The religious-secular earnings gap: differences in cognitive ability or different returns on cognitive ability?

Religious individuals tend to earn less than secular individuals. Existing explanations attribute this disparity to weaker career orientation among the religious, to life-course choices associated with lower lifetime earnings, or to earnings themselves influencing religiosity. Yet these accounts neglect a central determinant of earnings: cognitive ability. Cognitive ability is recognized as one of the strongest predictors of earnings and an attribute that also differs between more and less religious individuals. To this backdrop, this study examines (i) whether secular and religious individuals differ in cognitive ability and (ii) whether such differences account for part of the earnings gap or whether secular individuals receive higher labour-market returns from their cognitive ability. Using Swedish register data on men’s cognitive test scores, religious affiliation, and earnings, we apply Blinder-Oaxaca decompositions to separate the earnings gap into components attributable to differences in cognitive ability and differences in the returns to cognitive ability. The analyses show that non-affiliated men have higher average cognitive ability and higher earnings than Church-affiliated men. However, the earnings gap is driven less by differences in cognitive ability than by higher returns to cognitive ability among the non-affiliated, a pattern observed across most of working life. Together, these findings describe how the religious-secular earnings gap is associated with heterogeneity in the labour-market returns to cognitive ability across religious groups.