| Mobile telephone service has been one of the most successful mass market technologies over the last several decades. Its broad acceptance and ever-increasing penetration raise a number of interesting questions regarding the sorts of factors that motivate people to subscribe to the service. Moreover, it raises a number of questions for regulators of telephone services, including what mobile service implies for universal service requirements, the extent to which mobile wireless service serves as a competitive check on the prices of regulated incumbent fixed line carriers, and the potential for market power in the mobile telephone sector, itself. Thus, there is great importance for analysis of the demand for mobile telephone service. There is a substantial literature studying demand for fixed line telephone service; however, the literature on mobile telephone demand, although becoming more prevalent, is still relatively limited. In this dissertation, I develop a structural model of demand for wireless telephone service, which I derive from a quadratic utility function. I derive a likelihood function which permits simultaneous estimation of parameters affecting both the access choice and minutes-of-use choices of a consumer. In addition, the model allows for uncertainty in ex ante usage, which could lead to ex post optimization error. I estimate this model using simulated maximum likelihood from a sample of actual billing data for households which were part of a large survey. I find that demand for access is inelastic to changes in the monthly access fee, and to changes in the per-minute use charge. I also find that fixed line service is a substitute for wireless. Wireless demand is also correlated with youth and is associated with higher incomes, all else equal. Lastly, I show that a monthly access fee subsidy applied to wireless subscription, akin to the Federal Lifeline Assistance program, would help to achieve universal service objectives. In general, my results are consistent with those of prior studies. |
