Florida. She received her Ph.D. in Business Administration from Temple University. Professor Liu's research interests include: strategic cost and performance management, corporate governance, human capital, executive compensation, financial reporting quality, and the financial implications of information technology investment. Professor Liu has presented her research actively at the most prestigious international conferences in Accounting and Information Systems areas. Her work has published in International Journal of Operations and Production Management, Information & Management, Decision Sciences, and Benchmarking: An International Journal among others.
Baojun Gao is a Professor of Management Science at Wuhan University. He received his B.E., M.Sc., and PH.D. from Xi'an Jiaotong University. His research interests are in the areas of text analytics, information systems and social media. His research has appeared in Decision Support Systems, Information & Management, IEEE Transactions on Engineering Management, IEEE Transactions on SMC: Systems, Tourism Management, Electronic Commerce Research and Applications, China Economic Review, and Service Science.
David Yen is currently a Professor and Dean of JHJ School of Business at Texas Southern University. Professor Yen is active in research and has published books and articles which have appeared in ACM Transaction of MIS, Decision Support Systems, Information & Management, Decision Sciences, International Journal of Electronic Commerce, ACM SIG Data Base, Information Sciences, Communications of the ACM, Government Information Quarterly, IEEE IT Professionals, Information Society, Omega, International Journal of Organizational Computing and Electronic Commerce, and Communications of AIS among others. Professor Yen's research interests include data communications, electronic/mobile commerce, egovernment and e-health.