Public Policy and Market Competition: How the Master Settlement Agreement Changed the Cigarette Industry

Abstract

This paper investigates the large and unexpected increase in cigarette prices that followed the 1997 Master Settlement Agreement (MSA). We integrate key features of rational addiction theory into a discrete-choice model of the demand for a differentiated product. We find that following the MSA firms set prices on a more elastic region of their demand curves. Using these estimates, we predict prices that would be charged under a variety of industry structures and pricing rules. Under the assumptions of firms' perfect foresight and constant marginal costs, we fail to reject the hypothesis that firms collude on a dynamic pricing strategy.

How Do Smokers Respond to Cigarette Taxes? Evidence from China's Cigarette Industry

“…Our measure of smoking intensity for an average smoker masks considerable heterogeneity in smoking addiction. However, Ciliberto and Kuminoff () also assume smokers to be equally addicted and use a similar addiction measure in their study on the US cigarette industry.…”

“…However, very few studies in the literature on cigarette demand have treated cigarettes as differentiated products. Three notable exceptions are Tan (), Cilliberto and Kuminoff (), and Qi (), which apply differentiated product models to study firms’ behaviors in the US cigarette market. In contrast, we focus on the demand side.…”

A Random Coefficients Logit Analysis of the Counterfactual: A Merger and Divestiture in the Australian Cigarette Industry

“…The combination of tax changes and the high level of regulation prevented us from being able to predict exact price changes. In addition, we also provide new information on the extent of market power in the cigarette industry in addition the sole other brand-level study by Ciliberto and Kumino (2010). We would recommend further research into the applicability of the random coecients model to assessing price eects of a merger in industries which are unencumbered by the intense regulation of the cigarette industry.…”

“…The random coecient logit model is used as it is the main dierentiated products demand model used for estimating demand (Ackerberg et al, 2007), including for cigarettes (Ciliberto and Kumino, 2010). There are now several assessments of its performance in merger simulation, such as Peters (2006) and Weinberg and Hosken (forthcoming) but despite divestitures being a common tool of competition of competition authorities, Friberg and Romahn (2012) is the only other study assessing its performance in simulating divestitures.…”

“…indirect costs are included, which would be excluded in our estimates. A second reason could be that the Bertrand-Nash assumption is not appropriate for all brands and pack sizes Ciliberto and Kumino (2010). suggest allowing for addiction results in rms charging lower prices.…”