Whether or not to differentiate - or track - students according to ability has been debated over the years. In Sweden, secondary schools that practiced tracking and schools that did not practice tracking existed simultaneously from 1980 to 1997. This variation in tracking status between schools is used in a differences-in-differences approach. I estimate whether tracking math, or not, in Swedish secondary school had any effect on the probability of having graduated upper-secondary school, but also whether tracking had any consequence for the math grade in upper-secondary school. The results show that when considering the attainmentof upper-secondary education and the mean achievement in math, there are no effects oftracking. However, there are effects when estimating the probability of receiving a specific grade, i.e. fail, pass, pass with distinction or pass with special distinction. Tracked students,from families with low-educated parents, are more likely to fail math than similar students ina non-tracked environment.
2006. , 23 p.