The ever-rising cost of a college education in the United States is causing policymakers across the United States to consider ways to reduce those costs. With good reason. The amount of student debt has ballooned in recent years while new research shows that the students struggling the most with student debt are those who attended for-profit schools and community colleges many of whom were on the edge of attending or not. The federal government tries to encourage students to pursue higher education by providing loans —a subject for another day—but it also offers a number of incentives through the tax code via credits and deductions to lower the perceived price of college.

But looking at the research on these programs, federal policymakers would do well to rethink them. Tax credits and deductions could plausibly increase attendance at college and universities by reducing the price individuals face when paying for college. Think of it this way: If college applicants find out they will receive a deduction for going to school then the price of tuition seems to be lower and should spark more prospective students to “buy” college.

Of course, there’s the possibility that prospective students treat the tax credits and deductions as an increase in their income that doesn’t affect their schooling decisions. Rather, the credit just acts as a tax cut for people who already were planning to attend college. This second reaction is what happens more often than not, according to a clutch of recent research.

A working paper published earlier this week by the National Bureau of Economic Research finds that the federal tax deduction for college tuition has essential no effect on college attendance or a variety of other metrics on investment in higher education. Economists Caroline Hoxby of Stanford University and George Bulman of the University of California-Santa Cruz use a technique called “regression discontinuity” to determine if the tax deduction has any causal effect by comparing the actions of taxpayers just above and just below the income cut-off for the deduction.

What they find is not encouraging for those policymakers who hope the deduction increases investments in human capital via higher college attendance. The two researchers find no causal effect on “attending college (at all), attending full- versus part-time, attending four- versus two-year college, the resources experienced in college, the amount paid for college, or student loans.” In other words, it’s hard to see how the tax deduction does anything but act as a tax cut for those already set on attending college. The tax deduction studied by Hoxby and Bulman also has important distributional element: It is most valuable to tax filers with higher tax rates, but these are not the students who need the most support. Research from earlier this year by Hoxby and Bulman show very similar results for tax credits as well. Credits also appear to have no causal effect on investment in higher education.

Perhaps these results shouldn’t be too surprising. In a variety of areas of public policy, tax credits and deductions are used as backdoor means to promote diverse ends, including efforts to increase homeownership and retirement savings. Of course, there are very successful tax credits, among them the earned income tax credit, though even in this case some economists argue employers capture a significant amount of the value of the credit.

The reflex among policymakers to reach for the tax code as a means of providing greater access to important goods and services is a behavior they might want to change.