Targeting, Discretionary Funding, and the Provision of Local Public Goods: Evidence from Kenya
We elicit sitting politicians' preferences over two attributes of local public goods, opportunities for targeting and control of discretionary funding, by conducting an incentive-compatible choice experiment with 179 elected county councilors in rural Kenya. In our experiment, local politicians choose between different public goods packages that vary across two dimensions: whether or not the politician is able to target the good to the location of his choice, and whether he controls the discretionary funding associated with the project. Local officials put a high premium on opportunities for geographic targeting, but not on the ability to control the associated discretionary funding; local officials are particularly uninterested in controlling the funding mechanism (and taking on the associated maintenance responsibilities) when they are able to choose the public good's location. Decisions about where to install the public good suggest a combination of motives: councilors choose locations that generate relatively high social welfare, but favor locations in their home areas. Quantitative estimates suggest that users in one's home area count approximately twice as much as constituents who live further from the councilor.