Estimates of the Effect of Parents' Schooling on Children's Schooling Using Censored and Uncensored Samples

In this paper we estimate the impact of parental schooling on child schooling, focus on the problem that children who are still in school constitute censored observations, and evaluate three solutions to it: maximum likelihood approach, replacement of observed with expected years of schooling, and elimination of all school-aged children. Plug (2004) - a recent mobility study that relies on censored data - serves as an illustration. With updated and uncensored versions of previous samples, we re-examine Plug's estimates and test how the three correction methods deal with censored observations. The one that treats parental expectations as if they were realizations seems to fix the censoring problem quite well