A 2007 NBER working paper by Banerjee, Iyer and Somanathan presents the following facts: “In Nepal, access to schools is ten times better in the best districts compared to the worst. For Kenyan provinces, this ratio is 8:1; it is more than 2:1 for both Indian states and Russian regions and slightly over 1.5:1 for Chinese provinces. In contrast, regional differences are small in Mexico and Thailand and negligible in Vietnam.” Why is the case? What determines these differences in public good allocations? The Banerjee, Iyer and Somanathan paper reviews some of the root causes of why public good provision can be so variable within and across countries.
- Land Ownership. Foster and Rosenzweig (2000) investigate 245 villages in India between 1971 and 1982 and find that “investments in schooling were greatest in areas with a high fraction of landed relative to landless households.”
- Cultural Norms and Religious Beliefs. The Brahman caste are traditionally considered to be priestly elite in India. A paper by Banerjee and Somanathan (2006) finds that “in the early 1970s, the population share of Brahmans in a constituency is positively correlated with access to primary, middle and secondary schools, to post offices and to piped water.”
- Sex of Politician. The 73rd Amendment to the Indian Constitution requires that one third of all local election seats be reserved for women. A paper by Chattoppadhyay and Duflo (2004) concludes that “political reservation for women in local government results in greater provision of goods which women value, such as drinking water and roads.”
- Ethnic Fragmentation. There has been a lot of work done to see whether or not ethnic diversity will increase or decrease the amount of public goods provided in a given area. Alesina, Baqir and Easterly (1999) find that U.S. cities with more ethnic fragmentation “spend proportionally less on schooling, roads and trash pickups but more on health and police.” Banerjee and Somanathan look at data in India and “construct a fractionalization index of social heterogeneity using population shares of non-Hindu religions (Muslims, Christians, Sikhs, Jains, Parsis), as well as 185 distinct Hindu caste groups…Of the fifteen different public goods considered in 1971, they find that the social heterogeneity measure has a significant negative effect in six cases and positive in two.”
- Monitoring and Communication. A general finding for the experimental literature is that individuals generally give more money to public goods than predicted by the Nash Equilibrium, but if games are repeated, the donations converge to the stingy Nash equilibrium. An experiment by Cason and Khan (1999) allowed the experimental group to converse for a few minutes before they decide on their contributions. Individuals receive information on others donations levels after either every round or every six rounds. The authors find that “in the absence of communication, contributions with both types of monitoring are fairly similar and decline over time. In the presence of communication, overall contributions are much higher (about 80% of the tokens were invested in the group activity as opposed to a high of 40% in the no-communication case) and did not decline in later rounds.” Thus one can conclude that a monitoring mechanism operated within a group where communication is frequent can lead to high levels of public good contribution.
- Autocratic Rule. Many public good expansions have taken place under colonial or autocratic rule. The European powers often helped to provide schooling for the people living in their colonies. The British, French and Dutch invested more in local public goods than did the Belgian or Portuguese. During the 19th century, autocratic princely states in India often invested heavily in public goods despite no political pressure to do so. “The Travancore state in present-day Kerala is particularly well-known for its long tradition of enlightened rulers. In 1817, the Regent Gauri Parvathi Bai declared, ‘The state should defray the entire cost of the education of its people in order that there might be no backwardness in the spread of enlightenment among them, that by diffusion of education they might become better subjects and public servants and that the reputation of the state might be enhanced thereby.’ “
- Political Atmosphere. “After independence, the Kenyan leadership played up tribal loyalties for political reasons and little effort was put into building a Kenyan identity; in contrast, the Tanzanian leadership put a lot of emphasis on creating a single Tanzanian identity. This seems to have implications for public good provision: in the Busia region of Kenya, ethnic heterogeneity at the local level is negatively correlated with the quality of public goods (mainly schools), while in the nearby Meatu region of Tanzania, they are slightly positively correlated.”
Note: The term “public good” is used in a very general sense in the paper. Using a stricter definition, education has some public good elements if working with more educated individuals increases one’s productivity. Medical care generally is not seen to be a public good with the exception of services such immunization, but Public Health efforts—such as maintaining a clean water source—are more likely to fall within the public good arena.
- Banerjee, Iyer, Somanathan (2007) “Public Action for Public Goods” NBER Working Paper, #12911.