The provision of basic social infrastructure, such as schools, clinics, and latrines, is a crucial function of government in developing countries.
According to experts, these politically popular, locally targeted, and highly visible projects are also exactly the type of public goods that politicians seek to deliver in theories of pork barrel politics, tactical redistribution, and credit claiming.
Yet it is anecdotally common for governments in developing countries to start work on such projects but never actually complete them, leaving behind half-finished projects of no value to users and voters.
Though widely remarked upon and substantively important, the extent and causes of this phenomenon have received surprisingly little attention.
Viewed through the lens of distributive politics that has guided many studies on the political economy of public good provision in developing countries, non-completion is also theoretically puzzling: facing strong political incentives to deliver projects, why would governments use scarce resources to build only the bottom half of a school?
In a report titled ‘The Political Economy of Unfinished Development Projects: Corruption, Clientelism, or Collective Choice?’, Martin provides three plausible explanations for the phenomenon of project non-completion. First, non-completion may be due to corruption, either for private gain or to finance political activities. In this view, projects go unfinished because someone stole the money.
Second, non-completion could arise in theories of clientelism in which it may be sometimes rational for politicians to deliberately leave projects unfinished to increase voters’ incentives to re-elect them.
In the clientelism view, projects go unfinished because it was strategically optimal for the incumbent to leave them unfinished.
Third, non-completion could be the result of dynamically inconsistent collective choice processes over project distribution, in which political actors’ inability to sustain intertemporal bargains among each other leads to erratic project implementation.