Is the Internet a public good?

As an enabler of development, its primary value is for providing access to other basic goods and services

The growth of the Internet in the 1990s led to the fear of a new kind of social inequity in the form of the digital divide. A contrary view was that the new technology had the potential to overcome previously existing divides, provided government support was available in order to overcome the challenges of ICT (information and communications technology) in under-served areas. A view not examined enough is the relevance of the Internet in the context of the need to balance the allocation of scarce public money among different urgent priorities.

The European Union Universal Service Directive of 2002 suggested that a necessary condition for a service to be included within the ambit of universal service is that, in the light of social, economic and technological developments, the ability to use the service has become critical for social inclusion—that is, it is a consumption norm.

Some governments have tried to operationalize the notion by setting a trigger mechanism in the form of a minimum number of users of a service that would need to be crossed before provision of subsidies for its universalization is considered. While the number of Internet users in India has multiplied, the number of active users is still too low for the Internet to be considered a “consumption norm” that a government is obligated to provide.

Government funds can also be deployed to correct market failures and realize positive externalities. Since the 1980s, sophisticated statistical techniques have been used to establish cause-effect relationships between the adoption of new technologies—the mobile phone, the Internet and broadband—and gross domestic product (GDP) growth. In India, for example, such studies show that a 10% increase in Internet penetration can increase the GDP by 1.6% in the presence of a minimum penetration level of 25%. These studies have become the basis for calls for the use of public money for Internet access and broadband expansion.

However, there are three major conceptual issues with such policy conclusions. First, these studies do not claim that the new technologies will yield positive spillovers prior to the threshold level of penetration being reached. Many under-served areas lie far below the threshold. Second, these studies do not establish that the bang per buck of the new technology is greater than that for other basic inputs such as education, roads or health. Third, these econometric studies are typically carried out with either a country or a province within a country as the unit of analysis. Given the immense size and heterogeneity of the unit, the claims do not automatically extend to the sub-unit level—for example, to rural areas within a province.

Indeed, there are reasons to believe that the externalities may not accrue in many rural areas at their present level of development. Unlike roads, the provision of digital connectivity is not sufficient to ensure empowerment or even equitable inclusion of the target population. The reaping of benefits requires the ability and willingness to use the new technology on the part of the intended beneficiaries, relevant content and applications, and affordability. Mere access without a host of complementary inputs is unlikely to lead to positive spillovers.

As per the “enabler of development” rationale, the Internet has value not in and of itself but rather as a medium that gives access to other basic goods and services. Indeed, ICT for development projects cover many domains including healthcare, education, online government services and the provision of commodity price information to small producers.

Two implications emerge from this. First, the level of provision of the basic goods and services facilitated via ICT should adhere to some consumption norm. In the case of the provision of health services, for example, the government needs to aim for a level that at the minimum achieves the targets of the Millennium Development Goals. Second, since the provision of basic services using ICT is dependent on the availability of other complementary inputs, the decision on the level of a particular ICT service that is to be provided cannot be made without reference to the presence of other complementary inputs. Continuing with the healthcare example, the ICT network should develop in rough alignment with the complementary institutions, processes and skills needed to provide remote medical services.

The provision of connectivity can to some extent substitute for the other inputs. However, the substitutability peters out beyond a point and then the provision of advanced connectivity amounts to wastage of social resources.

There is also the view that the provision of connectivity will trigger the provision of complementary inputs and the development of ability to use (“build it and they will come”). However, the experience of several government schemes in India shows that there are limits to this rationale for advance build-out of connectivity.

The conclusions from this exploration are that “universal access” to the Internet need not be interpreted as “uniform access” and the build-out of networks should be aligned to the absorptive capacity of a region. The deliberations on the national optic fibre network and “free basics” could benefit from such a nuanced approach.