No abstract

No abstract

We estimate a mixed logit model using data on choices of broadband technologies by 94,388 subscribers to a single broadband operator in a European country on a monthly basis from January to December 2014. We find that valuation of DSL connection speed in the range between 1 and 8 MB/s is very similar. Moreover, in January 2014, the valuation of FttH connection with speed of 100 MB/s is not much different than of DSL connection with speed of 1 or 8 MB/s but it increased over time. The small initial difference in valuation of DSL and FttH connections may be because basic Internet needs of consumers such as emailing, reading news, shopping, browsing and even watching videos online could be satisfied with connection speed below 8 MB/s. We also find that consumers face significant switching costs when changing broadband tariffs, which are substantially higher when switching from DSL to FttH technology. According to counterfactual simulations based on our model estimates, switching costs between technologies are the main factor which slows down transition from DSL to FttH.

No abstract

In this paper, we study the impact of competition on the legacy copper network on the deployment of high-speed broadband. We first develop a theoretical model, which shows that the relation between the number of competitors and investment in a quality-improving technology can be positive if the quality of the new technology is high enough, and is negative otherwise. We test these theoretical predictions using data on broadband deployments in France in more than 36,000 local municipalities. First, using panel data over the period 2011-2014, we estimate a model of entry into local markets by alternative operators using local loop unbundling (LLU). Second, using cross-sectional data for the year 2015, we estimate how the number of LLU entrants impacts the deployment of high-speed broadband, controlling for the endogeneity of LLU entry. We find that a higher number of LLU competitors implies lower incentives to deploy and expand coverage of fast broadband, with speed of 30Mbps or more. By contrast, a higher number of local competitors has a positive effect on the incentive to deploy super-fast broadband, with speed of 100Mbps or more, but has no significant effect on the incentive to expand coverage of this technology.