Real Effects of a Widespread CSR Reporting Mandate: Evidence from the European Union's CSR Directive

• Published date: 01 September 2022
• Author: PETER FIECHTER,JÖRG‐MARKUS HITZ,NICO LEHMANN
• Date: 01 September 2022
• DOI: http://doi.org/10.1111/1475-679X.12424

DOI: 10.1111/1475-679X.12424

Journal of Accounting Research

Vol. 60 No. 4 September 2022

Printed in U.S.A.

Real Effects of a Widespread CSR

Reporting Mandate: Evidence from

the European Union’s CSR

Directive

PETER FIECHTER,∗JÖRG-MARKUS HITZ,†

AND NICO LEHMANN‡

Received 27 November 2018; accepted 4 January 2022

ABSTRACT

We investigate real effects of a widespread corporate social responsibility

(CSR) reporting mandate. In 2014, the European Union (EU) passed Direc-

tive 2014/95 (hereafter, “CSR Directive”), mandating large listed EU firms

to prepare annual nonfinancial reports beginning from fiscal year 2017 on-

ward. We document that firms within the scope of the directive respond

by increasing their CSR activities and that they start doing so before the

entry-into-force of the directive. These real effects are concentrated in firms

that are plausibly more strongly affected by the directive, that is, those with

previously low levels of both CSR reporting and CSR activities. Using various

alternative outcome variables (e.g., new CSR initiatives, improvements in CSR

∗University of Neuchatel; †University of Göttingen; ‡Erasmus University Rotterdam.

Accepted by Christian Leuz. For helpful comments, we thank two anonymous review-

ers, Ulf Brüggemann, Holger Daske, Michel Dubois, Joachim Gassen, Urska Kosi, Bruno

Lanz, Maximilian Müller, Thorsten Sellhorn, Sönke Sievers, Naomi Soderstrom, Scarlett

Xiaotong Song (discussant), and workshop participants at Humboldt University of Berlin,

Mannheim Business School, Nanyang Business School, University of Paderborn, the Sustain-

able Finance Research Seminar at University of Zurich, the 2017 AAC Convention in Cluj,

the VHB/IAAER Accounting Conference in Berlin, the 2018 EAA annual meeting in Mi-

lan, and the 2018 AAA annual meeting in Washington, DC. An online appendix to this

paper can be downloaded at http://research.chicagobooth.edu/arc/journal-of-accounting-

research/online-supplements.

1499

© 2022 The Chookaszian Accounting Research Center at the University of Chicago Booth School of

Business.

1500 p. fiechter, j.-m. hitz, and n. lehmann

infrastructure, or firm performance), we show that these real effects reflect

meaningful increases in CSR beyond firms’ potential attempts to “greenwash”

CSR performance. Finally, we conduct tests that increase our confidence that

the documented real effects are attributable to the CSR Directive and not

general EU trends in CSR.

JEL codes: G18, G38, K22, K32, L21, M14, M41, M48

Keywords: corporate social responsibility (CSR); disclosure regulation; real

effects; Directive 2014/95; European Union (EU); non-financial reporting

directive (NFRD)

1. Introduction

We investigate real effects of a widespread corporate social responsibility

(CSR) reporting mandate. In 2014, the European Union (EU) passed Di-

rective 2014/95 (hereafter, “CSR Directive”), which mandates that large

listed EU firms must prepare annual nonfinancial (CSR) reports. These re-

ports must include comprehensive information on policies, main risks, and

outcomes related to environmental matters, social and employee factors,

respect for human rights, anti-corruption issues, and diversity of the board

of directors. The EU regulator’s stated objective for this mandate was to

increase firms’ CSR transparency (Directive 2014/95, recital 1). Moreover,

consistent with a real effects objective, it can be inferred from the CSR Di-

rective’s basis for conclusions that the EU regulator regarded the mandate

as a policy tool to nudge firms’ toward pursuing more CSR activities, as “dis-

closure of non-financial information is vital for managing change toward a

sustainable global economy” (Directive 2014/95, recital 3) and “disclosure

of non-financial information helps the measuring, monitoring and manag-

ing of undertakings’ performance and their impact on society” (Directive

2014/95, recital 3).

Previous literature provides evidence of real effects for various CSR dis-

closure mandates: for single country settings such as China (Chen, Hung,

and Wang [2018]), for industries such as the U.S. mining sector (Chris-

tensen et al. [2017]), and for specific reporting outcomes such as green-

house gas emissions (Jouvenot and Krueger, [2020], Downar et al. [2021],

Tomar [2021]) or extraction payments (Rauter [2020]). Yet it is unclear

whether and how these findings translate to other countries and industries.

This is a particular concern for our EU setting, given that the CSR Direc-

tive represents an unprecedented act of supra-national disclosure regula-

tion, affecting many firms across different industries and countries.1Firms

within the scope of the directive are diverse in terms of industry, business

model, and location of operations, resulting in very different CSR reporting

1The CSR Directive was expected to affect approximately 6,000 firms across 28 EU member

states (European Commission [2014]).

real effects of a widespread csr reporting mandate 1501

issues.2This diversity in turn impedes stakeholders’ monitoring and bench-

marking of firms’ CSR. On top of this, firms may attempt to meet the CSR

reporting requirements by using boilerplate or “greenwashing” disclosures,

in particular, because enforcement of the directive is, if anything, in its in-

fancy. In this paper, we investigate the open empirical question whether the

CSR Directive resulted in real effects for EU firms.

We begin our analyses by examining whether firms within the scope of

the CSR Directive actually increased CSR transparency. The directive was

passed in April 2014 and applied for fiscal years 2017 onward, which means

that the reporting mandate came into effect in 2018. Therefore, we use

a difference-in-differences design and estimate yearly treatment effects for

all years from 2011 (the starting year of our sample) until 2018 (the year in

which the CSR reporting mandate came into effect), using 2013 as base year

(the year prior to the passage of the directive). As the CSR Directive applies

to all large, listed firms in the EU, we use a sample of propensity-score-

matched U.S. firms as benchmark group.3Our findings document that EU

firms within the scope of the regulation on average increased transparency

of CSR and that they began to do so before the reporting mandate came

into force in 2018.

In our main analyses, we investigate the effect of the CSR Directive on

firms’ CSR activities. Given the heterogeneity of firms and their CSR issues

within the scope of the directive, we select the CSR score provided by Thom-

son Reuters ASSET4 as our outcome variable. This score is a comprehensive

measure of CSR activities, combining publicly available information with a

proprietary analytical technology to capture an array of social and envi-

ronmental activities (e.g., Servaes and Tamayo [2013], Lys, Naughton, and

Wang [2015], Dai, Liang, and Ng [2021]).

Yearly difference-in-differences analyses with CSR activities as the out-

come variable yield three main insights. First, we find a positive treatment

effect for 2018 (relative to the base year 2013). This finding is consistent

with prior empirical evidence, which attributes observed real effects to such

mechanisms as stakeholder pressure following mandatory CSR disclosures

(see, e.g., Christensen, Hail, and Leuz [2021] for a more detailed discus-

sion) or to firms’ benchmarking of CSR performance against industry peers

(Tomar [2021]). Our second finding is that significant real effects mate-

rialized before CSR reporting became mandatory, that is, as early as 2016

2For example, manufacturers have production sites in emerging and development coun-

tries and therefore need to report, for example, on wastewater emissions and carbon emissions

along their supply chain. Meal delivery firms, in contrast, typically have a business focus on ur-

ban EU areas and are therefore required to provide disclosures, for example, on labor safety

and working conditions of their cyclist delivery staff.

3We select U.S. firms as control group because the United States did not adopt any market-

wide CSR-related disclosure mandates during our sample period (Ioannou and Serafeim

[2017], Christensen, Hail, and Leuz [2021]). See subsection 2.2 for a detailed discussion of

our control group choice and section 5 for tests using alternative control groups.