Comments and Observations Regarding the Relation Between Theory and Empirical Research in Contemporary Accounting Research

We offer some thoughts on the relation between theoretical and empirical accounting research in the context of causal inference, in response to two questions posed by Professor Ivan Marinovic, organizer of the 2014 Stanford University Graduate School of Business Causality conference. The two questions are: should causal inference be the objective of accounting research; and what is, and what should be, the relation between theory and empirical research in accounting? With regard to the latter, we point to two sources of difficulty: (1) confusion and disagreement about interpretation, advantages and disadvantages of various empirical identification strategies; and (2) a lack of progress on the part of empirical researchers in testing the implications of existing accounting theories and thereby providing discipline to those theories. We argue that published empirical accounting research relies too much on insufficiently precise verbal models or generic models that provide few or no new accounting-specific insights and tends to ignore recent advances made by theoretical researchers. As a result analytical models in accounting research are not sufficiently challenged by empirical research and analytical researchers have made slow progress in establishing a meaningful distinction between accounting information and other types of information provided by firms and their managers. Our concern is that accounting research is in danger of losing the healthy disciplining balance between theory and empirical research that is essential to any scientific field. Without this balance, the profession becomes a discipline of beliefs, rather than a discipline of scientific discovery.