We propose a new methodological framework for studying status exchange in marriage. Highlighted by recent debates on race- and beauty-status exchange among American couples, the conventional contingency table approach is prone to controversial model specifications and interpretations. The ambiguity and disagreement mainly concern two methodological issues: balancing the differential distribution of characteristics and identifying the exchange. Log-linear models rely on complicated multi-way interaction terms for balancing and identification simultaneously, which easily conflates the two and produces results too sophisticated to interpret. Instead, we employ gender-cohort-specific relative ranking to balance the status distribution, and non-parametric matching to identify homogeneous and heterogeneous couples for pair-wise comparisons holding one spouse’s characteristics equal. Our straightforward Exchange Index measures the average within-couple status difference between the matched couples. We study the race- and age-education exchange based on the 2000 US Census 5% microdata sample to communicate our improved methodological parsimony and flexibility with existing studies.
Presented in Session 120. Marriage, Assortative Mating and Inequality