of the role of money contributions in funding the mobilization of activists. The obvious impli-
cation is that not only the intensity of ethnic feelings matters, but the abundance of monetary
contributions is key as well. The fact that ethnic conflict is intrinsically multidimensional makes
the relationship between polarization and conflict subtle. If we focus on some indicator of ethnic
similarity or distance, it is indeed true that greater group homogeneity will be recorded as an
increase in the polarization of ethnic sentiments and will correspond to more intense conflict, as
expected. However, when the salient social cleavage is ethnicity, more within-group economic
homogeneity may actually reduce conflict, as the perverse synergy between those willing to
supply conflict resources and those willing to supply conflict labor is reduced.
Notice that this is not the case for class conflict. There, wealth determines both the ability to
contribute to the cause and the sense of identity and alienation. For economic classes, wealth is
the appropriate variable for the polarization/conflict nexus. For ethnic groups, it is not. Wealth
differences across groups may have very different implications in the two cases. One therefore
needs prior knowledge on the role of a particular dimension in a conflict in order to be able to
say whether increased “income polarization” will lead to higher conflict or not.
This discussion suggests two important features which are not part of the current exercise.
First, this paper assumes that the situation is inherently conflictual, and we make no attempt to
model group decisions to enter into conflict in the first place. These are undoubtedly important
issues. but are omitted here so as topermit clearer focus on the questions of intra-group and
inter-group heterogeneity.
To summarize, then, we construct a simple model of ethnic conflict. The model assumes
that there are inter-group antagonisms, the intensity of which can vary from individual to in-
dividual. We take no position on whether such antagonisms are economic, political, or indeed
primordially ethnic, but take it that they affect individual contributions to conflict. In addition,
individuals can directly participate as activists, but such participation needs to be compensated
(this includes “self-compensation”, such as the willing bearing of opportunity costs). This model
permits us to analyze different sorts of factors, which we’ve broadly classified into inter-group
and intra-group categories. Among other things, we show that
An increase in activism by one contending group will create escalation or deterrence depending
on whether that group was weaker or stronger to start with; this has implications for the relation
between polarization and conflict;
Balanced increases in the wealth of one group might reduce its activism: the higher money
contributions may be insufficient to cover the higher costs of mobilization; but
An increase in wealth inequality within one ethnic group may make inter-ethnic conflict more
violent: mobilizing the poor is cheaper, and the rich have more resources to implement that
mobilization.
2. The Model
We will use the notation H and M to denote two ethnic affiliations, often referring to these
as Hindu and Muslim respectively. [But the model that we lay down is more abstract and can
be applied to other conflictual situations quite easily.] There are nh and nm people of each
conviction, with nh + nm = n, the total population.
In this paper, we place attention on “religious policies”, to be thought of as actions or con-
cessions (or social attitudes) that are effectively tilted towards one religious group or the other.
To be concrete, suppose that there is some “status quo” policy to begin with, against which an
alternative proposal has been raised. For instance, fundamentalist Hindus might (and did) call