Making a killing or getting fleeced? Differing accounting models muddy student loan debate
Pablo Martinez Monsivais, Associated Press
In a rare Washington compromise last week, Congress passed and President Obama signed a bill that temporarily lowered interest rates for student loan borrowers.
The rate each new borrower will get will change each year, sliding up and down based on U.S. Treasury bond prices. If the economy picks up, rates will climb.
The relief was welcomed by all sides, but some continue to argue rates should go even lower because the federal government now administers them. The debate is muddied by two dramatically different estimates from the same reliable source.
Using one model, the government will make $184 billion on education loans in the next 10 years. Using another, it will lose $95 billion.
It's a $279 billion accounting dispute. They can't both be right, and the dispute has enormous implications not just for student loans but for all kinds of federal loan programs, including home loans. One side cites enormous profits, the other enormous losses. Those stuck in the middle have few tools to navigate the divide.
“Instead of helping our students, the government is making a profit on student loans,” Sen. Elizabeth Warren, D-Mass., said in July. “That is wrong. It is morally wrong. That is obscene.”
Warren sponsored an amendment, rejected by the Senate, that would have lowered the cap for undergraduate loans to 6.8 percent. Not everyone agrees with Warren that the government is making a killing on student loans. Some very serious voices are arguing that taxpayers actually are getting fleeced.
Warren has a point. The Congressional Budget Office estimated in June that under existing law the federal government would have raked in a $184 billion profit from student loans over the next 10 years. The figure was widely circulated in the student loan debate.
Since the CBO has long been widely recognized as the most objective and competent number cruncher around, Warren's use of its numbers should hardly be controversial.
But the CBO report also offered another number that got less attention — a $95 billion loss. That's what CBO projects using an alternate standard known as "fair-value accounting." Unlike standard government accounting, FVA factors market risk as private investors would, with the same perspective on risk.
On the same page and in the same table that the CBO says the government would likely earn $184 billion profit it also says that the government may instead lose $95 billion on the same loans over the same time period.
When Warren cited CBO numbers to bolster her claim that rates should come down, Glenn Kessler, the Fact Checker at the Washington Post, called her out. He awarded Warren two Pinocchios because she failed to mention that the CBO itself paired her more optimistic numbers with totally different fair-value estimates.
As Kessler noted, it is quite clear that the CBO favors the more skeptical model, but he still hesitated to make the call, indicating he did not really feel qualified to referee the competing accounting standards. His two Pinocchios out of four seemed almost a gesture of futility, arms thrown in the air, splitting the difference for lack of real clarity.
"We obviously are not going to take a position on the correct accounting method," Kessler wrote, adding that "ordinarily a CBO projection is worth its weight in gold. But this is an unusual situation. The respected and nonpartisan budget agency clearly has concerns about the method mandated by Congress — and the differences over the next 11 years are stunning."
The CBO position
Kessler did make at least one mistake; he wrote that the alternate CBO numbers were requested by congressional Republicans. In fact, the report makes clear that the request was bipartisan, coming from the chairman and ranking member of the Senate Budget Committee.
The CBO report lays out the reasoning behind the fair-value estimates, focusing on the premiums private investors would be required to take on the risk of that loan. The CBO focuses particularly on the "fact that investors demand additional compensation to accept the risk that losses may exceed those already reflected in the estimates of cash flows and that those losses may occur when resources are scarce and particularly valuable."
In other words, fair-value accounting adds a premium for the possibility that debtors will default in large numbers at the very moment that the investor is most in need of liquid assets.
Using that standard, the CBO examined 38 direct loan and loan guarantee programs. It reported that for 33 of the 38, switching to the fair-value accounting model moves the program out of profit and into the loss column for 2013.
A thought experiment
Jason Delisle, director of the Federal Education Budget Project for the New America Foundation, illustrates the CBO approach with a thought experiment.
According to Delisle, "the government's magical ability to lower cost of capital" is purely a function of its ability to tax and print money to make good on losses.
The only proper way to measure the risk borne by the taxpayer-as-investor, he argues, is to set aside the taxpayer guarantee and see how the market would price the risk without it.
"Let's say the government is going to finance the cash flow of its entire student loan portfolio by issuing special treasury bonds backed only by the cash flow from the student loans," Delisle proposed. His trust would already factor in estimated defaults, but it would not have access to any more federal money. It would have to be self-funding.
The bonds, Delisle argued, would not trade at treasury rates. "Suddenly the bond buyers would say, 'Hey, now I care about the risk of those student loans.’ ”
The only way the bonds would trade at treasury rates, according to Delisle, is if the government put an unlimited guarantee on the bonds, promising that if the bonds failed the taxpayers would make up the difference.
A responsible lender?
Not everyone agrees. Among the resistors are those who fear that fair-value accounting will make books look bad and constrain the government from pursuing social policy through lending.
"It's an empirical question," said Jared Bernstein, a senior fellow at the Center on Budget on Policy Priorities. "You have to look at the history of government lending and loan guarantees and see if, in fact, this has been a burden on the taxpayer."
Bernstein believes the track record of government loans has been good, pointing to a 2012 report by John Griffith and Richard Caperton for the left-leaning Center for American Progress. The authors argued that the U.S. government has historically "proven to be a safe and responsible lender," often overestimating risks rather than understating them.
Griffith and Caperton found that over the past 20 years, government loans of all types only cost taxpayers 94 cents for every $100 loaned. They acknowledge the huge hit the housing crisis took on Federal Housing Administration loans, but argue that the FHA actually weathered that storm better than most of its private counterparts. And setting aside the housing crisis, they found that government lending "actually overestimated the total costs to government by $3 billion over the past 20 years."
"Government underpricing of risk is a convenient theory for free-market ideologues but it runs contrary to the overwhelming evidence," Griffith and Caperton wrote.
- The rise and fall of the University of Phoenix
- Former wrestlers charged, assistant coach...
- U. students release net neutrality-themed...
- 83-year-old volunteer a 'classroom...
- Universities seek alternatives to expensive...
- Mitt Romney to Jacksonville University...
- Why is BYU honoring Robby George, and who is he?
- Samuelson honored as 5,648 graduate from BYU
- Former wrestlers charged, assistant... 23
- Why is BYU honoring Robby George, and... 22
- Mitt Romney to Jacksonville University... 9
- Universities seek alternatives to... 6
- Samuelson honored as 5,648 graduate... 5
- Pittsburgh hires Scott Barnes as... 4
- Amy Choate-Nielsen: How one community... 2
- Why college matters more today than 20... 1