Abstract
This article links recent conceptual theories regarding the determinants of fertility with research designs appropriate for testing those theories. The most important causal factors in these theories—typically social or cultural institutions, occasionally emergent properties of the collective behavior of individuals—are properly conceptualized at the macro level. Research designs must therefore feature variation at this level and are at a minimum comparative. Noncomparative micro-level research designs will be misspecified from the standpoint of theory. A case is also made for continuing to incorporate the observation of individual-level behavior into any comprehensive research design. A focus on macro determinants of fertility does not imply that fertility outcomes are determined at the institutional level. Instead, there remains some process whereby systemic properties are translated into individual behaviors. Recommendations for comparative community-level studies are discussed. The multilevel analysis framework is reviewed as a paradigm for the conceptual features of cross-contextual analysis.