[WS:] I am not sure about this. JHU receives over 50% of its revenue from this source, and only about 12% from tuition, and I suspect it is not that much different from other private research universities.
My own hypothesis is a bit different - it is a version of the Baumol effect http://en.wikipedia.org/wiki/Baumol's_cost_disease. Unis are labor intensive and cannot increase labor productivity the way industry does. In the "good old days" some teaching was subsidized by research and government, but with expansion of teaching (i.e. more students being admitted) the marginal increase of cost of that teaching would have to be covered solely by tuition increases, assuming that the research funding and government grants remain constant. If they are cut, however, the unis have to pass this decrease on the tuition in addition to covering marginal increase in the cost of teaching due to increased enrollment.
Another element is linked to competition among unis. To attract students they need to invest into various amenities (recreational facilities, celebrity faculty, services, etc.) and that has to be covered solely by tuition or private grants, since research grants (esp. government research grants) cannot be used for this purpose. The expansion into real estate market that you mention would fall into this category, I think.
Finally, there is competition form private industry. In order to attract celebrity faculty, they need pay competitive salaries, but they cannot get greater productivity out of them the way industry can (consistently with the Baumol effect.) This is particularly the case of fields like law, economics, medicine or engineering, where opportunities for industry employment are greater than in liberal arts or social sciences.
Wojtek