Empirically examining the role of religion in social phenomena requires accurate measures. This chapter has three purposes. First, it describes the variables used in this study to measure government involvement in religion (GIR). Second, it describes and critiques the types of variables used to measure religion in previous studies, as well as how they influenced this study's variable design. Unfortunately the discussion of past attempts is a short one because until recently religion has rarely been included in cross-national empirical studies of social and political phenomena. When it is included, the measures are often limited, crude and indirect. Third, this chapter reviews many of the empirical findings that correlate religion with social and political phenomena.

PAST MEASURES OF RELIGION

Past measures of religion generally use one of four methods, each of which measures a different aspect of religion. First, many measure some aspect of religious identity – whether individuals or populations are nominally members of specific religions. Second, some focus on religious diversity within a state. Third, many use survey data to measure religiosity – the extent to which individuals are religious. Fourth, some measure the involvement of social and political institutions and groups, including governments, in religion. I will discuss each of these types of variable.

Identity-based Variables

The majority of religion variables used by cross-national quantitative studies are identity-based variables. They measure religious identity in a number of ways.