Authors
Conway, Elaine
Affiliation
University of DerbyIssue Date
2019-05-16
Metadata
Show full item recordAbstract
With the increasing use of CSR ratings firms' data to guide ethical investing and to derive findings in academic studies, there has been a growth in the number of ratings firms. These firms use differing methodologies and data to derive their ratings. Therefore, it is important to understand whether ratings are commensurable, as decisions made on the basis of the ratings data used may differ. This paper assesses the level of agreement between two ratings firms, Bloomberg and CSR Hub across three main CSR subcategories and an overall score. It uses Lin’s concordance correlation coefficient and intraclass correlation coefficient on continuous ratings, and cross-tabulation and Cohen’s kappa on ranked ratings within a sample of 720 US and EU companies. For both continuous and ranked data, there is most agreement on Employees/Social, Community/Social and Overall categories and weaker agreement on Environmental and Governance categories. Firms in the German DAX are most consistently rated, as are large and medium-sized firms. These findings propose a degree of caution for investors and academics using only one rater as the basis for their decisions/inferences. Accounting practitioners should be aware their CSR disclosures result in differing ratings and should consider which raters their key investors use. This paper is original in the comprehensive range of methods used to analyse two ratings firms across all CSR sub-categories, in samples from both the US and EU.Citation
Conway, E. (2019) 'To agree or disagree? An analysis of CSR ratings firms'. Social and Environmental Accountability Journal, pp. 1-26. DOI: 10.1080/0969160X.2019.1613248.Publisher
Taylor and FrancisJournal
Social and Environmental Accountability JournalDOI
http://dx.doi.org/10.1080/0969160X.2019.1613248Additional Links
https://www.tandfonline.com/doi/full/10.1080/0969160X.2019.1613248Type
ArticleLanguage
enISSN
0969160XEISSN
21562245ae974a485f413a2113503eed53cd6c53
http://dx.doi.org/10.1080/0969160X.2019.1613248