Estimating and predicting merger effects and pass-through rates

My dissertation is centered around issues in antitrust, specifically the estimation and prediction of merger effects and pass-through rates in various industries. Library concern with publisher mergers has been long-standing, but the increased focus on mergers as a primary explanation for journal price inflation appears to be due to work by Mark McCabe. McCabe argues that mergers among journals that are seemingly independent in demand can be expected to result in price increases because libraries face portfolio-specific "budget constraints".;The first chapter of my dissertation replicates and extends McCabe's analysis by conducting a post-merger review of several mergers, including the Reed Elsevier's 2001 acquisition of Harcourt General, which was not examined by McCabe. I find that using the same difference-in-difference specification as McCabe, the Reed Elsevier/Harcourt General merger resulted in price decreases rather than the price increases predicted by McCabe and the Association of Research Libraries. Furthermore, using alternative specifications, I find both price increases and price decreases, even among prior mergers examined by McCabe. Thus, while mergers do affect prices significantly, the explanation is neither market power nor "budget" effects.;The second chapter of my dissertation analyzes fuel costs for passenger airlines. Over the last few years, higher fuel costs have wreaked havoc on the airline industry. Several smaller airlines ceased operating, Delta and Northwest merged, and most airlines implemented fuel surcharges to help recoup some of the increasing costs. This raises an interesting question: why can't the airlines pass through the increases in fuel costs to consumers? I argue that, as a nonsunk fixed cost, an increase in fuel costs can be passed through to consumers in a competitive industry only through reductions in capacity, a process which effected rapidly in a competitive industry is likely to result in major financial pain or bankruptcy.;This paper is one of the first to model the mechanism by which fuel cost increases are passed through to consumers and test that model by estimating the industry-wide pass-through rate. Consistent with that model's predictions for the conditions under which increases in nonsunk fixed costs affect prices, I find an industry-wide pass-through rate that is not statistically different from zero during periods when higher fuel costs do not trigger capacity changes, and a pass-through rate significantly above zero and almost unity during periods when higher fuel costs do trigger capacity changes.;The final chapter of my dissertation returns to predicting and estimating the effects of mergers. While the regulatory agencies, consultants and academics have all conducted ex-post merger reviews in a number of industries, these analyses have typically been conducted on mergers that were allowed to proceed (i.e., not deemed anticompetitive when they were reviewed by the Agencies) or modified. The question of how well the ex ante analyses used in the merger review process would have been able to predict the effects of a merger have not been addressed in the literature. This paper looks at mergers in the asphalt industry in Michigan that did not meet the Hart-Scott-Rodino thresholds and were not reviewed by the Agencies but one was later deemed anticompetitive. Three different ex ante methodologies are used to predict the price effect from the merger, and these predictions are then compared to the ex-post estimates of the actual effects to determine which model most accurately predicts merger effects in this industry. I find that while no single ex ant model predicts very well, combining the three methodologies into a screen may be the best predictor of competitive effects. This paper not only contributes to the literature by being one of the few to compare actual ex-post estimates to ex ante predictions, but it also contributes to the literature by being one of the only papers to calculate ex-post effects under an auction context. (Abstract shortened by UMI.)...