Interestingly the data shows otherwise. Factors such as parents' education, where one lives, age, race, family structure, IQ are measured and factored in. Two stats techniques are commonly applied. Ordinary Least Squares (OLS) is more traditional while Instrument Variables (IV) is a newer approach which attempts to account for statistical errors such as omitted and unknown variables in order to more strongly determine causality.
If you like stats check out this review of the literature: http://emlab.berkeley.edu/users/card/papers/return-to-school... which concludes that studies "Taken as a whole, however, the findings from the recent IV literature are remarkably consistent with Griliches' (1977) assessment of a much earlier set of studies, and point to a causal effect of education that is as big or bigger than the OLS estimated return, at least for people whose schooling choices are affected by the supply-side innovations that have been studied so far."
There are real economic reasons why governments invest in their citizens' schooling. It's not just because it sounds like a nice idea. The increased income potential of someone going to school pays off for a government higher earnings which is a primary indicator of an economy's health.