Time point X: education institution keeps records of who it has enrolled, including demographic records of each student. Country keeps demographic records of which demographic groups are currently earning what income (census and income tax records).
Time point Y, 45 years later: Lifetime income of each individual can be calculated, and demographic income distribution can be compared with the income distribution at time point X. For demographic groups which have increased in relative income earning rate compared to time X, the educational institution that had enrolled them at time point X is awarded a prize proportional to the growth, imaginable as a portion of the GDP growth that would not have occurred had they not been the educators.
Educational institutions are thus incentivized to think capitalistically: seek out underserved demographics which will offer the most bang for the buck, for which education will cause the most growth in income earning.
It may also provide incentive to define new demographic categories: the growth of a narrowly defined especially underserved demographic is not diluted by less growth of a more broadly defined demographic.
Weird unintended consequences might occur, especially since battling over proportion of income distribution is a zero-sum game.
45 years is a very long investment term, but the financial markets will figure it out.
Originally conceived just for primary education, because children don't have control over their education decisions. Adults in secondary education seem to have incentives to maximize the benefit of their education. Nevertheless, should the proposed scheme also be deployed for secondary education?