Catherine BROS

Caste rules prohibit the sharing of water between castes as a contact made by an untouchable with water ritually taints the source. Despite untouchability practices being outlawed by the Constitution of India, they are still vigorous and violently enforced. This is what is investigated in this article. The aim is to evidence a relationship between the number of acts of violence against untouchables and the way water is distributed. Our results show that, the more individual the source, the less scope there is for potential ritual pollution and hence, the lower the number of violent acts against untouchables. This is the first analysis, to the best of our knowledge, that quantifies and evidences the enforcement of untouchability rules with regard to water distribution. This paper underlines that water improvement programs have the nice side effect of alleviating caste based violence on top of improving sanitation.

A vast recent literature has stressed social fragmentation's negative impact on the provision of public goods. It has been established theoretically that social fragmentation engenders discord and thereby undermines public goods provision. Empirical research has produced mixed results about this relationship. On the one hand it rarely holds for all the goods and on another hand it appears attenuated at the micro-level. Three points ought to be considered. First, the negative role attributed to social fragmentation rests upon the actuality of a relationship between social antagonisms and ethnic diversity. Yet, such an actuality is to be proved. Second, should such a relationship exist, polarization indices would be more appropriate than the traditional fractionalization index used so far in the literature. Third, theoretical works have set aside the possibility of ethnic patronage in accessing public goods. Nevertheless, it is a central issue as patronage is common in developing countries. In this event, a positive relationship could be found between social fragmentation and the presence of public goods. This article aims at showing that such a positive relationship does exist, at least in parts of India, as a consequence of caste patronage. It also shown that polarization is irrelevant as social antagonisms do not seem to be an obstacle to the provision of public goods.

A vast recent literature has stressed social fragmentation's negative impact on the provision of public goods. This is a key issue, given that public goods availability has been reckoned as crucial to economic development, while developing countries' societies often exhibit high degrees of fragmentation. Although it has been well established both empirically and theoretically that fragmentation is detrimental to collective action, two caveats ought to be considered. First, a high level of social fragmentation is often associated with greater inequality, which, as Olson pointed out, may be beneficial to collective action. In Olson's argument, should most of the public goods benefits accrue to a small number of group members, they are encouraged to invest in group activities, given that their stakes in the collective action are quite high. Second, should access to publicly provided goods be restricted to the elite, a positive relationship may be found between fragmentation and ethnically based patronage. Given that both patronage and inequality are common in developing countries, it is surprising that fragmentation has never been found to have a positive effect on the provision of public goods. This article aims at showing that not only does this positive relationship exist, but it is linked to the presence of wealthy individuals who are in a position to deny access to public goods to other groups members.

The aim of this paper is to provide a detailed analysis of the process of segregation formation. The claim is that segregation does not originate from prejudice or exogenous psychological factors. Rather it is the product of strategic interactions among social groups in a setting where one group has captured power. While using a model featuring random matching and repeated games, it is shown that whenever one group seizes power, members of other groups will perceive additional value in forging long term relationships with the mighty. They will systematically cooperate with the latter either because it is in their interest to do so or because they do not have other choice. The mighty natural response to this yearning to cooperate is to refuse intergroup relationships. The dominated group will best reply to this new situation by in turn rejecting the relationships and a segregation equilibrium emerges. Segregation stems from the systematic cooperation by one group with another. However, not all societies that have experienced power captures converge towards segregation. It is shown that the proportion of individuals that are actually powerful within the mighty group determines convergence towards segregation.