Why Can’t We All Just Get Along? Structural Modelling and Natural Experiments in Merger Analysis

"We start with a discussion of the fundamental principles of science to set the background for the analysis of decision-making. By definition, all scientific statements must be falsifiable. Given the potential for multiple theories to be scientific, policy decisions turn on the evidence underpinning the various scientific theories. Validation (testing) and verification (applicability) are the two key concerns, with theories expected to survive some attempt at falsification prior to application. Section C provides a background for both the structuralist and experimentalist methodologies. In general, structural approaches offer broader insights that may not be applicable to the relevant fact situation, while experimental methods offer clear, but often limited, insights into the relevant question. Testing sits at the heart of experimental analysis, but is often an

afterthought to the structuralist’s focus on modelling and parameterisation. Section D moves on to merger analysis, with analyses of both the Cournot and Bertrand merger models, along with an introduction to the natural experiment approach that serves to underpin the experimentalist analysis."