This paper uses household survey (Sakernas) data from 2004 to estimate the determinants of earnings in Indonesia, a country where non-salaried work is widespread and where earnings data are available for salaried employees only. We deal with the selection bias by estimating a full-information...