In 1945 Friedman and Kuznets presented a radical proposal for financing professional education which Friedman 10 years later extended to higher education in general. Their proposal was, in essence, that individuals sell "share" in their future income streams in order to finance investment in themselves. This paper discusses why and how the Friedman proposal must be modified in order to take account of information costs and certain political and social consequences involved in "risk-rating" individuals. The recently introduced Yale Tuition Postponement Option is used to illustrate a number of difficult problems involved in actual implementation of the Friedman proposal. It is shown that the consequences of income-contingent loan repayment programs and their financial viability in the absence of subsidy depend crucially on a number of empirical aspects of the behavior of participants and potential participants which are, at present, poorly understood.