Currently, the debate on American colleges and universities is dominated by the cost of tuition. This debate misses the point.

Let’s look at two of America’s leading schools—Harvard University, a private research university, and Williams College, a private undergraduate college. In 2010, Harvard’s endowment income, investment income and gifts came to $158,106 per student. For Williams, that figure was $91,206. Net tuition (some students pay full tuition and others benefit from scholarships that lower the cost) at each school was $19,834 at Harvard and $19,994 at Williams. Harvard spends $151,014 per student—of which $79,380 goes to instruction, $48,256 goes to research and $5,320 to fellowships. Williams spends $79,536 per student—of which $63,172 goes to instruction.

Let’s compare that with a school dedicated to teaching low-income students: Grambling State University in Louisiana. In 2010, Grambling’s support from the state of Louisiana plus its income from investment and endowments came to $4,809 per full-time student. Net tuition was $4,310 a year. Grambling spends $20,460 per student—of which $13,856 goes to instruction, $83 to research and $2,012 to fellowships.

I offer these stark contrasts to make several points. First, tuition covers only a small portion of instruction at Harvard, Williams or Grambling. Supporting a university education takes a large amount of non-tuition revenue. But the stark differences in amounts of non-tuition revenue mean that if we want all students in America to have access to a quality education, we must either talk about increasing our public investment or accept we will be charging students to try to buy quality. And this puts a different light on comparing a percentage increase in tuition at Grambling with a percentage increase at Harvard. Between 2008 and 2013, the state of Louisiana cut funding per full-time college student by 42%. An $80 per student cut to Grambling passed on to students in a tuition raise (to try to keep up with Harvard’s resources to recruit faculty and provide classroom technology) would hike tuition by 1.8%; an $80 increase at Harvard would be a 0.4% increase. Dollars don’t guarantee quality, but the scale of difference between 5.7-to-1 and 4.55-to-1 on instructional budgets between Harvard, Williams and Grambling suggests an unbearable level of inequality.

Second, an important level of expenditures in research and fellowships support graduate education and research at Harvard and Grambling. America’s universities are our engine of innovation and technological growth. Whether it is Hewlett-Packard, Microsoft or Apple, our economy is powered by research coming from our colleges. It is what makes them a major services “exporter,” vital to our national trade balance. (When foreign students attend American universities, their tuition is counted as exporting educational services.) So investment in our universities is vital to generating patents and addressing our current account deficit.

Second, an important level of expenditures in research and fellowships support graduate education and research at Harvard and Grambling. America’s universities are our engine of innovation and technological growth. Whether it is Hewlett-Packard, Microsoft or Apple, our economy is powered by research coming from our colleges. It is what makes them a major services “exporter,” vital to our national trade balance. (When foreign students attend American universities, their tuition is counted as exporting educational services.) So investment in our universities is vital to generating patents and addressing our current account deficit.

In 1975, state and local governments contributed 60.3% of all expenditures on higher education. By 2010, that figure had fallen to 34.1%. We have been forcing public universities to compete with each other, and with private colleges, for non-tuition revenue. It is a race that was not level and is not level. To reduce costs, public four-year colleges have cut instructional costs per student since 2008, and they have been reducing research per student since 2004.

Some argue that shifting the cost burden to students and their parents with increasing student debt loads is workable, because the returns on a college degree remain high. While the wages of recent college graduates, adjusting for inflation, have fallen 7.7% since 2000, those with a high school diploma have seen wages fall 10.8%. So while the real earning capacity of college graduates is falling, they would be even worse off if they didn’t go to college. And while a bigger share of their earnings may go to pay college debts, they may still be better off—so long as the increase in debt isn’t bigger than their 73% wage advantage. I’ll excuse you if you think that’s an odd way of looking at things.

But is that the way to address funding our national need for sustaining our universities’ global competitiveness, and keeping up vital research or providing equal opportunity? The question is bigger than “is college debt sustainable as costs for individuals” because the benefits are bigger—having new patents, more exports and citizens smart enough to understand global warming debates.