In response to the question Diane Auer Jones poses below about student loan reforms, my colleague Ben Miller (who has a frighteningly detailed understanding of the intricacies of federal loan policy) offers the following:

Because the piece has a lot going on, let's go through it one step at a time. Jones writes:

On one hand the Administration is trying to garner support for another jobs bill, yet on the other, it wants to eliminate an entire industry.

No matter how you run the numbers, the entire student loan industry will NOT be eliminated. In the doomsday scenario, about one-third of the 35,000 jobs--about 11,550--could be lost, but it's also possible that jobs could be gained. No one knows for sure. If there are losses, they are going to be in two areas: marketing and origination. Traditionally, loan origination means just moving the funds from the lender to the recipient. The student loan companies have twisted this a bit to include a series of amorphous functions that do not have a level of cost or effectiveness associated with them. No matter what, the meat and potatoes of student loan companies--functions associated with repayment--will not be lost.

In fact, Jones notes this two lines later:

The Department of Education is complaining about the horrible behavior of FFEL lenders, yet on the other, they will hire a few of them to be the contractual servicers of the loans issued by the Department under the DL program.

It's true, lenders in the bank-based system have done some bad things in the past. But this was the result of trying to market directly to schools in the hopes of obtaining their loan volume, a desire that stems from a subsidy model in which the way to maximize profits is to take on as many borrowers as possible. That's why companies were so willing to pay off financial aid officials or obtain illicit subsidies.

That's why the administration's loan plan essentially defangs the companies. When hired as servicers, the lenders are paid a small per-unit fee, one which decreases as you get more volume. Moreover, the Department of Education decides loan volume allocation and has greater oversight of performance, making it possible to take away loans from companies that are not doing a good job.

Here's where the fun really starts:

The interest rates charged to students and parents in the DL program are well above current market rates.

Stafford loans have an interest rate of 6.8 percent and interest accrues but does not compound in school. Subsidized rates are even lower. Loans taken out by parents or grad students are either 7.9 percent or 8.5 percent. By contrast, Sallie Mae's Smart Option private loan--a product that should reflect the market rate--carries an interest rate of LIBOR + 9.75 percent and you pay interest immediately.At Chase, the rates vary, but can get as high as 10 percent.

And all of this assumes that you could even get a loan in the private market. The credit crunch significantly tightened lending standards meaning that students without a co-signor or a poor credit score are pretty much out of luck.

But it continues:

Nobody would take a mortgage at 8 percent today, yet Congress and the President don't think twice about charging hard-working, middle class families 8 percent for the money they borrow to help their kids get an education.

So here's a quick recap on how a mortgage, car loan, or other similar types of products work. The bank agrees to lend you a certain amount of money for that product and it in turn is held as collateral. Thus, if you default, the bank can easily recover some of its losses by just repossessing the house, car, etc. This keeps rates lower because there is a physical good associated with the loan.

Student loans do not work that way. Students are too young to have collateral and a history of credit, so it is a risk to lend to them. That's why private market rates are higher than the ones determined by the government.

Now, could you get a loan with your house as collateral and then use it to pay for college? Absolutely. But that's not a student loan. It's a home equity loan and it's the same as if you used those funds to remodel your house. For most families, that probably is a better deal than getting a student loan. But if you don't have a house or have no collateral, that isn't an option.  Try walking into the bank and just asking for a $20,000 loan with no collateral. My guess is you'll end up with an interest rate well above 8 percent.

Continuing, Jones says:

I hope everyone understands that the savings are not primarily the result of the elimination of lender subsidies, but instead are a reflection of the amount of money the government will earn by collecting interest and fees from borrowers.

I'll throw this one over to Jason Delisle at the New America Foundation, who had a definitive post on this issue based upon a similar claim in Alexander's op-ed:

In fact, CBO’s market cost estimates show that students with direct loans get a subsidy on their loans from the federal government, not the other way around. That means under 100 percent direct lending the government won't be "overcharging borrowers," nor will the loans make money for the government. Rather, students will get loans at terms more generous than those offered in the private market, which results in a cost to the government. Thus the $47 billion is purely and simply the private market value of the government subsidies offered to private lenders under the FFEL program. A switch to direct loans eliminates those subsidies.

Jones continues:

It is also important to understand that the CBO score does not include administrative costs associated with issuing and servicing the significantly increased Direct Loan volume.

Actually, CBO has talked about this cost before and it's about $7 billion more over 10 years. You can do the math, but even subtracting that amount (as earlier estimates have done) still yields substantial savings.

And it continues:

According to the New America Foundation, while the first year cost of a $3000 FFEL loan is $88, the first year cost of a $3000 DL is $2995.  At a time when we are borrowing nearly 50 cents of every Federal dollar spent, it seems unwise to spend $2995 when, instead, we could spend $88, especially if we have to borrow part or all of the $2995 from China.  The CBO score does not take into account the cost of borrowing money to issue loans, which will ultimately reduce and possibly even eliminate the "savings" associated with the conversion to an all DL program.

Essentially, Jones is claiming that the first-year cash flow of a direct loan is higher than a FFEL loan. As a result, the government has to borrow more money upfront to pay for the loans, which increases the cost of borrowing and eats into savings.

But CBO has already done a score of the loan program in which it looked at what the cost would be if the government had to borrow at a higher rate. That's how it came up with the $47 billion savings figure that's referenced in the quote from Delisle's post above. So even if borrowing costs went up, it's still a better deal, first-year cash flow notwithstanding.

Talking about first-year loan cash flow also ignores the fact that the government is already funding most of the FFEL Program by buying existing loans from lenders. No one is complaining about the added cost of acquiring those loans. You know why? Because it's cheaper for the government to buy the entire loan upfront than it is to let it sit on the lender's books--even if they have to make a single upfront payment that could require borrowing. In fact, the alternative floated by the loan industry would still continue the need for a large lump payment, but it would actually make it more expensive than paying for the loan itself thanks to the addition of several fees for lenders.

Now, there is one part of Jones' post that is worth noting--her concerns about the Pell Grant program. The costs there are increasing at an incredibly high rate and will probably need to be addressed at some point, either by changing eligibility terms or by coming to a more honest understanding about how much money we are prepared to spend on this program.

But citing Pell cost concerns as a reason not to take $47 billion in subsidies away from middlemen lenders seems downright ridiculous. It's better to spend that money on covering some of Pell's costs than just letting it disappear into loan companies.