Research Paper CSAE Working Papers WPS/2016-38
The twin instrument
Twin births are often construed as a natural experiment in the social and natural sciences on the premise that their occurrence is quasi-random. We present new population-level evidence challenging this premise. Using data on about 18 million births in 72 countries, we find that maternal condition is positively associated with twin birth. Thus studies using twins to instrument fertility will under-estimate impacts of fertility on parental investments and women’s labour supply. This is pertinent given recent research indicating these relationships are weak. Using developing country and US data, we demonstrate the bias and estimate bounds on the true parameter.