Three empirical essays on industries with creative content

In the motion picture industry most actors and actresses work for low fees, but a small group can command compensations of millions of dollars per movie. In the first chapter I test the consumer learning hypothesis: consumers choose films using information from other films with the same star actor or actress. Using DVD sales data for a sample of 206 films, I show that the release of a new movie increases sales of old movies with the same star. However, this result is driven by a subset of sequels and Oscar nominated films. I conclude that some effect other than consumer learning must drive the presence of superstars in the motion picture industry. In chapter two I provide a novel explanation for the optimality of the sequential release by showing that sequential release can be the equilibrium of a signaling game. Theater owners must choose which films to screen and studios signal which films are of high quality by delaying the home video release of those films until after the theatrical run. Sequential release is an effective form of signaling because it is costly. The properties of the equilibrium of the DVD release timing game I describe suggest that a theatrical premiere with release in other markets no sooner than three months later will remain the norm for big budget movies. One important economic consequence of the Internet and information technology has been the reduction in the fixed costs of distribution and a consequent increase in product variety. In chapter three I investigate the effect on the welfare of producers of creative content. I identify a set of books sold on Amazon.com for which the author's revenue can be easily computed. I use this data to parameterize a model of author entry and then estimate the welfare gains of authors and consumers from Print-on-Demand (POD) technology.