Model specification is a critical step in demographic research. In model specification, control variables serve as an important tool for ensuring that the estimate of a relationship between X and Y is properly specified. However, control variables have often been uncritically included and potentially misspecified in observational social science, risking several types of bias. Using the journal Demography as a window into the discipline, we reviewed the use of control variables in statistical models published in the journal from 1995 to 2020. Results show that control variables are usually under-explained, making it difficult to evaluate their quality. When they are explained, variables are often included with logic that is out-of-step with current causal inference literature. There were just marginal improvements over the study period, and results were similar across topics. We discuss the implications for demographic science and recommend that a more rigorous selection of control variables and clear descriptions of that decision process are detailed as part of demographic scholarship.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.