The Student Loan Tax

As student loan debt passed the $1 trillion mark, President Obama, speaking at Chapel Hill yesterday, called the upcoming interest rate hike on student loans a tax.  He didn’t tell the half of it.  Congress’ dirty secret is that the government makes a huge annual profit on student loans.   According to the scrupulously nonpartisan Congressional Budget Office, $37 billion will flow IN to Treasury from student loans made this fiscal year at the 3.4% rate (on a net present value basis and net of about $1.5 billion to administer them.)   The President’s current dispute with Congressional Republicans is about whether to increase this annual profit next year.  The interest rate that students pay on the basic “subsidized” loan is slated to rise from 3.4% this year to 6.8% next year, unless the lower rate is extended by Congress.

How does the government profit from student loans?  In two words, yield spread. 

Treasury can borrow money at 0.5% or less, and lends it to students at 3.4%.   Administrative costs are well below 1%. Prepayment risk is minimal; repayment stretches over many, many years, and the yield spread just keeps on coming.  Interest rate risk is also minimal, given that Treasury can issue debt in a range of maturities. 

What about the credit losses, you ask?  While many loans go into default (about 10% projected for 2013 loans), credit losses are relatively modest.  The Education Department assumes it will collect between 75% and 80% of defaulted loans (on a discounted NPV basis), using its supercreditor powers, especially wage garnishment and tax refund intercepts. There is no statute of limitations on student loans, and even bankruptcy discharge is difficult. The $37 billion Treasury profit for FY2012 is after allowing for estimated credit losses in the $5 billion range.

Treasury even profits on the loans made by banks and guaranteed by the government, although bank loans are costlier and less efficient than direct loans.  Guaranteed student loans generate revenue because the guarantee fees paid by lenders to Uncle Sam greatly exceed losses net of recoveries.  In other words, if student loan interest rates were set on a break-even basis, they would be much lower, perhaps around 1.5%, albeit rising as Treasury rates rise.  Recently adopted loan forgiveness programs, and the expansion of income-based repayment, may increase the cost  (or reduce the profit) of the student loan program in the long run, but so far CBO scores those loan modification costs as minimal.

So what are the President and Congress arguing about?  They are arguing about how much of the federal deficit to plug with student loan interest money.  The current “baseline” budget assumes that the rate will jump up to 6.8% for 2013 loans, yielding another $30 to $40 billion return to Treasury.  Congressional Republicans see the expiration of the rate cut as a given, and any extension as increasing the deficit (by reducing student borrowers’ subsidy to all other federal programs). They are demanding offsetting cuts in other programs before agreeing to keep the rate at 3.4%.

Interest rate policy has huge consequences for the American middle class.  Charging an above-cost rate in order to fund Pell grants for low-income students, for example, is justifiable on a theory akin to equity financing of human capital.  It is a cross-subsidy from successful college grads to needy college students whose future success is uncertain.  A similar case can be made for expanded subsidy and loan forgiveness programs for college graduates in low-paying but socially necessary occupations.  On the other hand, the rationale to tax student loan borrowers to fund tax cuts for the wealthy, subsidies for energy and agriculture or other unrelated federal expenditures, is less clear.

Comments

[LSTB]

"How does the government profit from student loans? In two words, yield spread. Treasury can borrow money at 0.5% or less, and lends it to students at 3.4%."

Actually, the government is probably losing money on student loans.

Due to the Federal Credit Reform Act, the government is required to use "accrual accounting" when it values the returns on its loans, as you demonstrate in the quote above. While this is better than "cash accounting," which makes loans appear as large expenditures in the year they're issued with a corresponding smaller return in subsequent years, accrual accounting is inferior to "fair-value accounting."

As the CBO states [http://www.cbo.gov/ftpdocs/113xx/doc11343/03-15-Student_Loan_Letter.pdf]:
"Although the FCRA methodology accounts for the average losses from defaults on loans, it does not include the cost of all of the risks that loans and loan guarantees impose on taxpayers. In particular, it does not include the cost of market risk—the risk that losses from defaults will be higher during periods of market stress, when resources are scarce and hence most valuable. The cost of market risk is excluded from estimates under FCRA because the law dictates that expected future cash flows be discounted at Treasury borrowing rates rather than at the higher rates that private investors would require to make the loans or guarantees (page 3)."

Indeed, when the CBO did score the switch from the FFELP to the Direct Loan Program [http://www.cbo.gov/sites/default/files/cbofiles/ftpdocs/110xx/doc11043/03-25-studentloans.pdf] under fair-value measurements, it projected that the government would *lose* 12 cents on the dollar on average between 2010 and 2020. Under accrual accounting it gained 9 percent. (Table 3) I have no idea if this is still accurate.

On top of this, there's the question of what students are buying with this debt. As far as the government is concerned, student loans work in fiscal terms, but if graduates' degrees don't improve their productivity and if they don't otherwise have higher wages to pay off the debt, then student debt is like signing up for indentured servitude to the government. Sure, IBR helps, but that just means taxpayers just subsidized a lot of programs, administrators, dorm rooms, and climbing walls for little in return.

[Alan White]

LSTB, I'm fully aware of the FCRA/fair value debate. Clearly administrative costs have to be added into FCRA profit/subsidy numbers, and I mention that. I have a problem with the whole market risk concept, and the related liquidity premium concept, as a measure of cost for a federal government loan program. The argument is essentially that Treasury should calculate its lending costs as if it were a private lender. If the cash flow projections for student loan repayments already reflect reasonable default and recovery assumptions, the higher market price for student loans reflects only things the Treasury doesn't need to pay for: the higher cost and liquidity risk of funding student loans via securitization, and the higher cost of capital in the private market. Treasury has lower funding costs and basically no liquidity risk, and those savings are currently going to the General Fund, not to student loan borrowers.