| In this dissertation, I explore the implications of indirect network effects. Network effects exist when one good (“hardware”) becomes more valuable to each individual user as the total number of users increases. For indirect network effects, the value increases because the greater number of users supports a wider variety of a complementary good (“software”). Much of the literature has discussed such indirect effects while modeling them as direct effects, wherein the utility a user obtains from a good increases directly with the total number of users of the good.; In Chapter 1, I construct a model with indirect network effects which I compare to an analogous model with direct network effects. I examine the propensity of software firms to standardize on the same hardware technology. The ways in which direct and indirect effects influence standardization are quite different: some parameter changes have opposite effects in the two models, and some factors which are irrelevant under direct effects are central under indirect effects. I illustrate the differences in predictions using a variety of examples, including computer hardware and software, VCRs and videotapes, and automobiles and service.; In Chapter 2, I focus on the strategic decisions of hardware firms when firms have market power. I analyze hardware firms' strategies for location in a product space and for compatibility with the competing hardware. I also consider the effects of a first-mover advantage in the hardware market. Again, there are significant differences in the implications of the direct and indirect models.; In Chapter 3, I use a model of indirect network effects to examine a current policy issue. I investigate the competition between suppliers of components of a system for which there are network effects among users. Bundling one of these components with an outside good reduces the cost to consumers of using the system. While this cost reduction is likely to be welfare-enhancing, bundling can also reduce welfare by decreasing innovation incentives. I use the model to evaluate Microsoft's bundling of Windows with Internet Explorer and its effect on competition with Netscape. |
