Consent to Data Linkage: Experimental Evidence from an Online Panel

Abstract

Longitudinal surveys have tended to have a strong focus on primary data collection, with most of the analytical information obtained directly from interviews. The next generation of longitudinal studies will likely make extensive use of linked data to augment survey responses. However, this will occur in the context of declining response rates to surveys and in consent to data linkage. Surveys on probability based online panels have several advantages in that they are cheaper and faster to implement than traditional longitudinal surveys but also have the capacity to deliver complex survey instruments. There are different types of information linked to survey data, with consent usually asked for specific sources of data. Evidence from observational studies where consent to link different types of data has been requested report consistent patterns of consent rates with consent to link highest for education records followed by health, with income or economic records having substantially lower levels of consent.