Minority Participation and Well-Being in Majority Catholic Nations: What Does it Mean to be a Religious Minority?

Italy was omitted from the final analysis because it contained no religiously active non-Catholics with available data on all of the variables included in the analysis.

All forms of Protestant Christianity make up the second largest denomination in 6 of the 10 countries included in the analyses, and 61.53 % of the cases in the sample belong to a Christian denomination. To be certain that our findings are not the product of the size of each case’s religious group or type (but rather are due to the size of the majority religion), we include a series of dummy controls for religious preference in each analysis.

In the United States, for instance, every religious denomination is a numerical minority. Although the USA is a predominantly Christian nation, differences between Catholics, mainline Protestants, conservative Protestants, and other Christian groups make the experience of Jews in the United States very different from the experience of Jews in Poland where the Roman Catholic Church makes up 97 % of the religious market.

Argentina is one of several Catholic nations with a “nominal” Catholic majority. In order to account for this difference, we re-ran the analyses with a dummy variable demarking each nation with a non-practicing Catholic monopoly as indicated by the CIA World Factbook. This variable was highly correlated with the Gini coefficient of income inequality and did not substantively change the results in any meaningful way.

This measure includes individuals who identified a religious preference but are not currently a member of a particular religious organization; it does not include atheists and the non-religious who were omitted from the analysis.

Significance is based on a Chi-square test using the frequency distributions (not shown) of active members and religious participation across the 10 countries in this analysis.

In additional analyses (not shown here), we included two other country-level control variables. The first is a measure of the government regulation of religion and the second is a series of dummy controls for geographic region of the world (Latin America, former east bloc, etc.). Both of these measures were significant predictors of individual well-being; however, due to issues of multicollinearity between the first measure and our measure of Catholic population share and the second measure and the Gini coefficient for income inequality, we omit these variables from the results we present here. Including the dummies for geographic region did not significantly alter the effects in the final model, nor did the Government Regulation Index. We chose to include the Gini coefficient rather than the dummies for geographic region in the models presented here because the former offers greater nuance between the 10 countries in the analysis. Likewise, the theoretical basis for this paper requires that we examine the effect of population share rather than the effect of the government regulation of religion. For a discussion of the latter, see Elliott and Hayward (2009).

Weighting follows the procedures suggested by the World Values Survey Association (2009).

For more on predictive margins, see Williams (2012).

To test for issues of multicollinearity due to the high number of interaction terms in our model, we also ran separate models testing the direct effect of active membership on each of our dependent variables for countries with a Catholic super-monopoly. These models (not shown) are identical to Models 1, 3, and 5 in Table 2. Consistent with our findings in Models 2, 4, and 6 in Table 2, the effect of active membership for non-Catholics residing in Catholic super-monopolies is large and negative.