for the razing of a mosque and the building of a temple on supposedly sacred ground. Or fun-
damentalist Muslims might (and did) decide to enforce problematic aspects of Muslim personal
law, pertaining to divorce or marriage.
In each of these cases, the acceptance of the newly “proposed” situation marks a change in
social attitudes, sometimes but not always mediated by state policy. Different people will, of
course, feel differently about the change. Let us use the individual-specific variable x to capture
the intensity of feelings regarding the proposed shift from the status quo. For the sake of con-
creteness, suppose that the proposed policy is favorable to the Hindu position, relative to the
status quo. Consequently, in the Hindu camp, the variable x will stand for how strongly an indi-
vidual feels about supporting the shift, while in the Muslim camp, x will stand for how strongly
the individual feels about opposing it.5
The different groups will voice their support or dissent regarding the proposed shift in reli-
gious policy. Such activism may take the form of open debate or discussion, but more likely it
will take the form of demonstrations, processions, and riots, and occasionally looting, rape and
murder. “Activists” are needed to engage in these activities. Denote by Ai, i = h, m the number
of individuals of each creed involved in such “activism” and denote by p the probability that
the proposed policy shift will indeed be successful. We assume that p depends on Ah and Am in
the following way:
(1)
p(Ah , Am) =
ψ(Ah)
ψ(Am) + ψ(Ah),
with ψ strictly increasing and concave and ψ(0) = 0.
The individuals in our model vary in three ways. First, as already mentioned, they are H
or M . Second, and also discussed, they can vary in their feelings about the proposed policy;
this is captured by the variable x. Finally, individuals may also vary in the resources under their
command; call this w. To fix ideas we shall think of w as the earning capacity of the individual. It
will be used not only to proxy his wealth, but also as a measure of the opportunity cost incurred
by that individual if he becomes an activist. So in summary, an individual is characterized by his
religion, his religious attitudes and his resources. With religious affiliation given, we will refer
to any combination of the remaining two characteristics as a “type”. Use z = (x, w), i = h, m, to
denote a type. There are ni (z) individuals of each type for each group i.
We assume that each individual makes a decision about how much resources (personal or
financial or both) to contribute to activism. To solve this problem we need to convert units of
money to units of activism. We take up this issue in some detail below; for now, simply assume
that each unit of activism needs to be financed by an amount si . This amount can vary across
the two groups. Then a typical H -individual of type z will seek to choose r to maximize
(2)
px + u(w - r)
where p is given by (1), Ah can be affected by the contribution r, and u(w-r) is the utility of con-
sumption when contributing to the religious cause an amount r by an individual with resources
w. [An analogous expression with 1 - p in place ofp is maximized by the M -individuals.]
We assume fairly standard things about the utility function: that it is increasing and strictly
concave in consumption —and hence convex in r— and that individuals cannot contribute more
than their earning capacity. Notice that the concavity of u implies that increased wealth reduces
the marginal utility cost of contributions.6
5This is a bleak view indeed. Many tolerant Hindus might deplore the proposed shift as well. The model is easily
amended to take this into account.
6Formally, we assume that u is a smooth function with u'(w — r) > 0, u' (w — r) → ∞ as r → w, and u''(w — r) < 0.