Measuring IPV From Various Population Administrative Data Sources

• In Event: New Developments in Research on IPV and Sexual Abuse in Data-Intense Denmark

Abstract

Measurement issues loom in existing research on intimate partner violence (IPV), and how researchers define “partners” and which data sources they use for measuring IPV may be profoundly important for the conclusions they can draw. This paper uses extraordinarily precise administrative data from Denmark to draw out the consequences of definitions and data sources in the study of IPV. Combinations of daily records of marital status, cohabitation, and co-parenting allow us to observe several definitions of partners and cleanly distinguish between current and ex-partners. And combinations of daily police records and emergency room records allow us to apply several measures of IPV. With these data, we show how women exposed to IPV (by different definitions) are distributed across these different definitions of “partners”, and we benchmark the distribution against the population distribution. And we present results for the prevalence of IPV – again using several definitions of partners and IPV – across the 2001-2021 period to document time trends. Results indicate that the distribution of types of relationship is very different among those exposed to IPV and the population; that as much as 50-70 percent of recorded IPV is committed by ex-partners, not current partners; that the prevalence of IPV varies greatly by data source used; and that across all definitions, the prevalence of IPV as obtained from police and emergency room records increased over the 2001-2021 period.

Authors

• Lars H. Andersen, ROCKWOOL Foundation

• Melanie Lautrup, ROCKWOOL Foundation

• Sara Wakefield, Rutgers University-Newark

• Christopher Wildeman, Duke University