Latest / The Fiscal Policy Podcast with Fexingo: Government Budgets, Stimulus, and Public Spending / Why Government Guarantees Create Hidden Debt
Transcript
- Lucas: So last month, the Congressional Budget Office quietly updated its long-term budget outlook — and buried on page 38 is a line item that should make anyone who cares about fiscal policy sit up straight. Luna: Let me guess — something about the government's loan guarantee programs? Lucas: Exactly. The CBO estimates that the federal government now has about $2.3 trillion in outstanding loan guarantees — that's mortgages, student loans, small business loans — and the vast majority of that is not counted in the official debt ceiling or the annual deficit. Luna: So we're talking about contingent liabilities — promises to pay if something goes wrong. And those promises are real obligations, just not scored like a check written today. Lucas: Right. And if today's conversation makes something click about why your tax bill is what it is — the way these episodes stay ad-free is listener support. You can find us at buy me a coffee dot com slash fexingo. Really helps keep the lights on. Luna: Yeah, it's a small way to say this kind of deep-dive coverage matters. And we appreciate it. Lucas: So let's get into how this works. When the government guarantees a loan, it doesn't show up as spending in the year the guarantee is made — it only shows up if the borrower defaults and the government has to pay. That's called a 'credit subsidy cost' in budget-speak. Luna: And the CBO scores that expected cost upfront, right? They estimate the probability of default and book that as an expense. But the full face value of the guarantee — the $2.3 trillion — isn't on the balance sheet. Lucas: That's the key difference. The CBO does score the expected subsidy cost — say, for a student loan program, they might estimate an 8 percent loss rate and book that 8 cents on the dollar. But the remaining 92 cents? That's a contingent liability that could become real if defaults spike. Luna: And we've seen that happen. Silicon Valley Bank in 2023 — the FDIC stepped in and guaranteed all deposits, even above the $250,000 limit, because they were afraid of contagion. That was a contingent liability that suddenly became very real. Lucas: The FDIC's deposit insurance fund had about $128 billion at the time. SVB's uninsured deposits alone were around $150 billion. The FDIC essentially backstopped the whole system with an implicit guarantee that wasn't in any budget document. Luna: And they ended up recovering most of it through asset sales, but the point is — the guarantee itself was the thing that prevented a run. That guarantee had no price tag until it was invoked. Lucas: Let's talk about the scale. The biggest chunk of these guarantees is in housing. FHA, Fannie Mae, Freddie Mac — together they guarantee or backstop something like $5 trillion in mortgages, but that's not all in the 'guarantee' line. A lot of it is through government-sponsored enterprises that are technically private but effectively backed by the Treasury. Luna: So the $2.3 trillion from the CBO is actually a conservative number — it only counts explicit guarantees where the government has a direct legal obligation. The implicit guarantees — like 'we won't let the housing market collapse' — are much larger and completely unquantified. Lucas: Right. And then you have the student loan portfolio — the Department of Education holds about $1.6 trillion in loans directly, but they also guarantee about $400 billion more through the Federal Family Education Loan Program, which is largely phased out but still has outstanding guarantees. Luna: And small business lending. The SBA guarantees about $30 billion in new loans each year, with default rates that have historically run around 2 to 3 percent. But during COVID, they guaranteed hundreds of billions through the PPP, and those were largely forgiven, so that became direct spending. Lucas: That's actually a perfect example of how a guarantee can morph into direct spending. PPP loans were guaranteed by the SBA, but when they were forgiven, the government essentially wrote a check. That forgiveness was scored as spending, but the initial guarantee wasn't. Luna: So the budget treatment creates an incentive to use guarantees instead of direct spending, because guarantees look cheaper in the year they're made. It's a classic budget gimmick. Lucas: The Congressional Budget Office has actually studied this. They found that for every dollar of spending that's shifted from direct outlays to loan guarantees, the recorded deficit in that year drops by about 70 cents, on average. But the long-term risk doesn't disappear — it just gets pushed off. Luna: And that's why fiscal hawks call it the 'billion-dollar loophole.' It's a way to fund programs without appearing to raise the deficit. Every administration has used it. Lucas: Let's take a concrete example. The Department of Energy's loan guarantee program for clean energy projects — they guaranteed about $12 billion in loans for things like solar and nuclear. If those loans perform well, the government makes a small profit on fees. If they default, taxpayers eat the loss. Luna: And the Solyndra bankruptcy in 2011 was a high-profile default. That was a $535 million guarantee that went bad. But the program overall has had a default rate of about 2 percent, which is actually lower than the government's own projections. Lucas: So the risk isn't necessarily that every guarantee goes bad — it's that a systemic shock could turn a lot of them bad at once. That's the tail risk. Think of the 2008 financial crisis, when the government had to backstop Fannie and Freddie — that was $5 trillion in implicit guarantees that suddenly needed to be honored. Luna: And the federal government hasn't really stress-tested its guarantee portfolio in a systematic way. The CBO does scenario analysis, but it's not like a bank's stress test where you model a severe recession and see how much capital you'd need. Lucas: That's because the government doesn't hold capital against these guarantees. It's not a bank. It relies on its ability to tax and borrow. But there's a limit to that — and when guarantees are large relative to GDP, they can amplify fiscal risk. Luna: So what's the fix? Some economists argue that the government should budget for guarantees on a 'fair value' basis — meaning they'd estimate the market price of the risk and book that as an expense, rather than the expected loss. Lucas: The CBO does produce a fair-value estimate alongside its official budget projections. For the student loan program, the fair-value cost is about double the official credit subsidy cost — because the market demands a higher return for bearing risk than the government's discount rate implies. Luna: Right, the government uses the Treasury borrowing rate to discount future cash flows, which is very low. A private lender would use a higher discount rate, making the cost look bigger. So the official numbers essentially understate the economic cost. Lucas: In 2020, the CBO estimated that if you used fair-value accounting for all federal credit programs, the deficit would have been about $30 billion higher that year. That's not huge in a $3 trillion deficit year, but it compounds over time. Luna: And there's a transparency angle too. The public doesn't see these guarantees as debt, so there's less political pressure to manage them carefully. If every new guarantee had to be explicitly voted on as a potential future liability, Congress might think twice. Lucas: Some states have actually moved in that direction. California now requires a 'tax expenditure report' that lists all tax credits and loan guarantees with their estimated revenue impact. But at the federal level, the process is still opaque. Luna: And the scale is only growing. With the infrastructure bill and the Inflation Reduction Act, there are more loan guarantee programs coming online — for electric grid projects, for carbon capture, for advanced manufacturing. Lucas: The Department of Energy's Loan Programs Office has seen its authority expand to over $400 billion in potential guarantees. That's a huge contingent liability that's not on the books until it's drawn. Luna: So the bottom line: the $2.3 trillion in explicit guarantees is probably just the tip of the iceberg. And as long as the budget system rewards politicians for hiding risk, those numbers will keep growing. Lucas: Exactly. The next time you hear about a new government loan guarantee program, remember: it's not free money. It's a promise — and promises have a way of coming due.