The Creation and Performance of Classification Schemes: Rating Systems in United States Broker-Dealers 1993-2000.

Abstract

In the past few years, organizational researchers have found a renewed interest in categorization and its consequences. Most research focuses on the consequences of categorization for the categorized objects, arguing that a clear identity is important for users to understand (and thus value) an object. In my dissertation, I shift the perspective from the users and the categorized objects to focus on the creator of the classification scheme. I question the commonly held assumption that categories are created to reduce ambiguity and structure the world for users. In the first part of my dissertation, I suggest conditions under which a categorizer might create an ambiguous classification scheme rather than an unambiguous one. I argue that the ambiguity of the classification scheme depends on the categorizer’s relationship with the objects it rates, its relationship with many different types of users, and its status relative to other categorizers within the industry. By doing so, I move the focus away from the results of categorization to the antecedent of it. Users frequently question whether an object has been correctly placed in a classification scheme, but they must also question whether the scheme itself has been created strategically by the categorizer. In the second part of my dissertation, I examine the performance of the categorizer as a result of its classification scheme. While strategic behavior may help an organization achieve its objectives, in the long run, there may be negative consequences for a categorizer when the meaning of a classification scheme cannot be understood. I argue that overly ambiguous classification schemes lead to decreases in customer accounts. In addition, I measure the linguistic network position of individual schemes to illuminate how the usage of different schemes within an organizational field affects all firms in the industry. I use equity rating systems (for example, “buy, sell, hold”) in the United States during the years 1993-1999 as a setting in which to test my hypotheses.