Comments

Alexander Galetovic: Internet diffusion is a hot topic in Latin America. Presidents, policymakers, and politicians are eager to speak about their plans and schemes to put their countries ahead in this strategic race and eliminate the digital divide. Ambitious targets are set and spectacular penetration rates promised. The sobering conclusion of Estache, Manacorda, and Valletti is that a lot of work remains to be done. Despite impressive advances in telecommunications indicators in the 1990s, penetration rates are still far from those in developed countries, and the gap is unlikely to close anytime soon. Not surprisingly, the authors trace the cause of the gap to differences in per capita income and, to a lesser extent, income inequality. But they also argue that part of the blame lies with defective regulation, particularly access rules. Their main conclusion is that substantially improving regulation, in general, and access rules, in particular, is key to cutting the cost of connecting to the Internet. I organize my comments around three topics: the main unresolved regulatory issues in the telecommunications industry, particularly in Latin America; recommendations for improving regulation; and the relevance of affordability and the digital divide.

Most of the issues that the authors discuss are relevant not only for Internet access and diffusion, but also for telecommunications regulation in general. In particular, their arguments also apply to the regulation of local call rates charged by dominant fixed phone companies, an issue that will still be with us for some time.

What Is to Be Regulated?

As the authors show, most people access Internet and other telecommunication services through a fixed phone network, which is thus an essential facility: you cannot reach users connected to a given fixed network unless the owner grants you access to it. ¹ Competition can occur at two different levels. The first level is between service providers such as long distance companies, Internet service providers (ISPs), cell phone companies, or call backs who use the network to reach users. Experience shows that these activities are fairly competitive. Regulators need only to ensure that providers can access the fixed phone network both technically and at reasonable cost; all else will follow. Second, fixed networks themselves can compete for end users—the so-called facility-based competition. In Santiago, Chile, for example, there are zones in which four different companies (one of them a cable company) offer fixed phone services, with each company using its own physical network. Nevertheless, regulation is needed even when fixed networks compete. As the authors discuss, network externalities imply that a dominant company with most of the subscribers can easily derail entry by others simply by denying them access to their larger pool of clients; this concern is very relevant, because in all countries incumbents are former legal monopolies. In addition, each fixed network has monopoly power on access to their subscribers, and companies may easily collude to set high reciprocal access charges, which soften competition. ² On top of that, there is far less agreement whether facilities-based competition is desirable in the first place, because it is inevitably duplicative and thus prevents firms from taking full advantage of scale economies.

The authors are right in pointing out that improving access regulation is central as far as Internet diffusion is concerned. Their examination of regulatory practices and regimes across the continent reveals that with some exceptions, governments essentially privatized first and have yet to design adequate regulatory regimes. Thus, access rules are still not clear in most countries, and governments have not developed the capacity to fix access charges. The consequence is that the cost of providing Internet services (and hence the price of the service) is still too high, while investments in fixed digital lines are probably being delayed because of uncertainty about regulatory rules. This distortion means that access to the Internet is sub-optimally low. While the evidence presented in the paper falls short of proving their claims, well-informed observers would basically agree with their assessment. I am less convinced, however, by their...