Essays on industrial organization, intellectual property, and econometrics

This dissertation contains essays on industrial organization, intellectual property, and econometrics. The first essay discusses the joint operating arrangement (JOA) between two of Colorado's major daily newspapers, the Denver Post and the Rocky Mountain News. It proposes a simple bargaining model showing that the newspapers' assessments of their financial conditions are truthfully revealed in their proposed JOA profit-sharing plan. The analysis casts doubt on the contention that the Rocky Mountain News would have failed in the absence of the JOA. The model provides some guidelines for distinguishing applications for JOAs that are in fact simply attempts to seek collusive profits. In particular, when the parties to a JOA propose to divide their profits evenly, it implies that they think each has similar survival prospects. In such cases the JOA application should be rejected.;The second essay shows that when estimating time-series regressions of unknown functional form, multicollinearity can make regressions very sensitive to small misspecifications of nonlinear relationships. This essay shows that a simple seminonparametric estimator can help to solve this problem. Specifically, adding a flexible time trend to the regression model will give significantly less biased OLS estimators, even if the model remains misspecified. This is because the trend is now fitted directly, so the other explanatory variables need to fit only fluctuations around the trend. A consistency result is proven, and Monte Carlo experiments are used to illustrate the approach.;The third essay discusses open source software and intellectual property protection. It explores how the new institution of open source software facilitates the diffusion of knowledge while at the same time providing incentives for programmers. The essay first summarizes the literature on four major theories about open source software. Case studies of seven open source projects are conducted. The results provide support for the use-value theory, the voluntary contribution mechanism theory, and the spin-off theory, while providing less support for the reputation theory. The empirical results suggest that effort is allocated efficiently in the sense that more resources are devoted to more popular projects. This chapter also brings forward interesting questions for future research.