Originally posted on Yahoo! Finance.

By Andrew P. Kelly

On July 1, the interest rate on roughly one out of every three new student loans is set to double, from 3.4 percent to 6.8 percent. Unless Congress steps in to stop it.

Having déjà vu? You should be: We had this same conversation last year. Except this time, the arguments from student advocates have become even more divorced from fiscal reality.

To refresh your memory: In 2007 Congress set the interest rate on subsidized Stafford Loans, which largely go to low-income borrowers, at 3.4 percent. The rate was due to sunset last July, but Congress, the President–even Nominee Romney–all sought to out-pander one another to extend the policy for a year. The temporary fix cost taxpayers about $6 billion.

But surely the benefits to borrowers were huge, right? Hardly. All that fuss saved them about $9 a month.

What will happen this time around? In his new budget, President Obama has suggested tying rates on all student loans to the government’s cost of borrowing. But of course, presidential budgets are just glorified suggestions until Congress acts on them. Meanwhile, student advocates have already dusted off the “don’t double my rate” hash-tag that got so much airtime in last year’s campaign.

But as the New York Times‘ Tamar Lewin points out, the latest arguments also harp on the purported profits the federal government makes off of student loan programs. Because the government earns money on its lending, advocates argue, policymakers should not let interest rates go higher, but should keep them low.

In the policy brief at the center of Lewin’s story, the Public Interest Research Group, Young Invincibles, and the U.S. Student Association lay out the argument:

Recent projections from the Congressional Budget Office, the official scorekeeper for federal programs, indicate that federal educational lending now carries a “negative cost subsidy” of 36.48 percent for 2013. On average, every dollar lent will yield more than 36 cents of profit to the federal government. . ..

The federal student loan program as it currently operates is the opposite of a low cost program to student loan borrowers; it makes billions in revenue yearly.

Terry Hartle from the American Council on Education (a leading higher education trade association) broke it down more practically in the Times:

If the numbers are accurate, the government will make more money on student loans than Ford makes on automobiles,” he said. “Using student loans to create a profit center is not what anybody intended.

There’s just one problem with these numbers: they’re entirely dependent on what method the Congressional Budget Office (CBO) chooses to use in accounting for the cost of loan programs. Traditionally, the CBO has used a procedure laid out in the Federal Credit Reform Act (FCRA) of 1990 to assess the cost of federal loan programs. This method produces the gaudy student loan surpluses cited by student advocates (and displayed here, in CBO’s 2013 forecast for student loans).

But by the CBO’s own admission, the FCRA procedures don’t account for the cost of “market risk,” or the potential for economic forces out of the government’s control to affect the ability of borrowers to pay back their loans. When the economy goes south, for instance, borrowers are more likely to be delinquent or to default, leading to lower recovery rates and higher collection costs for the government.

Luckily, the CBO has recently experimented with another method–“fair value accounting”–that better acknowledges the cost of market risk. What difference does it make? A lot, it turns out. The six education loan programs go from a surplus of $36.3 billion under the government’s traditional accounting methods to a surplus of just $5.5 billion under fair-value in FY 2013 (see Table 1 here), or from about 32.1 cents for every dollar lent to 4.9 cents. That’s just for one year, and doesn’t count administrative costs. Map the difference over the standard 10-year budget window and we’re talking big money.

Which method is right? As Jason Delisle of New America (a proponent of the fair value approach) argued last November, if you take leading, non-partisan finance experts as a litmus test, the answer is pretty clear:

Even if it is Republican lawmakers who have taken up the mantle of reform for fair-value estimates, the non-partisan views of organizations like the CBO, the Federal Reserve Bank of Kansas City, and the Financial Economists Roundtable should make it clear that this is hardly a one-sided partisan issue.

The first step toward making our student aid system more sustainable and effective is to acknowledge, as fully as possible, what each program costs us each year, and whether those dollars are well-spent. Until then, we’ll continue to make policy based on politics and deadlines, not sound accounting and common sense.

In that spirit, policymakers and journalists must recognize that the Ford-esque “profits” attached to student loans may be more fiction than fact, an artifact of the way Congress has chosen to measure the cost of federal loan programs.