Abstract
Recent findings of non-linearities in the returns to education are cited as evidence of sheepskin effects, but evidence from Guatemala suggests otherwise. While non-linearities in the returns to education exist, it is argued that they are associated with ‘threshold’ levels.