Skip to main navigation Skip to search Skip to main content

Accounting for Corporate Sustainability: Conceptualization, measurement, and disclosure

Research output: Thesis › Ph.D. thesis

79 Downloads (Pure)

Abstract

Sustainability has been widely discussed during the past decade at the political level and amongorganizations worldwide. This has led to a continuing increase in research in the field (Grewal &Serafeim, 2020), but knowledge on disclosing and measuring corporate sustainabilityperformance (CSP) remains unstructured. The concept of sustainability emerged in the 1970s,focusing on resource consumption, population growth, and environmental losses (Poveda, 2011).Subsequently, the idea has evolved and is now referred to as the 21st-century business paradigm(Garcia, Cintra, Cássia, Torres, & Lima, 2016). Today, corporate sustainability (CS) is oftenviewed based on three elements; social, environmental, and economic performance, also calledthe Triple Bottom Line (Elkington, 1998). External and internal stakeholders increasingly demandtransparency and communication of activities related to CS (O'Dwyer, Unerman, & Hession,2005). Once the CSP has been concretized through measurement, this constitutes valuableinformation that many organizations wish to communicate to their stakeholders. This could be togain competitive advantage (Greening & Turban, 2000), legitimacy (Michelon & Parbonetti,2012), increase their brand value (Dhaliwal, Li, Tsang, & Yang, 2011), or gain extra profit (Gao& Bansal, 2013; Revell, Stokes, & Chen, 2010; Schaltegger & Hörisch, 2017).

Therefore, this dissertation investigates measurements and disclosures of CSP, with the following overarching research question:

What affects the measurement and communication of corporate sustainability?

The research question is answered through three independent papers based on archival data fromAmerican Fortune 500 and S&P 500 organizations. The overall research question is concretizedinto three sub-questions examined in the three papers. The sub-questions are:
1) What are the trade-off effects of different corporate sustainability performancemeasurements on precision and accuracy?
2) How can computer-aided text analysis and natural language processing measure corporatesustainability as a construct?
3) How does organizational size or media visibility, jointly and individually, influence thelevel of sustainability disclosure? 

Paper 1 develops a conceptual framework that categorizes, explains, and evaluates available measurement tools according to their accuracy and precision, and provides a guideline for their context-specific application. Paper 1 is a systematic literature review based on an initial sample of 1,415 papers. The final sample consists of 74 papers, suggesting four categories of measurements: 1) isolated indicators, 2) indicator frameworks, 3) sustainability balanced scorecards (SBSC); 4) sustainable performance measurement systems (SPMS). The analysis reveals that isolated indicators are inaccurate and imprecise, limiting the application to organizations with delimited, specific measurements of parts of CSP due to the risk of a GIGOeffect (i.e., measurements based on low-quality data will always end up producing low-quality output). CSP indicator frameworks are imprecise but accurate, making them applicable for organizations handling a more significant amount of CSP data. However, Greensplashing is a risk and occurs when an organization presents many (primarily simple) CSP indicators that are not connected to the organization and do not fit well with the industry, organization, strategy, and culture. In contrast, SBSCs is precise but inaccurate and valuable for organizations desiring a comprehensive strategic management tool with limited capacity to handle sustainability issues. However, the streetlight effect composes a risk if researchers study what is easy to study and managers measure what is easy to measure. The ideal CSP measurement is identified as SPMS, as these are both precise and accurate. SPMS is helpful for organizations with complex, comprehensive, connected, and tailored indicators but are methodologically challenging.

Paper 2 applies features from Computer-Aided Text Analysis (CATA) and Natural Language Processing (NLP) for developing and validating a measurement of CS. Today, CS as a construct is often used in research. Still, proxies for CS are criticized for being non-transparent in terms of their methodology, for being convergent, and for having significant data gaps (Kotsantonis & Serafeim, 2019). Paper 2 utilizes the capabilities of CATA and NLP to develop an alternative measurement of CS based on the development of data dictionaries and a word-counting algorithm. This provides an opportunity to determine what events or concepts are discussed in a document. Paper 2 provides important methodological contributions as it illustrates how to develop a measurement based on CATA and NLP. Additionally, it applies the measurement to an empirical sample and demonstrates the validity and reliability.

Paper 3 investigates the distinct and joint effects of organizational size and media visibility on sustainability disclosures (SD). Previous research tends to use these interchangeably and restricts analysis of the impact of legitimacy pressure per se (Schreck & Raithel, 2018). Paper 3 examines a sample of S&P 500 organizations from 2011-2020 by collecting ESG scores, visibility scores, and financial data. The results indicate a significant positive impact of media visibility on all levels (environmental (ENV), social (SOC), and governance (GOV)) of SDs and a positive and significant effect of organizational size on all disclosure levels but SOC. The effect size for media visibility is largest for ENV disclosure and weakest for GOV disclosure. The picture differs for organizational size as the most substantial effect is on GOV-related disclosures. This highlights the different and isolated effects of media visibility and organizational size on SD and the importance of handling these as separate concepts. In addition, paper 3 finds a significant mediating impact of media visibility and a moderating effect of polluting industries. 

Overall, the dissertation provides systematic archival-based research that contributes to the understanding and application of measurement methods of CSP and the communication thereof. It provides an in-depth and nuanced picture of the field. First, the dissertation contributes by structuring the existing literature and its descriptive characteristics. This provides an overview of the current research situation and highlights missing elements that can be researched in the future. Second, the dissertation contributes by providing a nuanced analysis of the trade-off between accuracy and precision that often arises when measuring CSP. Third, the dissertation offers methodological contributions by developing a new method for measuring or quantifying CSP or CS transparency. The technique can be applied to all written text and constitutes a contribution to research on sustainability and other topics. The method can also be used in practice. Fourth, the dissertation contributes empirically by providing new insights to the relationships between organizational size, media visibility, and disclosure of sustainable activities.
Original languageEnglish
Awarding Institution
  • University of Southern Denmark
Supervisors/Advisors
  • Lueg, Rainer, Supervisor
  • Liempd, Dennis van, Supervisor
Date of defence28. Oct 2022
Publisher
DOIs
Publication statusPublished - 28. Oct 2022

Fingerprint

Dive into the research topics of 'Accounting for Corporate Sustainability: Conceptualization, measurement, and disclosure'. Together they form a unique fingerprint.

Cite this