This dissertation is comprised of three essays. The first two study the welfare effects of dynamic ticket pricing, meaning when a seller adjusts the price of a given seat at a given event as that event approaches. The third studies the effectiveness of a particular merger policy. In the first chapter, I analyze two motivations for dynamic ticket pricing: (1) price discrimination across consumer arrival time and (2) re-optimizing due to changes in the perceived product, in this case due to team performance. I estimate a flexible model of Major League Baseball ticket demand that takes both forces into account, then solve for optimal pricing over simulated team performance paths. I use an original data set of daily sales, prices, and product characteristics for over 400 games. I find that, on average, product changes affect price more strongly until the final week before a game, when the shift on consumer types plays a larger role. Prices therefore tend to oscillate and then increase, consistent with observed behavior. Resellers play a key role, dropping prices and dampening the franchise's final-week price increase. I find that dynamic pricing leads to substantial revenue gains compared to a pricing policy which depends on date of purchase but not dynamic product characteristics. In aggregate consumers lose, with those low-willingness-to-pay consumers who happen to face higher prices being hit the hardest. While the first chapter centers on an agent with high market power, the second chapter analyzes the dynamic pricing problem faced by agents with comparatively little market power: ticket resellers. Not only do they face steep demand curves but, unlike franchises, they have extremely limited inventory: most listings offer only two tickets. More than half of two-ticket resellers practice uniform pricing, suggesting a price adjustment cost. I assume that each of these resellers posted her price expecting never to change it, and by estimating demand and assuming optimal uniform pricing I recover each reseller's "scrap value," the value of still having tickets after markets have closed on the day of the game. The scrap value turns out to be negative for 41% of listings, suggesting either risk aversion or bounded rationality. Assuming that these scrap values are actually zero, I then simulate each reseller's expected revenue under a counterfactual where that reseller uses full dynamic pricing. The increase in expected revenue is greatest for those with zero scrap value, followed far behind by those with scrap value above zero and less than face value, followed by those with scrap value above face value. In the third chapter, Keaton Miller and I examine the effectiveness of U.S. pre-merger notification policy by studying the acquisition behaviors of cable telecommunication companies. We construct a novel dataset of acquisitions in the cable industry from 2000-2012, making use of the large sample and the ability to reasonably define the sets of actual potential mergers. We find that the Hart-Scott-Rodino disclosure threshold only affects firm behavior when acquiring firms whose geographic coverage overlaps with their own. In other words, The disclosure threshold appears to be successful in discouraging or preventing anticompetitive acquisitions.