We investigated the causal effect of college education on smoking, drinking, sleeping, and cancer screening behavior in Japan. To estimate said effect, we leveraged a unique instrument in which a mismatch between Japanese superstition and school year in 1967 leading to an increase in college...