Interracial unions in Brazil, 1980-2010: does religion matter?

Luciene Longo, Instituto Brasileiro de Geografia e Estatística (IBGE)
Paula Miranda-Ribeiro, Cedeplar, Universidade Federal de Minas Gerais
Ana Hermeto, Cedeplar, UFMG
Thomas W. Pullum, Demographic and Health Surveys

Brazil has experienced important changes in religious affiliation in the last four decades. In 1970, 92% of the population declared to be Catholic. Thirty years later, the proportion had fallen to nearly 74%. The proportion of Pentecostals almost tripled (5% to 15%), whereas those who declared to have no religion raised from 1% to 7%. In addition, the country has the largest African descent population outside Africa. In 2009, 44% of the Brazilians considered themselves “pardos” or mixed. However, there is a high proportion of endogamous marriages, with low rates of marital exchange among different race groups. The objective of this paper is twofold. First, we investigate the differences between interracial marriages according to religious affiliation between 1980 and 2010. Second, we verify whether female religious affiliation compensates for racial differences. Data come from the Brazilian demographic census of 1980, 1991 and 2000 (IPUMS-International data). We analyze women 20-29 years-old and their husbands/partners. Formal marriages and consensual unions are investigated separately. Loglinear model results comparing 1980 and 2000 suggest that, when the husband/partner and his wife/partner have the same religion, there is practically no change in the percentage distribution of marriages by race, compared to the overall distribution. However, when marriages between partners of different religious affiliations were analyzed, there was a decrease in all homogamous marriages, whereas interracial marriages increased. Although there was a raise in interracial unions between 1980 and 2000, heterogamous unions by religion have not changed in the same pace. Therefore, the strength of religion is highly relevant to unions in general but religion does not compensate for racial differences in Brazil. The next step is analyzing the last decade, based on data from the 2010 census, to be released in the beginning of 2012.