Conventional wisdom says that expansions in federal student aid will result in a more affordable and equitable postsecondary education system. While this belief has motivated massive expansions of federal aid in the recent past, rapidly increasing tuition and student loan default rates are raising questions about this approach.

In a new study, I review the basic statistics and recent research, and conclude that: 1) further increases in federal support for higher education are likely to be counter-productive because they lead to higher tuition for all students, and 2) the system of student loans should be reformed to mitigate the student debt problem.

Average gross tuition and fees for undergraduate studies increased more than three-fold in constant dollars from 1980 through 2014—even faster than the rate of increase in health care prices. The increase is widespread across several types of higher education institutions: private non-profit, public two-year and four-year—although private for-profit institutions have recently seen a decrease.

At the same time, government support for higher education in the form of tax benefits, grants (veterans and Pell), and loans has exploded since 1994, and especially since 2000. Grants and loans totaled almost $170 billion in 2014 (constant dollars), up from just over $50 billion in 1994.

Whereas earlier studies showed mixed results, recent studies using refined data and techniques consistently find that increased federal support for higher education leads to a significant increase in tuition and decrease in institutional aid. In fact, there is some indication that state aid for higher education is itself negatively related to the extent and nature of federal government support.

Increases in federal support for higher education have been focused on student loans, yet these increases are actually detrimental to the finances of many post-secondary students. One study found that students’ college decisions were almost as responsive to the offer of loans as to grants, even though they have to pay back loans.

Different measures of student loan default rates and repayment burdens uniformly show significant increases in recent years (as much as a doubling), across all types of institutions and students. Furthermore, the expanded use of deferment, forbearance, and, especially, income-related repayment plans hides an even larger increase in losses to taxpayers on these loans than the rising default rates would indicate.

Students from nontraditional backgrounds seem to have been harmed most by the increase in federal student loan amounts available. They have not seen increased incomes as workers, often have not completed their educations, and are much more likely to default on their loans, while missing out on job-related income and training while enrolled in college.

In addition, enrollment in higher education, as a percent of the young adult population, and in the supply of college-educated workers as a percent of the workforce, has steadily increased over the past four decades. Almost 70 percent of recent high school graduates are currently enrolled in some type of college. The increase is most notable for public two-year and for-profit colleges. Except at the top 10 percent of colleges and universities, average student quality in higher education institutions has declined.

The earnings premium for a college education over a high school education has held steady over the past 25 years. This contradicts the claim by some economists that the relative supply of college-educated workers has been dropping and the earnings premium increasing. It also debunks the claim as an explanation for increased income inequality, or a justification for even more government support for higher education.

Further increases in federal support for higher education are not needed and indeed are likely to be counter-productive because they lead to higher tuition for all students. The resultant increase in tuition, decline in average student quality, stagnation in wages, and increase in student loan defaults should lead policy makers to question whether the massive increase in federal support for higher education is achieving its goals.

It is quite likely that reducing subsidized student loan availability to upper middle-income families could be beneficial to the system by leading to a general lowering of tuition and reducing the loan burden on future workers.

The entire system of student loans, especially repayment options, needs to be rationalized and redesigned; in particular, the salience of loan repayment to students needs to be strengthened and losses to taxpayers reduced.

There are many indications that a system of universal higher education, which is where we have been heading, is wasteful. This is particularly true if college completion is mainly a signal of perseverance and effort more than actual preparation. Rather, a robust parallel system of on-the-job training, apprenticeships, and youthful practical work experience is needed—supported by changes in federal laws and regulations, such as lowering the minimum wage for younger workers, to encourage its creation.