A study in Canada cited by Alan Johnson when Education Secretary found that the introduction of tighter provincial restrictions on leaving school between 1920 and 1990 had helped in raising both average attainment and average incomes. The study found that students compelled to attend an extra year of school experienced an average increase in annual income of about 12%. It also found that compulsory schooling is closely associated with significant benefits.
Mike doesn't link to the study, but I suspect he means this pdf. And if he does, there's a problem, because it shows the exact opposite of what Mike and Alan Johnson would like. Here's page 17:
Average grade attainment is 0.24 points higher with a school-leaving age of 14 compared to a lower or no school leaving age...Raising the dropout age to 15 versus 14 exhibits an even larger increase in average grade attainment (0.63 points). However, the effect from increasing the limit age again to age 16 is negative. (emphasis added).
This is exactly what the law of diminishing returns predicts. If someone has little education, there's a decent gain from giving them more. But the more schooling they have, the smaller are the marginal gains from more of it.
Worse still, the paper suggests there might be negative spillovers from higher school leaving ages. It finds (table 2) that raising the leaving age to 16 has a small but significant negative effect upon students' chances of getting degrees. This is consistent with the possibility that forcing people to stay at school disrupts the learning experience for those who want to learn.
There might be a case for raising the education leaving age to 18. But it's not to be found in the paper cited by Mike or Alan Johnson. I hope this isn't supposed to pass for evidence-based policy-making.