Editing Ethnic conflict (section)

== Public goods provision ==
A major source of ethnic conflict in multi-ethnic democracies is over the access to state patronage. Conflicts over state resources between ethnic groups can increase the likelihood of ethnic violence. In ethnically divided societies, demand for public goods decreases as each ethnic group derives more utility from benefits targeted at their ethnic group in particular.<ref>{{Cite journal|last=Kolev, Wang|date=2010|title=Ethnic Group Divisions and Clientelism|ssrn=1644406|journal=APSA Annual Meeting Paper}}</ref> These benefits would be less valued if all other ethnic groups had access to them. Targeted benefits are more appealing because ethnic groups can solidify or heighten their social and economic status relative to other ethnic groups whereas broad programmatic policies will not improve their relative worth. Politicians and political parties in turn, have an incentive to favor co-ethnics in their distribution of material benefits. Over the long run, ethnic conflict over access to state benefits is likely to lead to the ethnification of political parties and the party system as a whole where the political salience of ethnic identity increase leading to a self-fulfilling equilibrium: If politicians only distribute benefits on an ethnic basis, voters will see themselves primarily belonging to an ethnic group and view politicians the same way. They will only vote for the politician belonging to the same ethnic group. In turn, politicians will refrain from providing public goods because it will not serve them well electorally to provide services to people not belonging to their ethnic group. In democratizing societies, this could lead to ethnic outbidding and lead to extreme politicians pushing out moderate co-ethnics.<ref name=":19" /> Patronage politics and ethnic politics eventually reinforce each other, leading to what Chandra terms a "patronage democracy".<ref>{{Cite book|url=http://politics.as.nyu.edu/docs/IO/2587/Chandra2007-CountingHeads.pdf|title=Patrons, clients, and policies: Patterns of democratic accountability and political competition|last=Kitschelt|first=Herbert|publisher=Cambridge University Press|year=2007}}</ref>

The existence of patronage networks between local politicians and ethnic groups make it easier for politicians to mobilize ethnic groups and instigate ethnic violence for electoral gain since the neighborhood or city is already polarized along ethnic lines. The dependence of ethnic groups on their co-ethnic local politician for access to state resources is likely to make them more responsive to calls of violence against other ethnic groups.<ref name="Berenschot2010">{{Cite journal|last=Berenschot|first=Ward|date=2010|title=The Spatial Distribution of Riots: Patronage and the Instigation of Communal Violence in Gujarat, India|journal=World Development|doi=10.1016/j.worlddev.2009.11.029|volume=39|issue=2|pages=221–230}}</ref> Therefore, the existence of these local patronage channels generates incentives for ethnic groups to engage in politically motivated violence.<ref name="Berenschot2010" />

While the link between ethnic heterogeneity and under provision of public goods is generally accepted, there is little consensus around the causal mechanism underlying this relationship. To identify possible causal stories, Humphreys and Habyarimana ran a series of behavioral games in Kampala, Uganda, that involved several local participants completing joint tasks and allocating money amongst them.<ref name=":14">{{Cite journal|last=Habyaimana|first=James|date=November 2007|title=Why does ethnic diversity undermine public goods provision?|url=http://www.columbia.edu/~mh2245/papers1/HHPW.pdf|journal=American Political Science Review|doi=10.1017/S0003055407070499|volume=101|issue=4|pages=709–725|s2cid=4498060|access-date=2016-05-07|archive-date=2017-01-22|archive-url=https://web.archive.org/web/20170122192951/http://www.columbia.edu/~mh2245/papers1/HHPW.pdf|url-status=dead}}</ref> Contrary to the conventional wisdom, they find that participants did not favor the welfare of their co-ethnics disproportionately. It was only when anonymity was removed and everyone's ethnicity was known did co-ethnics decide to favor each other. Humphreys and Habyarimana argue that cooperation among co-ethnics is primarily driven by reciprocity norms that tend to be stronger among co-ethnics.<ref name=":14" /> The possibility of social sanctions compelled those who would not otherwise cooperate with their co-ethnics to do so. The authors find no evidence to suggest that co-ethnics display a greater degree of altruism towards each other or have the same preferences. Ethnic cooperation takes place because co-ethnics have common social networks and therefore can monitor each other and can threaten to socially sanction any transgressors.<ref name=":14" />