Price Developments in Post-Cartel Periods

Jacob Berget

Student thesis: Master thesis

Abstract

This paper examines price developments in post-cartel period, a topicthat hitherto has not been subject to much academic research. For thatpurpose, the research question is approached from both a theoreticaland an empirical angle, in an attempt to uncover the dynamics ofpost-collusive pricing behavior. Following the introduction, the secondchapter is dedicated to a theoretical analysis that combines classicand behavioral economics, industrial organization, and sociologyand extends existing models in order to establish five hypotheses onpost-cartel pricing. All hypotheses point towards the fact that postcartelprices in general are above the competitive level, that couldotherwise have been expected but for the existence of the cartel. Thefive hypotheses are as follows:1. Given non-efficient markets, the immediate post-cartel pricesare expected to be above the competitive level2. Post-cartel prices are expected to be higher in markets wherethe post-cartel period is characterized by a litigation phase3. Residual collusion can lead to abnormally high post-cartel prices4. As price variance is generally larger during competitive periodsthan under collusive ones, a non-competitive post-cartel periodis expected to exhibit below-average price variation5. Due to reciprocity, post-cartel periods triggered by whistle blowingor deviation from an agreement, are expected to be characterizedby above-average competitionIn chapter three, part of the theoretical findings are tested empiricallyusing publicly available data on the case of the German cement cartelfrom 1991 to 2002. Pricing behavior is analyzed using conduct parameters- an estimate of market power - following the method laid outby Bresnahan (1982) and Lau (1982). Alas, the empirical analysis doesnot support nor reject the theoretical predictions. Finally, recommendationsare given as to how the remaining hypotheses can be tested empirically.

EducationsMSc in Applied Economics and Finance, (Graduate Programme) Final Thesis
LanguageEnglish
Publication date2014
Number of pages78
SupervisorsMarcus Asplund