Data from the Youth Cohort Study of Britain are used to estimate hazar
d functions for time to first job by school leavers. Participants in t
he Youth Training Scheme (YTS) are compared with non-participants. Whe
n time taken to obtain a job is modelled, YTS trainees obtain jobs at
a slower rate than non-trainees even when time spent on YTS is exclude
d. However, female YTS trainees obtain 'good' jobs at a faster rate th
an non-trainees when time spent on YTS is excluded. Our main results a
re robust to the inclusion of a selectivity effect, and to the inclusi
on of local labour market conditions as a time-dependent covariate.