This paper provides new estimates of the effect of Catholic school attendance on high schoolcompletion and university commencement and completion for Australian students. First, an
instrumental variables approach is adopted where the probability of Catholic affiliation is used as an instrument. Consistent with the recent US literature, results based on this
instrument are mixed. Instead, bounds are placed on the Catholic school effect using the assumption of equality between selection on observables and unobservables. The effect of
Catholic school attendance is found to be smaller than previous results and negative treatment effects cannot be ruled out. Recent improvements in public school outcomes may have
contributed to the smaller Catholic school effects.