CEO overconfidence and CSR decoupling

• Author: Steve Sauerwald,Weichieh Su
• Date: 01 July 2019
• DOI: http://doi.org/10.1111/corg.12279
• Published date: 01 July 2019

ORIGINAL ARTICLE

CEO overconfidence and CSR decoupling

Steve Sauerwald

1

|Weichieh Su

2

1

Department of Managerial Studies, University

of Illinois at Chicago, Chicago, Illinois, USA

2

Department of International Business,

National Chengchi University, Taipei, Taiwan

Correspondence

Steve Sauerwald, University of Illinois at

Chicago, Department of Managerial Studies,

601 S Morgan Street 2210 UH, Chicago, IL

60607, USA.

Email: ssauerw@uic.edu

Funding information

UIC College of Business Dean's Summer

Research Grant; Ministry of Science and

Technology, Taiwan, Grant/Award Number:

MOST 105‐2410‐H‐004‐167‐MY4

Abstract

Research question/issue: This study examines whether there is decoupling

between how firms communicate about corporate social responsibility (CSR) and

what firms do in terms of CSR. We argue that this CSR decoupling is driven by the

CEOs' cognitive biases. Specifically, we propose that overconfident CEOs increase

CSR decoupling.

Research findings/insights: We tested our arguments in a sample of S&P 500 firms

for the period of 2006–2014. We find that CEO overconfidence is positively related

to the decoupling between the optimistic tone of CSR reporting and the firm's actual

corporate social performance. However, the board of directors mitigates the effect of

CEO overconfidence on CSR decoupling when outside directors have CSR expertise

and ownership incentives.

Theoretical/academic implications: Previous studies have suggested that CSR

decoupling is a function of opportunistic management that can be constrained by

external monitoring. We examine CSR decoupling as a function of cognitive biases

(such as overconfidence) that can be constrained by internal monitoring.

Practitioner/policy implications: This study provides insights into the conditions

when CSR information released by the firm is symbolic. Practitioners may prevent such

symbolic CSR reporting by imposing effective oversight by the board of directors.

KEYWORDS

Corporate Governance, Overconfident CEOs, CSR Reporting, Symbolic Management, Decoupling

1|INTRODUCTION

Firms increasingly invest in corporate social responsibility (CSR)—

understood as “the responsibility of enterprises for their impacts on

society”(European Commission, 2011, p. 6)—to improve stakeholder

relationships (Porter & Kramer, 2011), insure against idiosyncratic firm

risks (Orlitzky & Benjamin, 2001), and ultimately increase financial per-

formance (Barnett, 2007; Flammer, 2015; McWilliams & Siegel, 2000;

Su & Tsang, 2015; Surroca, Tribó, & Waddock, 2010; Wang, Choi, & Li,

2008). Given the important role of CSR, it is imperative for stake-

holders to understand a firm's CSR actions and performance. Firms

therefore increasingly issue CSR reports to inform stakeholders about

their CSR activities (Beauchamp & O'Connor, 2012; Crilly, Hansen, &

Zollo, 2016; Gibson & O'Donovan, 2007; Tata & Prasad, 2014).

CSR reports should convey CSR information reliably and correctly,

but recent studies found variations in the accuracy of CSR reports that

distort information provided to stakeholders (Hooghiemstra, 2000;

Tata & Prasad, 2014). A prominent reason for the potential misinfor-

mation in CSR reports is the use of symbolic CSR management

(Michelon, Pilonato, & Ricceri, 2015). Symbolic management may lead

to CSR reporting that is decoupled from the reality of corporate social

performance (CSP; Hawn & Ioannou, 2016). Such decoupling may pro-

vide an incomplete and one‐sided CSR image of the firm (Kim & Lyon,

2015). Understanding what accounts for CSR decoupling—defined as

the gap between how firms communicate about CSR and what firms

do in terms of CSR—is therefore an important research issue. Research

so far primarily examined the causes of CSR decoupling emanating

from outside the firm, such as the threat of monitoring by external

Received: 22 January 2018 Revised: 11 March 2019 Accepted: 12 March 2019

DOI: 10.1111/corg.12279

Corp Govern Int Rev. 2019;27:283–300. © 2019 John Wiley & Sons Ltdwileyonlinelibrary.com/journal/corg 283

stakeholders (Lyon & Maxwell, 2011; Marquis & Qian, 2014; Schons &

Steinmeier, 2016).

However, we know little about the internal forces that determine

CSR decoupling. CEOs may have incentives to engage in CSR

decoupling when the actual CSR performance of their companies falls

short of the desired CSR image (Deegan, 2002). We suggest that CEOs'

cognitive biases affect the degree of CSR decoupling. Recent attention

points to overconfidence as one of the most powerful and prevalent

cognitive biases among CEOs (Chen, Crossland, & Luo, 2015; Tang,

Li, & Yang, 2015; Tang, Qian, Chen, & Shen, 2015). We propose that

overconfident CEOs have a biased perception of their abilities to influ-

ence their firm's CSR activities. This cognitive bias may lead CEOs to

communicate a CSR picture that is overly optimistic when compared

with the CSR performance of the firm, resulting in CSR decoupling.

Given the risks of painting an inaccurate picture of the firm's CSR

activities (Bromley & Powell, 2012; Lyon & Maxwell, 2011), firms may

implement internal corporate governance mechanisms to mitigate CSR

decoupling. We propose two corporate governance conditions that

moderate overconfident CEOs' use of CSR decoupling. First, we argue

that outside directors who experience effective CSR practices in other

boardrooms may reduce the effect of CEO overconfidence on CSR

decoupling because such directors can use their CSR expertise to eval-

uate, discuss, and monitor CSR in the focal firm. Second, we argue that

outside directors with high levels of stock ownership also reduce the

effect of CEO overconfidence on CSR decoupling because ownership

motivates outside directors to protect firms from biased CSR

communications.

CSR reporting is an ideal setting to investigate CSR decoupling

because it creates benefits for both firms and managers. CSR reporting

creates benefits for firms by reducing capital costs (Dhaliwal, Li, Tsang,

& Yang, 2011). CEOs may benefit from CSR personally by impressing

members of the corporate elite (Marquis, Davis, & Glynn, 2013),

protecting their jobs after engaging in earnings management (Prior,

Surroca, & Tribó, 2008) and securing perks and admiration from their

peers (Atkinson & Galaskiewicz, 1988). Thus, firms and CEOs are moti-

vated to show good CSR actions and performance regardless of actual

CSR performance, which, if performance is lacking, may promote a

decoupling of CSR reporting and actual CSR performance. On the

basis of a sample of 1,003 CSR reports issued between 2006 and

2014, we find that CEO overconfidence increases CSR decoupling.

We further find that internal governance mechanisms such as boards

of directors can mitigate this tendency when outside directors are

qualified and motivated.

Our study makes two contributions. First, we contribute to the

decoupling literature by exploring the potential for symbolic manage-

ment based on the CEO's cognitive biases (Bromley & Powell, 2012).

Previous studies suggested that decoupling may be a risk for firms

(Kim & Lyon, 2015; Marquis & Qian, 2014), but little is known whether

the CEO's cognitive biases directly account for decoupling. We argue

that overconfident CEOs promote symbolic expressions and thus

increase CSR decoupling. Our study, therefore, resonates with prior

studies finding that decoupling serves the interests of powerful corpo-

rate leaders (Westphal & Zajac, 1998, 2001). Second, we contribute to

the corporate governance literature by examining the board's effec-

tiveness in constraining CSR decoupling. Previous studies found that

external governance mechanisms such as state monitoring effectively

curbs symbolic CSR (Marquis & Qian, 2014). However, we know little

about whether firms can proactively reduce symbolic CSR practices by

imposing internal governance mechanisms, such as effective board

monitoring. Our findings not only confirm the effectiveness of incen-

tive mechanisms such as outside director ownership (Atkinson &

Galaskiewicz, 1988), but also show that board expertise in the CSR

domain can reduce CSR decoupling (Hillman & Dalziel, 2003).

2|THEORY AND HYPOTHESES

DEVELOPMENT

2.1 |CEOs and symbolic management

Research on symbolic management suggests that firms often decouple

their actual practices from espoused policies (Westphal & Zajac, 1998,

2001). In other words, what firms claim to be doing may not accurately

reflect what they are actually doing (Meyer & Rowan, 1977). CEOs

play a crucial role in decoupling decisions. For instance, CEOs may

engage in symbolic management to protect their discretion by

appearing to fulfill stakeholder expectations (Westphal & Zajac,

2013). This research traditionally examines how CEOs decouple their

policies from practice to appear complying with institutional norms

(such as shareholder value norms [i.e., Fiss & Zajac, 2004] and corpo-

rate governance standards [i.e., Westphal & Zajac, 1998]).

Some research has also started to examine how CEOs' disposi-

tional characteristics may lead to the use of symbolic management

to protect their self‐image. For instance, top managers may decouple

policies from practice to protect their ideological beliefs (Tilcsik,

2010). Besides CEOs' dispositional attributes, cognitive biases—that

is, systematic errors in judgment—may also play a key role in

decoupling policy from practice. However, cognitive biases have

received little attention in the study of symbolic management

(Westphal & Zajac, 2013). In this study, we focus on the cognitive bias

overconfidence and its effect on CSR decoupling.

2.2 |CEO overconfidence and CSR decoupling

CEO overconfidence refers to a cognitive bias that leads individuals to

overestimate their abilities to produce a desired outcome (Russo &

Schoemaker, 1992). An emerging stream of research has examined

the effects of CEO overconfidence on strategic decisions, primarily

focusing on improving financial performance. For instance, overconfi-

dence induces CEOs to increase their firm's willingness to take risk (Li

& Tang, 2010; Simon & Houghton, 2003) and invest in innovation

(Hirshleifer, Low, & Teoh, 2012). CEO overconfidence may also lead

to more mergers and acquisitions (Malmendier & Tate, 2008) and

reduce the CEO's ability to learn from past mistakes (Chen et al.,

2015).

284 SAUERWALD AND SU