The ABCs of empirical corporate (governance) research

• Author: Renée B. Adams
• DOI: http://doi.org/10.1111/corg.12229
• Date: 01 November 2017
• Published date: 01 November 2017

SPECIAL ISSUE ARTICLE

The ABCs of empirical corporate (governance) research

Renée B. Adams

University of New South Wales, Sydney, NSW

2052, Australia

Correspondence

Renée Adams, University of New South Wales,

Sydney, NSW 2052, Australia

Email: renee.adams@unsw.edu.au

Abstract

Manuscript Type: Commentary

Research Question/Issue: I ask how academics can do better empirical research in corpo-

rate governance and other fields.

Research Findings/Insights: I argue that academics can do better research by using old

methods better. I provide a list of recommendations I label as ABCs as a reminder that one can

write excellent papers if one gets the basics right. The ABCs also serve as a reminder that one

must master the basics before one can credibly use new techniques.

Theoretical/Academic Implications: Following the recommendations I provide can

improve research by making the research more transparent and easier to replicate.

Practitioner/Policy Implications: Better research improves policy‐making and practice,

especially if improvements in research practices lead to research that is easier to read.

KEYWORDS

Corporate Governance, Research, Methods, Assumptions, Causality, Writing

As scientists, we are conditioned to constantly look for new methods

to advance our knowledge. But, sometimes we can achieve our goals

more easily by using old methods better. This commentary lists some

ABCs of empirical work that I think are useful to remember. The term

“ABCs”is deliberate. It serves as a reminder that one can write

excellent papers if one gets the basics right. It is also a reminder that

one must master the basics before one can credibly use new

techniques. As Hamermesh (2000, p. 378) says: “Before we resort to

wizardry, we should be certain that we do not add confusion by

making mistakes with simpler techniques.”

These ABCs are based on my experience, my reading of the

literature, and on my extensive experience as a referee and editor.

Until it became untenable, I used to have a policy that I would referee

at least once for any journal that asked me. As a result, I have refereed

a substantial number of papers at more than 40 journals at all levels of

impact across a variety of different disciplines (e.g. finance, economics,

accounting, management, and psychology). I've also been an associate

editor, special issue editor, and now editor. As a result, I've followed

the progress of a substantial number of papers at the journals and

know what referees complain about.

My ABCsare not exhaustive, nor arethey meant to be. Doing empir-

ical workis a craft (e.g. Hamermesh,2000). Real worlddata typicallydoes

not satisfythe assumptions of the models.Often, we do not even know

what the right model is, i.e. there is model uncertainty. As a result, the

sameproblem can be approachedin different ways.Compelling empirical

work combines methodological tools with creativity, critical thinking,

intuition,judgement,taste, and writing skills.As such, it would be difficult,

andeven counterproductive,to try toformalize an exhaustivelist of ABCs

for doing empirical work. The “ABCs”depend, in large part, on the

researchquestion. As our questions evolve,so will the ABCs.

THE ABCS

A. Assumptions: Discuss them!

Referee complaints about the validity of instrumental variables or other

features of the paper's identification str ategy are central to most

rejections. A surprising number of these rejections could have been

avoided if the authors had discussed the assumptions necessary for

identification. As soon as one verbalizes the arguments why instrumental

variables or treatments are exogenous, one often realizes they are not.

This realization can inform the search for a better instrument or “shock”

and ultimately help authors write a more compelling paper.

B. Big data: More data is not always better!

In this era of big data, we are conditioned to think more data is always

better. But, it is not necessarily true. Some phenomena are better

DOI: 10.1111/corg.12229

Corp Govern Int Rev. 2017;25:461–464. © 2017 John Wiley & Sons Ltdwileyonlinelibrary.com/journal/corg 461