Mapping the corporate governance scholarship: Current state and future directions

• Published date: 01 January 2023
• Author: Nitesh Pandey,Christian Andres,Satish Kumar
• Date: 01 January 2023
• DOI: http://doi.org/10.1111/corg.12444

REVIEW ARTICLE

Mapping the corporate governance scholarship: Current state

and future directions

Nitesh Pandey

1

| Christian Andres

2

| Satish Kumar

1

1

Department of Management Studies,

Malaviya National Institute of Technology,

Jaipur, India

2

WHU–Otto Beisheim School of Management,

Vallendar, Germany

Correspondence

Christian Andres, WHU–Otto Beisheim School

of Management, Burgplatz 2, 56179 Vallendar,

Germany.

Email: christian.andres@whu.edu

Abstract

Research Question/Issues: We conduct a comprehensive analysis of all scholarly

publications in the field of corporate governance (CG) since the Enron scandal of

2001. Using bibliometric methods, we identify prominent themes that have served as

the foundation of CG research, prominent topics in the field along with their tempo-

ral developments, and recent trends in CG research. In addition, we identify the

authors and journals that have had the biggest impact in the field.

Research Findings/Insights: We document that the number of annual publications in

CG has increased by a factor of eight since the early 2000s, with research articles

being published in a wide variety of general interest journals, as well as in outlets spe-

cialized in the field of CG. We identify six research themes as the foundation of CG

research: the theoretical foundations of CG, ownership, CG mechanisms and firm

outcomes, disclosures, the board of directors, and family firms. We further find

19 major bibliographic keyword clusters that have been explored by researchers in

the past 20 years. Research on corporate social responsibility (CSR) and sustainabil-

ity, governance mechanisms, control mechanisms and disclosures, board diversity,

CG in family firms, and CG in the Chinese context are the most dynamic areas of

research in recent years.

Theoretical/Academic Implications: We systematically analyze the literature on CG

and outline theoretical foundations, structuring academic contributions by keyword

clusters. Links between keyword clusters as well as the most dynamic research areas

are identified. The analysis provides guidance for researchers regarding suitable out-

lets for the different thematic clusters and helps as a basis to identify research

opportunities.

Practitioner/Policy Implications: We document growing research interest in the field

of CSR and sustainability. The growing body of literature in these areas can inform

CG and CSR policies in the future. Our comprehensive bibliographic analysis provides

an overview of CG research themes across research disciplines and streams of the

literature.

Received: 2 February 2021 Revised: 17 March 2022 Accepted: 23 March 2022

DOI: 10.1111/corg.12444

This is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits use and distribution in any

medium, provided the original work is properly cited, the use is non-commercial and no modifications or adaptations are made.

© 2022 The Authors. Corporate Governance: An International Review published by John Wiley & Sons Ltd.

Corp Govern Int Rev. 2023;31:127–160. wileyonlinelibrary.com/journal/corg 127

KEYWORDS

corporate governance, bibliometric analysis, citations analysis, systematic review,

bibliographic clustering

1|INTRODUCTION

Corporate governance (CG) as a distinct and systematic field of study

has emerged relatively recently in the academic literature (Kumar &

Zattoni, 2019). The field stands at the intersection of many research

areas, including microeconomics, organizational economics, organiza-

tional theory, information theory, law, accounting, finance, manage-

ment, psychology, sociology, and politics (Turnbull, 1997). In addition

to a public debate about the governance requirements of corporations

and the resulting regulatory changes (e.g., the Cadbury Report of

1992, the Sarbanes–Oxley Act of 2002, and the Higgs Report of

2003), a number of corporate scandals—starting with high-profile

events such as Enron, Tyco, and WorldCom—have not only raised the

awareness of investors but also stirred the interest of academics in

this field. With empirical evidence establishing a causal link between

CG and shareholder wealth (Bebchuk et al., 2009), researchers have

explored a broad range of CG mechanisms and their implications for

both firm “input”decisions (e.g., investments and research and devel-

opment) and firm outcomes (notably measures of accounting and mar-

ket performance).

Using bibliometric techniques, this paper analyzes the intellectual

structure of scholarly research in the field of CG. Specifically, we map

and connect the development of individual keywords and thematic

clusters in all academic journal publications covered by the Web of

Science platform since the Enron scandal of 2001. We thereby pro-

vide a systematic overview of the literature as a whole and are able to

identify temporal developments in the relative interest and impor-

tance of individual research themes. To fulfill this aim, we pursue

(at least) three objectives: First, we identify the key thematic clusters

that serve as the foundation of CG research since 2001 and provide a

brief overview of the literature of each cluster. Second, we analyze

keyword occurrences and their interrelations, allowing for a deeper

look into subtopics, time trends in their importance, and connections

between thematic keywords. Third, we identify the topics that have

been at the forefront of recent developments in the academic litera-

ture. To fulfill these objectives, we use methods of bibliometric analy-

sis that enable us to handle large amounts of publication data (Donthu

et al., 2021; Ramos-Rodríguez & Ruíz-Navarro, 2004). Specifically, we

use co-citation to find themes that have served as building blocks for

academic CG research since 2001 and keyword co-occurrence to

identify major topics in the field. To identify thematic clusters and

topics currently being pursued in research, we follow Andersen (2019)

and apply bibliographic coupling to articles published within the last

3 years (2018–2020).

Our paper is related to a number of recent papers in the area of

CG that apply bibliometric research methods to conduct a structured

literature review. Tunger and Eulerich (2018) review the field of CG

with a focus on German articles. They find six major thematic clusters

dominating research in German CG. Zheng and Kouwenberg (2019)

present a bibliometric review of the CG literature on the board of

directors. Using publication data from 1996 to 2018, they present the

state of research with a focus on individual board attributes, pointing

out the multidisciplinary nature of board-related research. Other

methodologies, such as systematic reviews and meta-analysis, have

been used, with a focus on individual subtopics within the field

(e.g., Carcello et al., 2011; Chalmers et al., 2019; García-Meca &

Sánchez-Ballesta, 2009; Hoobler et al., 2018; Post & Byron, 2015;

Terjesen et al., 2009; Uhde et al., 2017). The present paper is the first

comprehensive analysis of the body of CG research and is thereby

able to map the breadth of the current literature and to identify inter-

relations between individual aspects of CG research. In contrast to lit-

erature reviews—which typically provide very detailed overviews but

are necessarily focused on specific topics or subfields of the

literature—we are able to identify themes that bind together the dif-

ferent substreams of the literature and act as the core of the

discipline.

We begin, in Section 2, with a discussion of our methodology and

a structured overview of the sources of our literature analysis.

Section 3contains a source analysis and identifies the most impactful

journals and researchers in the field. Section 4uses co-citation analy-

sis to identify six thematic clusters and briefly summarizes the most

impactful publications of each cluster. Section 5explores research

topics in more detail on the basis of keyword co-occurrence and doc-

uments the temporal development in the number of publications for

each of the 19 keyword clusters identified in the analysis. Section 6

uses bibliographic coupling, analyzing contributions published over

the past 3 years to identify new developments in the field. We con-

clude with a brief overview of our findings.

2|METHODOLOGY

This study uses bibliometric methodologies to analyze the field of CG

research. The bibliometric method is the application of quantitative

tools to bibliographic data (Broadus, 1987). Bibliometric analysis has

been considered a legitimate method of scientific review in many

fields of study (Ellegaard & Wallin, 2015; Kumar, Pandey, et al., 2021;

Kumar, Sureka, et al., 2021), including management (Donthu

et al., 2021). Due to its quantitative nature, it facilitates the analysis of

large quantities of bibliographic data while minimizing potential biases

(Burton et al., 2020). The most prominent bias in this context is inter-

pretation bias, which implies that authors from different academic

backgrounds could look at the literature differently, thereby adding a

subjective component. This bias is likely if the results of a literature

128 PANDEY ET AL.

review are qualitative (MacCoun, 1998). With the use of bibliometric

methods, authors can summarize the literature using quantitative

tools and minimize interpretation bias.

We use co-citation analysis, keyword co-occurrence, and biblio-

graphic coupling to fulfill our research objectives. The co-citation

technique is based on the idea that papers cited together are similar

in content (Donthu et al., 2021). This analysis is useful in finding major

themes in a body of work (Liu, Yin et al., 2015) and thereby

identifying the intellectual structure of a field (Rossetto et al., 2018).

Our analysis uses co-citation to find themes that have served as

building blocks for academic CG research in the 20 years since the

Enron scandal.

Keyword co-occurrence analysis (Callon et al., 1983) is based on

the assumption that the appearance of certain words together across

different documents indicates their relatedness on a conceptual level.

Author-chosen keywords in any publication are the set of words that

are used to express its central themes (Zou et al., 2018). These words

are considered important by the authors (Pesta et al., 2018) and thus

represent their intent (Comerio & Strozzi, 2019). The analysis of key-

words and their co-occurrence can be instrumental in understanding a

field of study (Castriotta et al., 2019). We use keyword co-occurrence

to identify more specific research topics. The approach is similar to

that of Hutton et al. (2021) who analyze research trends in corporate

finance by reviewing articles published in the Journal of Corporate

Finance based on keyword analysis. Doing so their study presents the

development of research trends over the journal's 25-year history

while outlining fruitful directions for future research in the area of

corporate finance.

Bibliographic coupling, or co-referencing, analysis works under

the assumption that the similarity between two documents will

depend upon their shared literature references (Kessler, 1963; Kumar

et al., 2020; Mukherjee et al., 2021; Weinberg, 1974). The develop-

ment of any scientific field depends on the knowledge that precedes

it (Samiee et al., 2015), and the contributions of any study are based

on the literature accessed to conduct it (Hoffman & Holbrook, 1993).

The prior knowledge generated in the field is often acknowledged in

the form of literature references. Therefore, two documents accessing

the same sources of knowledge, that is, that share literature refer-

ences, must have similarities in themes and topics. We use biblio-

graphic coupling for the analysis of more recent research articles

(2018–2020) to identify topics that are currently of interest to

researchers. The reason for this methodological choice is that the

number of articles is much smaller than the overall corpus, facilitating

the creation of article clusters. In addition, the focus of the third part

of our analysis is on articles that have been published fairly recently.

These articles might thus not yet have appeared in reference lists,

which is a necessary requirement for co-citation analysis. In the case

of keyword co-occurrence, some keywords are very general

(e.g., performance) and are used in multiple contexts, requiring an

examination of the articles they appear in to derive any meaning

(Chang et al., 2015). Bibliographic coupling focuses on the articles

themselves and is therefore preferable if a relatively smaller number

of articles are to be linked and summarized.

To identify the relative importance of research themes, this study

uses measures from network analysis (Andersen, 2019). Specifically,

the degree of centrality, which is the number of nodes a given node is

connected to, and eigenvector centrality, which is a measure of the

relative importance of a node in the network, are employed. Eigenvec-

tor centrality is based on the assumption that a node connected to

other highly connected nodes will carry a great deal of information

about the network. In addition, co-citations, keyword network, and

bibliographic coupling networks are divided into clusters based on

their similarities (number of co-citations, number of keyword co-

occurrences, and number of shared literature references, respectively),

using Newman and Girvan's (2004) algorithm.

The articles considered in this analysis are obtained using the key-

word corporate governance in the Web of Science database in May

2021.

1

This choice is motivated by the coverage of high-quality

sources in Web of Science and prior work (e.g., Baker et al., 2020;

Kurzhals et al., 2020; Linnenluecke, 2017; Lu et al., 2012; Mas-Tur

et al., 2020; Poje & Groff, 2021; Soto-Simeone et al., 2020). The sea-

rch is restricted to articles published between 2001 and 2020, with

the language restricted to English. This search results in 16,996 docu-

ments (called articles hereafter). We further apply subject area filters

with results restricted to the Web of Science categories of business

finance, business, management, economics, ethics, political science,

and interdisciplinary social sciences, which results in 13,663 articles.

We then apply a quality filter and consider only articles published in

journals listed in the Academic Journal Guide (2018, hereafter AJG),

2

published by the Chartered Association of Business Schools. This step

leads to a final set of 12,498 articles.

3

In the co-citation analysis (Section 4), which is used to identify

major research themes, we consider only articles with at least 100 cita-

tions. There are no methodological guides for choosing a specific cita-

tion threshold, the network visualization being the sole concern

behind the threshold choice (Eom, 2009; Hota et al., 2020). Previous

studies have used the stress value to determine the goodness of fit

for their network (e.g., Hota et al., 2020), but, as noted by Chabowski

et al. (2013), stress values can be influenced by the removal and/or

addition of studies, which may make the configurations less meaning-

ful. In the resulting network, after merging duplicate entries, we

obtain a network containing 661 articles that represent the most

impactful publications in the field. These articles are then divided into

clusters, using the clustering algorithm of Newman and Girvan (2004),

which results in the formation of six thematic clusters. Each thematic

cluster is interpreted using its 20–30 most cited articles (depending

upon the size of the cluster).

We use the default threshold of five occurrences in VOSviewer

(van Eck & Waltman, 2010) to generate the keyword co-occurrence

network (Section 5), and we exclude keywords that are the plural form

or abbreviations of others (e.g., top management team,top manage-

ment teams, and TMT are merged). Network matrices are calculated

for the resulting network of 1196 keywords. For the analysis of recent

research fronts (Section 6), we focus on articles published between

2018 and 2020 and cluster them using bibliographic coupling.

Figure 1presents the research design for this study.

PANDEY ET AL.129