|Sector:||Competition Economics | Competition, Regulation and Business | Mergers|
|Document type:||LE publication (working papers and economic briefs)|
In the context of merger analysis, statistical inference is often used in the guise of a hypothesis test, where the test is on whether the merger will be harmful to consumers (and/or possibly to other competitors). The aim of this Working Paper is to review the methodological choices regarding the use of statistical inference. We focus on the implications of the choices of two main components of a hypothesis test: the choice of the null and alternative hypotheses; and the choice of the level of significance of the test.