Religious parties are a puzzle for political scientists. Some scholars consider them to be ideological and antidemocratic. Others find them to be opportunistic and mediating. One approach in literature is to resolve these contradictions by separating ideological religious parties from mediating Christian Democratic parties. However, one can imagine that religious parties advocate support for religion and, at the same time, maximize their vote share by adapting to their electorate’s (non-)religiosity. In this case, religious parties would be rational actors that adapt their strategy situationally, but in a principled manner. With the rise of religious parties in the Arab world and the persistent influence of religious parties in the West, it is important to understand how religious parties act, based on their societal context.
In my paper, I develop a model of religious parties, building on the economic theory of democracy (Anthony Downs) and the economic theory of religion (Laurence Iannaccone). Using a Monte Carlo simulation, I theoretically demonstrate how beneficial it is for parties to address religious voters in comparison with non-religious ones, even in a pluralistic society. Afterward, I differentiate policies related to religion with respect to the voter’s individual costs and benefits. Status-granting policies on religion impose indivisible costs and benefits on voters. Redistributive policies on religion impose divisible costs and benefits. Religious freedom policies impose costs and benefits in relation to the religious setup of a society and the voter’s (non-)religion. If religious parties are principled vote maximizers, their advocacy for status-granting policies should correlate positively with the share of religious voters in the electorate, as these policies are too costly for non-religious voters. In contrast, their advocacy for redistributive policies should rather be independent of social religiosity, as these policies are less costly for non-religious voters. In more secular societies, in particular, religious freedom policies should also be easy to implement.
These hypotheses are tested in the “laboratory” of the constitution-building processes of the German states. This approach has two major advantages: First, constitutions are a far better indicator for the outcome of political decisions than the “cheap talk” (Laver/Schofield) of political declarations or hard-to-explain budgetary fluctuations. Second, the German states vary strongly with regard to their confessional setup and the majorities in their constitutional assemblies. Some states’ societies are predominantly Catholic or Protestant, others are near-completely non-religious. Some constitutional assemblies have a dominant religious party (even in secular states), while, in others, religious parties are negligible. A multiple regression analysis shows the correlations of the number of precedents of the policy on religion of the categories built in the model, the religiosity of the population, the strength of the religious parties in the constitutional assemblies, and some control variables.
The results show—as predicted—that status-granting policies correlate strongest with the religiosity of the population, while redistributive policies correlate with the religiosity of the population and the strength of religious parties in the constitutional assemblies. Thus, religious parties—at least in the case of the German states—are highly responsive, but nonetheless purposeful parties.