High survey nonresponse in unemployment duration studies may have a strong effect on inference if exit from unemployment affects the chance of nonresponse (nonresponse causality). In rotational studies large part of the nonresponse results from panel attritions. A method to test the presence of the causality mechanism for rotating panel designs is proposed and its asymptotic consisten... https://www.thebanyantreers.shop/product-category/coaster/