El 25 de junio se llevará a cabo un nuevo seminario de investigación del Programa de Población, a cargo de Matías Dodel, docente del Departamento de Ciencias Sociales de la Universidad Católica del Uruguay, y Andrés Wilkins, docente del Departamento de Sociología de nuestra Facultad.
Será a las 13 horas en la sala 2 y 3 del piso 5 de Facultad de Ciencias Sociales.
Disentangling love from age in the digital world: a quasi-experimental approach to Age–Period–Cohort in online couple formation
This article examines how age, historical period, and cohort membership shape online couple formation in contemporary societies, addressing the classic Age–Period–Cohort (APC) dilemma within the context of digital partner search. Drawing on a life-course perspective, the study situates online dating within broader transformations in intimate relationships while critically engaging with existing theoretical assumptions about individualization and the deinstitutionalization of marriage. Empirically, the study analyzes online and offline couple formation using harmonized survey data (Generations and Gender Surveys, second round) from the capital cities of Scandinavian (Sweden and Denmark), Eastern European (Estonia and the Czech Republic), and South American (Argentina and Uruguay) countries. The unit of analysis is the individual person-year, enabling a dynamic examination of partnership transitions from 15 through age 50. Rather than relying on a conventional APC model—which faces well-known identification problems—the analysis exploits a quasi-experimental design. By leveraging cross-national variation in internet diffusion, the study observes individuals from the same birth cohort at the same age facing substantially different levels of internet penetration across countries. Survival analysis and hierarchical linear models (HLM) are used to compare patterns of online versus offline couple formation across cohorts and countries. Descriptive results show a clear cohort gradient: younger cohorts are substantially more likely to have formed couples online than older cohorts. Cross-national differences are pronounced among older cohorts, reflecting unequal patterns of internet diffusion, but tend to converge among younger cohorts. Multilevel models confirm that age at event and cohort membership remain strong predictors of online couple formation, even after accounting for internet penetration. Greater internet availability also increases the likelihood of forming couples online. Robustness checks reinforce these findings: when the same models are estimated for couple formation as a whole (online and offline combined), age and cohort effects remain structurally similar, but the internet penetration effect collapses to near-zero (OR = 1.001), indicating that connectivity specifically redirects partner search toward digital venues rather than accelerating couple formation in general. A complementary specification incorporating respondents’ previous-partner trajectories shows that life-course dynamics are not fully captured by chronological age: each additional previous partner increases the odds of online couple formation, while age, cohort, and internet effects remain stable. Overall, the findings suggest that online dating does not fundamentally disrupt age- and cohort-based patterns of couple formation, but rather operates within them.