Voluntary Assurance of Voluntary CSR Disclosure

• Date: 01 February 2017
• Author: Mark Bagnoli,Susan G. Watts
• DOI: http://doi.org/10.1111/jems.12171
• Published date: 01 February 2017

Voluntary Assurance of Voluntary CSR Disclosure

MARK BAGNOLI AND SUSAN G. WATTS

Krannert Graduate School of Management

Purdue University

West Lafayette IN 47907

mbagnoli@purdue.edu,swatts@purdue.edu

We study a firm’s decisions to engage in socially responsible activities, voluntarily report on

them, and purchase external assurance of the report. In our signaling model, neither firm type

nor the level of activity is observed. We show that if voluntary assurance is not too expensive,

the firm that engages in more socially responsible activities purchases external assurance and

thus “selects” a separating equilibrium. As a result, CSR reports can be used to infer the level

of activity and this causes all firms to engage in more socially responsible activity. Further,

when voluntary assurance is required to support a separating equilibrium, greater monitoring

by social activists increases the chosen quality of voluntary assurance—voluntary assurance and

monitoring by social activists are complements, not substitutes.

1. Introduction

Many firms engage in socially responsible activities (e.g., reducing environmental dam-

age, curbing water usage, ensuring safe working conditions, and supporting charitable

or community activities). Disclosure of a firm’s socially responsible activities is not (gen-

erally) mandatory in the United States.1However, some firms voluntarily provide these

disclosures, and some also voluntarily hire third parties to assure some or all of the

content of the disclosures. In their bi-annual survey, KPMG (2013) reports that 86% of

their sample of U.S. firms report on their socially responsible activities, up from 83% in

2011, and that 23% received some type of independent assurance, up from 13% in 2011.2

These results are similar to those reported by Casey and Grenier (2015) who analyze

2,649 corporate social responsibility (CSR) reports published between 1993 and 2010.

They find that only 230 are independently assured.

Of those CSR reports that are voluntarily assured, the quality of assurance seems

to vary considerably in practice. Firms purchase assurance services from a wide variety

of providers, and assurers offer various levels of assurance quality depending on the

Wethank the Editor, Co-Editor, and two anonymous referees for very helpful comments and suggestions. We

also thank Rashad Abdel-Khalik, Anil Arya, Ramji Balakrishnan, Quintao Fan, PJ Hoffman, Tom Lyon,Bob

Magee, Brian Mittendorf, Lin Nan, KoEun Park, Rick Young, and participants at the 2014 Midwest Research

Conference for helpful discussions. Wegratefully acknowledge the financial support provided by the Krannert

Graduate School of Management and Purdue University.

1. The Security and Exchange Commission (SEC) has just implemented a requirement that firms

disclose their use of conflict minerals as mandated by the Dodd–Frank Act (http://www.sec.gov/

News/PressRelease/Detail/PressRelease/1365171484002#.U4oJjiimU40) and has offeredinterpretative guid-

ance regarding disclosure of climate change risks and other environmental risks (http://www.sec.gov/

news/press/2010/2010-15.htm). Despite this, the vast majority of the information in a typical CSR report

is not required to be disclosed in the United States.

2. The Global Reporting Initiative (GRI) finds that only 10% of GRI-based sustainability reports received

independent assurance in 2011 (GRI, 2013).

C2016 Wiley Periodicals, Inc.

Journal of Economics & Management Strategy, Volume26, Number 1, Spring 2017, 205–230

206 Journal of Economics & Management Strategy

extent of their assurance (depth and breadth) and the quality (reputation) of the assurer

itself.3

The objective of this paper is to develop a model of a firm’s decisions to engage in

socially responsible activities, provide a report on these activities, and have the report

professionally assured if it is provided. We develop a role for these voluntarily provided

CSR reports and the voluntary assurance of their contents by relying on a natural

information asymmetry: the firm knows its chosen level of socially responsible activities

but the audience for the report (investors, creditors, consumers, employees, and/or

supply chain partners) does not observe this choice.4We show that this asymmetry

provides the firm with incentives to voluntarily disclose its socially responsible activities

but does not create sufficient incentives to report truthfully.

Despite this, the firm’s CSR report will have value if (at least) some of the audience

can use it to infer the actual level of the firm’s socially responsible activities. We show

that such a separating equilibrium is possible only if the probability of identifying any

mischaracterization in the firm’s CSR report is sufficiently high. This can occur if there

is sufficient monitoring by outsiders such as social activists, but when it does not, we

show that there is an endogenous demand for voluntary third-party assurance of the

voluntary disclosure.

Specifically, firm types with greater incentives to engage in socially responsible

activities have incentives to purchase professional assurance of their CSR reports. Doing

so raises the probability that any mischaracterizations of its activities are identified. Be-

cause assurance is costly, the firm benefits only if its payoff in a separating equilibrium

exceeds its payoff in a pooling equilibrium. When assurance is not too expensive, we

show that this type of firm does, in fact, acquire third-party assurance and “selects” the

separating equilibrium. Further, it does so by choosing the quality of assurance just suf-

ficient to support a separating equilibrium. In other words, the demand for voluntarily

obtained professional assurance of a voluntary disclosure arises because it allows the

purchasing firm to use its CSR report to distinguish itself from less socially responsible

firms. These same incentives also motivate the firm to choose a sufficiently high quality

of assurance to support its efforts to distinguish its level of socially responsible activi-

ties. Thus, if firms face different levels of monitoring by outsiders, our model suggests

that some firms voluntarily assure their voluntary disclosures, others simply provide

voluntary disclosures, and the rest choose not to offer any voluntary disclosure of their

socially responsible activities.

Our analysis has a number of implications. First, we show that firms can and do

distinguish themselves with their CSR reports. In the separating equilibrium, the CSR

report’s audience appropriately infers that firms engage in more socially responsible

activities when they claim to do so. Interestingly, the conditions that allow the audience to

infer the firm’s activities from its CSR reportalso ensure that the most socially responsible

firms exaggerate their activities in their report, consistent with the empirical findings in

Marquis and Toffel (2014).

3. Companies offering assurance include international accounting firms (Deloitte, EY,KPMG, and PwC),

boutique or niche consulting firms and engineering services firms (e.g., Bureau Veritas, Environmental Re-

sources Management Certification and VerificationServices [ERM CVS], TruCost and TwoTomorrows). Some

client firms choose a review rather than an examination rather than a verification, and some choose to purchase

varying levels of assurance for different parts of their CSR reports.

4. In this paper we use the terminology “CSR report” even though in practice, firms use a variety of names

for these reports: social responsibility reports, sustainability reports, ESG reports, and citizenship reports,

among others.