Languages and earnings management

https://doi.org/10.1016/j.jacceco.2017.04.001Get rights and content

Abstract

We predict that managers of firms in countries where languages do not require speakers to grammatically mark future events perceive future consequences of earnings management to be more imminent, and therefore they are less likely to engage in earnings management. Using data from 38 countries, we find that accrual-based earnings management and real earnings management are less prevalent where there is weaker time disassociation in the language. Our study is the first to examine the relation between the grammatical structure of languages and financial reporting characteristics, and it extends the literature on the effect of informal institutions on corporate actions.

Introduction

In this paper, we investigate how languages are associated with earnings management in different countries. Languages differ in the way they encode time.1 Following the linguistics literature, we separate languages into two types based on the way they encode time: strong future-time reference (FTR) languages, such as English, and weak FTR languages, such as German. Strong FTR languages require speakers to mark the timing of events in a distinct way, whereas weak FTR languages do not. In a weak FTR language, future events are talked about in the present tense. This may lead these speakers to perceive future events to be relatively less distant. According to Dahl (2000) and Thieroff (2000), marking future events mandatorily through future tense, for example, by using the verb “will” in English, reduces a person's concern about the future, because it increases the psychological distance from and reduces the psychological importance of the future.

Consistent with this argument, Chen (2013) finds that when the grammatical structure of a language disassociates the future from the present, speakers of the language also disassociate the future from the present in their behavior. Specifically, Chen (2013) shows that people whose language does not require them to grammatically mark future events (weak FTR languages) save more than those whose language requires them to grammatically mark future events (strong FTR languages). He also shows that people speaking weak FTR languages engage in more future-oriented behavior, such as more exercise and less smoking, than those speaking strong FTR languages. Given that weak FTR languages reduce psychological distance from the future (Dahl, 2000, Thieroff, 2000), managers in countries with weak FTR languages are likely to perceive future consequences of earnings management, such as possible restatements, enforcement actions, litigations, and dismissals, to be more imminent. Thus, we argue that firms in countries with weak FTR languages are likely to engage in less earnings management than firms in countries with strong FTR languages.

Using a large sample of firms from 38 countries, we examine whether accrual-based and real earnings management are less prevalent in countries with weak FTR languages than in countries with strong FTR languages. After controlling for various properties of formal institutions (legal origin, economic growth, ownership concentration, and creditor rights) and country-specific cultural characteristics (uncertainty avoidance, masculinity, individualism, and power distance), we show that firms associated with weak FTR languages engage in less accrual-based and real earnings management than firms in countries with strong FTR languages. Through country-level regressions, we also show that weak FTR languages are associated with more extensive external equity markets and less earnings management aggregated at the country level.

Although we control for various country-specific characteristics, our regressions are fundamentally cross-country and may omit important differences between countries not captured by these controls. To further isolate linguistic effects from confounds that vary on the country level, such as taxes, institutions, and capital markets, we also conduct a within-country analysis based on the birthplace information of U.S. firms’ CEOs. We code the languages by CEOs’ countries of origin. Because we focus on U.S. firms only, we are able to control for the effects of formal institutions that differ across countries. Fernández (2011) suggests that when individuals emigrate from their native country to a new country, their cultural beliefs and values travel with them, but their external economic and institutional environments are left behind. We find that CEOs born in countries with weak FTR languages engage in less earnings management than CEOs born in countries with strong FTR languages, confirming our results in cross-country regressions.

Our results are robust to excluding U.S. firms and to excluding firms in Belgium and Singapore, where a significant percentage of population uses different languages. Our results are also robust when we replace a dichotomous classification of FTR in languages with continuous measures based on a word-frequency analysis of online texts. Additional controls for cross-country differences in insider trading restrictions and compliance between financial and tax reporting do not change the tenor of the results. We further find that the negative relation between weak FTR languages and accrual-based earnings management is less pronounced for firms that issue American Depository Receipts (ADRs), suggesting that international exposure attenuates the relation between languages and earnings management.

Linguistically induced bias in time perception and lower precision of beliefs about time can lead weak FTR speakers to apply lower discount rates (Chen, 2013), making future cash flows and earnings relatively more valuable to weak FTR speakers than to strong FTR speakers. Under this scenario, weak FTR speakers are less likely to undertake income-increasing earnings management that shifts earnings from the future to the present, but could engage in more earnings management that shifts earnings from present to future periods. Concerns about the future negative consequences of earnings management, however, encourage weak FTR speakers to avoid both income-creasing and income-decreasing earnings management. Our evidence shows that both income-increasing and income-decreasing earnings management are negatively associated with weak FTR languages, suggesting that concerns about future negative consequences of earnings management dominate the discount rate effect.

Our study makes several important contributions to the literature. First, we contribute to the emerging literature that examines the effects of informal institutions on corporate behavior. Although studies that examine the effects of formal institutions on corporate policies are abundant, evidence on the effects of informal institutions, such as culture, values, and religion, is relatively scarce. Prior studies that investigate the effects of informal institutions focus on religion (e.g., McGuire et al., 2012) and culture (e.g., Han et al., 2010). Our study is the first to show the systematic relation between languages and earnings management.

Second, our study is the first to combine accounting and the grammatical structure of languages, more specifically, how languages mark time, and attempts to build a link between languages and cross-country variances in earnings management. A few recent studies examine the relation between linguistic complexity and disclosure. Lundholm et al. (2014), for example, find that foreign firms listed on U.S. stock exchanges write clearer text in the Management Discussion and Analysis section of their 10-Ks and write more readable text in their earnings press releases than do comparable U.S. firms. Brochet et al. (2016) find a negative relation between linguistic opacity and the investor reaction to conference calls. While these studies focus on how linguistic complexity associates managerial choice of and investors’ reactions to disclosure, we add to the literature by examining how the grammatical structure of languages relates to financial reporting characteristics.

Finally, by identifying an important non-market factor that is significantly associated with both accrual-based and real earnings management, we also contribute to the earnings management literature. We show that linguistically induced bias influences earnings management practices across countries.

The remainder of the paper is organized as follows. Section 2 provides the literature review and hypothesis development. Section 3 discusses the data and research design. We present the results in Section 4. Section 5 concludes the paper.

Section snippets

Literature review and hypothesis development

During their first ten to twelve years, human beings learn how to live, absorbing necessary information from their environment through language (Hofstede et al., 2010). Language is an element that influences human beings earlier than any other societal elements, such as religion, culture, and formal institutions. Given this, language could have a considerable effect on human behavior.

Languages differ as to how they mark future events. English marks the future with either “will” or “be going

Sample and data

We obtain data on FTR of each language from the European Science Foundation's Typology of Languages in Europe (EUROTYP) project (Chen, 2013). Future-time reference is a focal area of the EUROTYP Theme Group on Tense and Aspect, which studies the typological and areal distribution of grammaticalized FTR. Chen (2013) notes that the EUROTYP project is the most extensive typological program to study the cross-linguistic grammaticalization of FTR. Weak-FTR languages are those that do not require the

FTR of languages and earnings management

Table 4 presents the result from estimating Eq. (2), which links accrual-based earnings management and the FTR of languages, as well as the result from estimating Eq. (4), which links real earnings management and the FTR of languages. We cluster standard errors by country in these regressions and other cross-country regressions. In the first column, where the dependent variable is the absolute value of discretionary accruals (ABSDA), Weak FTR is negatively related to ABSDA at the 1%

Conclusion

Prior studies that examine cross-country variations in earnings management focus on legal institutions (Leuz et al., 2003) and informal institutions, such as culture (Han et al., 2010). Although language is an element that influences human beings earlier than other elements, such as culture, religion, and formal institutions, and therefore has considerable effects on human behavior (Hofstede et al., 2010), not much is known as to how language relates to financial reporting characteristics of

References (42)

  • A. Sweeney

    Debt covenant violations and managers' accounting responses

    J. Account. Econ.

    (1994)
  • C.L. Becker et al.

    The Effect of audit quality on earnings management

    Cont. Account. Res.

    (1998)
  • B. Blaylock et al.

    The association between book-tax conformity and earnings management

    Rev. Account. Stud.

    (2015)
  • F. Brochet et al.

    The capital market consequences of language barriers in the conference calls of non-U.S. firms

    Account. Rev.

    (2016)
  • M. Chen

    The effect of language on economic behavior: evidence from savings rates, health behaviors, and retirement assets

    Am. Econ. Rev.

    (2013)
  • S. Chen et al.

    Language and Corporate Decision Making

    (2015)
  • D.A. Cohen et al.

    Real and accrual-based earnings management in the pre and post Sarbanes-Oxley periods

    Account. Rev.

    (2008)
  • B. Comrie

    Tense

    (1985)
  • Ö. Dahl

    The grammar of future time reference in European languages

  • H. Desai et al.

    The reputational penalty for aggressive accounting: earnings restatements and management turnover

    Account. Rev.

    (2006)
  • D.J. Denis et al.

    Insider Trading Restriction and Top Executive Compensation

    (2013)
  • Cited by (0)

    We thank John Core (the editor), Rodrigo Verdi (the referee), Kee-Hong Bae, Shirley Daniel, Roger Debreceny, Koren Jo, Boochun Jung, Jun-Koo Kang, Sang-Hyop Lee, Zhonggao Lin, Dawn Matsumoto, O. Nicholas Ordway, Kunsu Park, Sangsoo Park, Ghon Rhee, Siew Hong Teoh, Chong Wang, David Yang, Donghoon Yang, and Jun Zheng for insightful comments. Yongtae Kim acknowledges financial support from Robert and Barbara McCullough Family Chair Professorship. Jian Zhou gratefully acknowledges the research support from Lloyd Fujie / Deloitte Foundation Professorship and the summer research support from Shidler College of Business at University of Hawaii at Manoa.

    View full text