Does religiosity enhance the quality of management earnings forecasts?

• DOI: http://doi.org/10.1111/jbfa.12446
• Author: Luo He,Lamia Chourou,Ligang Zhong
• Date: 01 July 2020
• Published date: 01 July 2020

DOI: 10.1111/jbfa.12446

Does religiosity enhance the quality

of management earnings forecasts?

Lamia Chourou1Luo He2Ligang Zhong3

1TelferSchool of Management, University of

Ottawa, 55 Laurier AvenueEast, Ottawa, Ontario

K1N 6N5, Canada

2John Molson School of Business, Concordia

University, 1455 de MaisonneuveBlvd. West,

Montreal, Quebec H2G 1M8, Canada

3Odette School of Business, University of

Windsor, 401 Sunset Avenue,Windsor, Ontario

N9B 3P4, Canada

Correspondence

LigangZhong, Odette School of Business, Uni-

versityof Windsor, 401 Sunset Avenue,Windsor,

OntarioN9B 3P4, Canada.

Email:ligang.zhong@uwindsor.ca

FundingInformation

LamiaChourou acknowledges financial sup-

portfrom the Social Sciences and Humanities

ResearchCouncil of Canada (430-2017-01148).

LuoHe acknowledges financial support from

FondsQuébécois de la Recherche sur la Société et

laCulture (2013-NP-168704) and the Facilities

OptimizationProgram at Concordia University.

LigangZhong acknowledges financial support

fromthe Research & Teaching Innovation Fundat

theUniversity of Windsor.

Abstract

This study investigates whether firms located in areas with higher

levels of religiosity disclose higher-quality management earnings

forecasts than do other firms. Using a US sample of 4,655 firm-year

observations over the period 2001 to 2014, we find that firms

headquartered in counties with higher proportions of religious

adherents issue earnings forecasts that are less optimistically biased

and that the effect of religiosity is concentrated in firms with weak

monitoring mechanisms. We also find that religiosity mitigates pes-

simistic bias in management earnings forecasts, but only for those

issued by firms operating in low litigation industries. This result

suggests that when the litigation risk is high, both ethicality and

risk aversion are at work and their competing effects likely offset

each other.Additionally, we document that forecasts issued by firms

in more religious areas trigger stronger stock price reactions than

those issued by other firms and that the effect is limited to forecasts

containing optimistic bias. Overall, our results show that religiosity

enhances the quality of management earnings forecasts, but the

effect varies based on different conditions.

KEYWORDS

Ethicality,management earnings forecasts, market reaction, religios-

ity,risk aversion, voluntary disclosure

JEL CLASSIFICATION

G30 (corporate finance and governance), M14 (corporate culture),

M41 (accounting), Z12 (religion)

1INTRODUCTION

The role of religion in economics has long intrigued researchers. Overa century of research has generated an extensive

body of literature at both the macro-level, regarding aggregate economic development (e.g., Anderson, 1988; Barro

& McCleary,2003; La Porta, Lopez-de-Silanes, Shleifer, & Vishny,1999; Weber, 1905), and the micro-level, concerning

individuals (e.g., Chiswick, 1983; Gruber,2005; Halek & Eisenhauer, 2001; Miller & Hoffmann, 1995). The abundance

of research echoes Iannaccone’s (1998, p. 1465) view that ‘studies of religion promise to enhance economics’. The

910 c

2020 John Wiley & Sons Ltd wileyonlinelibrary.com/journal/jbfa JBus Fin Acc. 2020;47:910–948.

CHOUROU ETAL.911

effect of religion at the firm level, however,has only emerged recently as a nascent avenue of research (e.g., Callen &

Fang, 2015; Dyreng, Mayew,& Williams, 2012; Hilary & Hui, 2009; McGuire, Omer, & Sharp, 2012).

Our study investigates the impact of religiosity on firms’ voluntary disclosure of financial information. In particular,

we examine whether US firms headquartered in counties with higher levels of religiosity issue less biased earnings

forecasts and whether stock markets react more strongly to earnings news contained in the forecasts issued by these

firms. Additionally,we examine whether the strength of firm monitoring affects the impact of religiosity on the forecast

qualityofafirm.

Severalrecent studies have examined the link between religiosity and the quality of information contained in finan-

cial statements (Dyreng et al., 2012; Grullon, Kanatas, & Weston,2010; Kanagaretnam, Lobo, & Wang, 2015a; McGuire

et al., 2012). Yetthe topic of voluntary disclosure has been barely investigated. The only exception to this that we are

aware of is Dyreng et al. (2012). Although their primary focus is the quality of financial statements, they find that firms

located in more religious areas report a higher proportion of bad news in press releases. Our study differs from Dyreng

et al. (2012) in that we concentrate on management earnings forecasts (MEFs), the most prominent performance pre-

diction that managers voluntarily provide. MEFs have been widely used as a proxyfor corporate voluntary disclosure

in the prior literature (e.g.,Baginski, Hassell, & Kimbrough, 2002; Healy & Palepu, 2001; Li & Yang, 2016; Nagar, Nanda,

& Wysocki, 2003).

Whether the positive impact of religiosity on the quality of financial statements as documented in prior litera-

ture can simply transfer to MEFs is unclear for three reasons. First, financial statement reporting and MEF disclosure

exhibit different reporting patterns that convey different implications. In MEF disclosure, both optimistic and pes-

simistic biases are common. Managers are generally inclined to appear optimistic about the firm’s future performance

to attractinvestors (Bergman & Roychowdhury, 2008; Rogers & Stocken, 2005), but managers are also motivated to be

pessimistic in their forecast disclosure, so they can shape the market’s expectationsinto an attainable earnings target

that they are able to meet or beat (Matsumoto, 2002; Richardson, Teoh,& Wysocki, 2004). In contrast, in the context

of financial statement reporting, managers’ general inclination to appear positive and their common desire to meet or

beat the market’s expectations both translateinto the same tendency: to manage earnings upward. 1Therefore, what

we have learned from prior studies about financial statement quality does not provide a complete picture. Our finding

that religiosity has differential effects on optimistic and pessimistic biases in MEFs furnishes additional insight into the

full disclosure practices of a firm.

Second, the quality of financial statements is jointly determined by managers and auditors, whereas MEFs are gen-

erally based on unaudited internal reports (Feng, Li, & McVay, 2009). In the absence of auditors, it is questionable

whether managers would behave in the same manner. Forexample, McGuire et al. (2012) document that religiosity

has a negative association with abnormal accruals but a positive association with real earnings management, suggest-

ing that managers in more religious areas conduct less accruals manipulation but more real activities earnings man-

agement. A potential reason for managers’ preference of the latter is that auditors cannot second-guess real earnings

management (Graham, Harvey,& Rajgopal, 2005). More recently, Brockman, Campbell, Lee, and Salas (2019) find that

firms with internally-promoted CEOs provide higher quality MEF disclosure than do other firms, presumably because

such CEOs have a deep understanding of their firm. In contrast, they find no evidence that higher quality in volun-

tary disclosure of MEFs extends to the quality of mandatory reporting such as accruals quality and restatements. The

authors attribute the difference to heavy scrutiny byauditors and other monitors in mandatory reporting.

1Priorstudies also have found evidence consistent with downward earnings management, which, however, occurs only in certain circumstances and/oris con-

ductedby particular types of firms. For example, earnings are managed downward when firms of some industries undergo import relief investigations (Jones,

1991), when very large firms want to reduce political costs (Watts & Zimmerman, 1990), or when managers desire to decrease stock prices before insider

stockpurchases (Badertscher, Phillips, Pincus, & Rego, 2009). In contrast, upward earnings management is much more prevalent, likely because in most firms,

managers are motivated to, for example,seek job security and promotions, maximize compensations, andmaintain good relations with various stakeholders

(Graham,Harvey, & Rajgopal, 2005; Kothari, Shu, & Wysocki, 2009). All these widespread motivations could propel upwardearnings management. Consistent

withthe view that upward earnings management is far more common than downward management, both Srinivasan (2005) and Dyreng, Mayew, and Williams

(2012)report a much larger fraction of restatements that reveal overstated rather than understated earnings.

912 CHOUROU ETAL.

Third, little is known so far about the relationship between voluntary disclosure and mandatory reporting of a firm

(Beyer,Cohen, Lys, & Walther, 2010). It is unclear whether managers tend to use voluntary disclosure as a substitute

for or a complement to mandatory financial reporting.2Hence, we do not know whether the same effect of religiosity

on financial statements simply transfers to MEFs. Examining voluntary disclosure in addition to mandatory reporting

enables us to understand the corporate information environment in its totality (Beyeret al., 2010).

We expect that religiosity in the area where a firm’s headquarters is located affects the quality of the firm’s earn-

ings forecasts, because religion is an important dimension of social norms, and social norms strongly influence human

behavior (Kohlberg, 1984; Sunstein, 1996). A firm’s earnings forecasts are mainly prepared and issued byaccountants

and managers working at the headquarters. Because managers and accountants often live near the headquarters, they

are likely to be connected to the area’s culturalvalues and social norms (Grullon et al., 2010; Hilary & Hui, 2009). Even

if the accountants and managers are not religious themselves, they are still exposed tothe influence of the area’s reli-

gious groups across various social activities and interactions.

Prior literature (e.g., Hilary & Hui, 2009; Longenecker,McKinney, & Moore, 2004; Miller & Hoffmann, 1995) identi-

fies two traitsof religious people, ethicality and risk aversion, that could potentially affect managers’ discretion in their

earnings forecasts. These two traits likely produce competing forces on the quality of MEFs. We expect ethicality to

improve MEF quality,because honesty is an essential component of ethicality. Choi and Pae (2011) find that ethicality

is tied to honest, high-quality financial reporting and disclosure. In the context of MEFs, honesty would contribute to

unbiased forecasts that convey a manager’s true understanding of the firm’s underlying economic performance. The

effect of risk aversion on MEF quality,however, is not straightforward. On the one hand, risk aversion is expected to

curb managers’ optimistic bias in depicting an overly rosy picture of the firm to outsiders, which would improve fore-

cast quality.On the other hand, risk aversion likely results in pessimistic bias because of motivations such as minimizing

litigation risk. Pessimistic bias decreases forecast quality by making MEFs deviatedownward from the truthful predic-

tion. Given the competing effects of ethicality and risk aversion, the overall impact of religiosity on MEF quality is an

empirical question.

Using a US sample of 4,655 firm-year observations overthe period 2001 to 2014, we find that firms headquartered

in counties with a higher proportion of religious adherents issue earnings forecasts that are less optimistically biased.

We also find that religiosity curbs pessimistic bias in MEFs issued by firms operating in low litigation industries. How-

ever, we find no evidence that religiosity has a discernible impact on pessimistic bias in forecasts released by firms

operating in high litigation industries. In addition, we document that the effect of religiosity in reducing bias is concen-

trated in firms with weak internal and external monitoring mechanisms. Furthermore,we find that earnings forecasts

issuedby firms in more religious areas trigger stronger stock price reactions than do forecasts issued by other firms and

thatthe effectis limited to forecasts containing optimistic bias. Overall, our results show that religion matters to corpo-

rate voluntarydisclosure: higher levels of religiosity contribute to higher-quality management forecasts that investors

value.3Our findings are robust to a battery of sensitivity tests, including controlling for a series of macro-level vari-

ables that correlate with religiosity,employing the two-stage Heckman regression models to address the self-selection

bias, distinguishing between two religious groups (Catholics vs. Protestants), controlling for past forecasting accuracy

in market reaction tests, using alternativemeasures of the key variables, and dealing with a database coverage issue.

Our study makes several contributions. First, to the best of our knowledge, our study is the first to examine the

role of religion in shaping the quality of MEFs. While ample evidence indicates that religiosity contributes to higher-

quality mandatory financial reporting (Dyreng et al., 2012; Grullon et al., 2010; Kanagaretnam et al., 2015a; McGuire

et al., 2012), little is known about the role that religiosity plays in voluntary disclosure. Moreover,our study provides

2Wethank an anonymous referee for suggesting this point.

3We also considered MEF quantity.While religiosity may affect MEF quantity, the direction is unclear because ethicality and risk aversion work in opposite

directions. Weexpect that ethicality encourages more frequent MEFs because managers in highly religious areas attempt to be more transparent. However,

we expectrisk aversion to result in less frequent MEFs because more frequent disclosure can encourage litigation if management cannot meet expectations.

Wetest the effect of religiosity on MEF frequency and find no statistically significant result in the full sample nor when we divide the sample on litigation risk

orfirm monitoring strength. We leave further exploration of religiosity and MEF quantity to future research.