In the study of ethnicity on public goods provision, the concept of ethnicity has
largely been under specified, resulting in ambiguity in what specific attributes of
ethnicity can prove to be deleterious to public goods provision. This dissertation focuses
on how two specific aspects of ethnicity, rigidity of ethnic boundaries and internal ethnic
social organization, affect preferences for public goods provision. Fearon (1999) argues
that nonporous ethnic boundaries facilitate forming minimum winning coalitions based
on ethnic identity as they more easily exclude others from sharing benefits. Hence, I
argue that this lowers trust between ethnic groups as they fear that whoever comes into
power will hoard government resources. It is thus not inherent antipathy as posited in
many works on ethnic politics but political competition that drives preferences for private
over public goods in diverse polities. Using Pakistan’s recent devolution as a natural
experiment, I show using in-depth surveys that introducing political competition at the
union council level of local government led to a perceived increase in political
significance of local kinship identities. Comparing a homogeneous union council in
southern Punjab with an ethnically diverse union council I find that the homogeneous
polity is more likely to vote by ethnicity, prefer private goods over public goods, and
prefer public goods provision in the regime before the local government system. Yet,
when asked who should benefit from a hypothetical public goods project, they were as
iii
likely to stipulate the entire community irrespective of identity as was the homogeneous
polity, illustrating that it is not inherent antipathy that leads to politicization of ethnicity,
but the fear of being locked out of politics.