Freedom of choice adds value to public goods

Significance Public goods, from tangible properties to intangible services, benefit all. They are produced or maintained through widespread participation in public goods provision. Low participation rates are therefore a looming threat that has motivated countless searches for ways to elicit participation. Recent theory suggests that social networks, as woven by personal relationships, are instrumental. We organized a social dilemma game experiment to investigate whether player participation in public goods provision depends on the global characteristics of social networks or the ability to freely choose among local public goods within a player’s network neighborhood. Our results demonstrate the importance of the latter factor, thus favoring bottom-up public goods provision that gives individuals a say in decision-making.
Public goods, ranging from judiciary to sanitation to parkland, permeate daily life. They have been a subject of intense interdisciplinary study, with a traditional focus being on participation levels in isolated public goods games (PGGs) as opposed to a more recent focus on participation in PGGs embedded into complex social networks. We merged the two perspectives by arranging voluntary participants into one of three network configurations, upon which volunteers played a number of iterated PGGs within their network neighborhood. The purpose was to test whether the topology of social networks or a freedom to express preferences for some local public goods over others affect participation. The results show that changes in social networks are of little consequence, yet volunteers significantly increase participation when they freely express preferences. Surprisingly, the increase in participation happens from the very beginning of the game experiment, before any information about how others play can be gathered. Such information does get used later in the game as volunteers seek to correlate contributions with higher returns, thus adding significant value to public goods overall. These results are ascribable to a small number of behavioral phenotypes, and suggest that societies may be better off with bottom-up schemes for public goods provision.