The aim of the paper is to investigate the nature of public goods (PGs) from a policy-oriented perspective, based on the capability approach to well-being. The reason for such an inquiry is that most policy discussions are centered on the relative efficiency of the actors who are supposed to provide the good - public or private – whereas the paper contends that the main issue is how to define PGs.

Samuelson’s definition of PG is based on non-rivalry and non-excludability and on a restrictive notion of good, of its enjoyment and of the government. This restrictiveness – which depends on a market-centered notion of good – also underlies much of the Coasian discussion of PGs. Subsequent contributions led to an extensive taxonomy, which is at odds both with the two defining features of a conventional PG and with the efficiency criterion that ought to underlie its provision. As a result, it is important to redefine PGs and to investigate how this affects decisions concerning their provision.

The paper focuses on the following issues. First, a proper definition of PGs requires a focus on what goals are pursued. Thus, if we switch from utility maximization to capabilities – as defined by Sen – we get a broader definition. For instance, the first approach would conceive of “health care” in terms of pharmaceuticals, medical services and medical infrastructures to cure diseases whereas the second might also include norms - concerning the organization of society – that prevent the very insurgence of diseases.

The shift from the provision of goods to the establishment of appropriate institutional arrangements raises a second issue, relative to the demand for PGs. As long as the latter meet basic needs, they are generally acknowledged to be useful and a government can effectively account for their provision. In countries with higher standards of living and more diversified needs, however, the benefits of PGs may appear to be less extensive and apparently less important: consider vaccinations in some rich countries.

A third issue arises from the above discussion. There is a potential inconsistency between the requirements for a capabilities-centered inclusive society and dispersed individual and social interests. This may lead to the undersupply of PGs. What appears to be an inevitable process raises the question about how policy-makers should relate to the dual nature of the citizenship: beneficiaries and voters.