Data Collectionhard
0:00.0

A researcher wants to estimate the causal effect of college attendance on lifetime earnings. Family wealth confounds this relationship (wealthy families both facilitate college and lead to higher earnings regardless of education). The researcher proposes using "distance to the nearest university" as an instrumental variable for college attendance. Which assumptions must this IV satisfy for valid causal inference?