Measuring Agreement Among Ranks: Sustainability Application

Is it possible to compare rankings from different sources when the individual rankings of the top x elements differ? To investigate this question, 2015 sustainable rankings from 4 sources that have ranked the top globally most sustainable corporations are considered (Corporate Knights, Fortune's World's Most Admired Companies, Newsweek's Green Rankings, and Harris). These rankings are analyzed using common rank comparison methods (Spearman's ?, Kendall's t). Then, they are analyzed to see if the sources ranking the data are doing so at random or if there is a specific pattern of agreement (Kendall's W and a method by Alvo, Cabilio & Feigin (1982)). The insights from these methods as well as possible limitations are considered. A truly sustainable corporation would transcend all definitions and be good for the environment and the people relying on the company. This paper will attempt to identify data points that tend to cluster close together in one or more groups, thereby justifying the feasibility of identifying sets of companies that are truly the “most” sustainable.