I got some very interesting responses to my post on Wednesday about the total amount of student loans outstanding — and I’ve learned a lot about the limitations of official statistics as a result.
The first thing to say is that the New York Fed figures I published for total student loans are not accurate — they understate the truth of the matter. The New York Fed is aware of this fact, and will revise its numbers for Q2 2011 in coming weeks; when it releases the Q3 numbers, they will reflect the new methodology and will be substantially higher than what we’re seeing right now.
And yes, the new numbers will show that student-loan debt exceeds credit-card debt.
But they won’t show student-loan debt at $1 trillion.
It turns out that the $1 trillion number is not new — it appeared in the lede of a NYT story back in April, which attributed the number to Mark Kantrowitz. Kantrowitz, who runs the websites FinAid.org and Fastweb.com. Kantrowitz, described by the Richmond Fed as “a leading resource on student financial aid”, has been pushing his own estimates of student-loan debt for a while; he supplied the data for the chart which accompanied the NYT story, and indeed published a very similar chart back in August 2010.
The USA Today story I criticized on Wednesday was interesting because it’s the first story to come up with the $1 trillion number which doesn’t attribute it to Kantrowitz. Instead, it attributed the number to the New York Fed — but the New York Fed has never published a figure for student loans anywhere near $1 trillion.
Kantrowitz claimed in an email to Reuters that “the US Department of Education has confirmed total federal education debt in NSLDS at over $800 billion”. NSLDS is the National Student Loan Data System, but the Department of Education doesn’t publish numbers for the total amount of debt in that system, and it’s unclear where or when this confirmation happened. Kantrowitz also pointed us to the numbers in the president’s budget for FY 2012. If you want to be daunted, go to this page and check it out: it’s absolutely enormous.
I got Nick Rizzo to try to pull out the line items for the cumulative balance of student loans outstanding. There are two main sources for federal student loans — the Federal Family Educational Loan Program, or FFELP, which is being phased out, and the Federal Direct Student Loan Program, or FDSLP. The FFELP has about $390 billion in total loans outstanding –$77 billion in Stafford loans, $81 billion in unsubsidized Stafford loans, $21 billion in PLUS loans, and $211 billion in consolidation loans. PLUS loans are actually loans to parents, rather than to students, but they still count as education debt.
The FDSLP is up to $220 billion in loans outstanding at this point — $58 billion in Stafford, $59 billion in unsubsidized Stafford, $20 billion in PLUS, and $83 billion in consolidation loans.
Add the two together, and you get to $610 billion. There are a few tiny other loan programs in there as well, but nothing which really moves the needle much past that point.
Still, this is just federal loans. These numbers don’t include private-sector student loans at all, and already they’re above the $550 billion that the Fed claimed was the total of all student loans outstanding in the country.
The Fed knew this — Kantrowitz is no shrinking violet when it comes to sharing his findings — and so they embarked on a quest to find out why their student-loan number was so small.
Now it turns out that the data feeding the NY Fed’s household debt and credit report is bought in — the Fed has a contract with Experian, the credit-report company. Experian takes a nationally representative random sample of the credit reports that it runs, breaks out the various different forms of debt in those reports (mortgage, home equity lines, auto loans, credit cards, student loans), and passes that information on to the NY Fed, which compiles the report by extrapolating that data to the nation as a whole. And although Experian has provided data going back to 1999, the New York Fed has only actually published this data series for less than two years — the first release came for the first quarter of 2010.
When it became obvious that the student-loan totals were too low, the NY Fed and Experian started looking at the data again. And eventually the culprit was found. There was a bucket of random obligations called “Miscellaneous”, which included things like utility bills, child support, and alimony. And it turns out that if you went burrowing in that miscellaneous debt, there was actually a pile of weirdly-categorized student loans in there.
When the NY Fed restates the Q2 figures, and from Q3 onwards, that pile of student loans will get included in the bigger student-loan figure, where it belongs. And the number will rise, probably by a couple of hundred billion dollars. Not enough to bring it to $1 trillion, but enough to make it bigger than total credit-card debt.
The problem is that when the new numbers are restated, the old numbers won’t be. There’s a lot of digging into formerly-miscellaneous line items involved here, and a lot of very old data which was never provided to the NY Fed which needs to be resuscitated and disaggregated. That’s a laborious process, and one which will cost a lot of money. The NY Fed wants to be able to publish the full series in an accurate manner, but it won’t be able to do that for a while, if ever.
For the time being, we’re just going to have a spike in the series at Q2 2011, when it stops being inaccurate and starts being accurate. We won’t be able to tell, for instance, how fast student-loan debt has been growing. Which is a bad thing.
And more generally, it’s probably suboptimal that the best public data series for student-loan debt comes from a sample of credit reports from Experian. Since the government owns the vast majority of student loans, why can’t it just publish accurate data for total federal student loans outstanding? That would certainly be easier than having to piece such things together from dozens of disparate line items in the annual budget. I suspect, too, that the government also has pretty good data on private student loans, as well. But putting a data series together is a big undertaking: you have to find the data going back quite a ways, and then you have to commit to updating it on a regular basis. I suppose that no one really cares enough to have made that happen, ever — with the result that no one really knows for sure just how much America owes in student loans.
But when the NY Fed releases its new numbers, those will probably be the best we’ve got. Even if they’re flawed.