Latest / The Fiscal Policy Podcast with Fexingo: Government Budgets, Stimulus, and Public Spending / How the Federal Government Misses Billions in Improper Payments
Transcript
- Lucas: So last week, the Treasury released its annual report on improper payments for fiscal year 2025, and the headline number is staggering — the government made over two hundred billion dollars in payments that shouldn't have happened. Luna: Two hundred billion. That's more than what many countries spend on their entire military budget. How is that still happening? Lucas: That's exactly what I want to dig into. And before we go deeper, I want to pause for a second — if you're a regular listener and you've gotten something out of walking through the economy with us, a couple of dollars a month genuinely keeps these episodes going. You can find us at buy me a coffee dot com slash fexingo, and it makes a real difference. Luna: Yeah, it's a small gesture that keeps us ad-free and independent. Lucas: So, improper payments — that's a catch-all term for overpayments, underpayments, and payments made to the wrong person or for the wrong reason. The two hundred billion figure represents an error rate of about six percent across the government's total outlays. That's actually an improvement from a decade ago, when it was closer to ten percent, but it's still enormous in absolute terms. Luna: And where does most of that come from? I imagine healthcare is a big chunk. Lucas: You'd be right. Medicare and Medicaid together account for about seventy percent of all improper payments. In fiscal year 2025, Medicare alone had roughly ninety billion dollars in improper payments, and Medicaid wasn't far behind at about eighty billion. The rest is spread across unemployment insurance, earned income tax credit, and a bunch of other programs. Luna: So when you hear about 'Medicare fraud' in the news, is that what they're talking about? These improper payments? Lucas: Not exactly. Improper payments include fraud, but they also include simple errors — like a provider submitting a claim with a missing code, or the government paying twice for the same service. In fact, the majority of improper payments are what they call 'unable to determine' — meaning the paperwork wasn't sufficient to even tell if the payment was proper. Luna: That's kind of alarming. So the government doesn't even know how much it's overpaying because the documentation is so bad? Lucas: That's exactly the problem. And that's what the Payment Integrity Information Act was supposed to fix. It requires agencies to identify programs at high risk for improper payments, estimate the amount, and then take corrective action. But here's the thing — the Government Accountability Office has flagged that the estimates are often unreliable, and agencies frequently don't meet the reporting deadlines. Luna: So the law is on the books, but it's not being enforced. What's the actual consequence for an agency that fails to comply? Lucas: Technically, OMB is supposed to work with them to fix it, and the Inspector General can get involved. But there's no real financial penalty. The agency just has to submit a plan for improvement. And you can see the cycle repeating year after year — the same programs show up on the high-risk list, the same corrective actions are promised, and the numbers stay high. Luna: I remember reading about unemployment insurance during the pandemic — that was a mess. Did that contribute to the current numbers? Lucas: Huge. In fiscal year 2025, unemployment insurance had about fifteen billion in improper payments, but that's down from a peak of over a hundred billion in 2021 when the system was overwhelmed. The problem was that states were trying to push out claims as fast as possible, and fraudsters took advantage of the chaos. The Department of Labor's Office of Inspector General found that at least twenty billion dollars in pandemic-era unemployment payments were likely fraudulent. Luna: Twenty billion. And that's just one program. So what's the government actually doing about this? Is there any reform that's showing promise? Lucas: There are a few things. One is the Do Not Pay initiative, which is a centralized database that agencies are supposed to check before making payments. It's helped catch some things, but it's only as good as the data it draws from. Another is the recent push for more data analytics — using machine learning to flag unusual patterns in claims. And then there's the pandemic lessons learned, where some states completely revamped their identity verification processes. Luna: Can you give me an example of a state that did it well? Lucas: Sure. Idaho, for instance, implemented a system that cross-checks claims against multiple databases in real time. They cut their improper payment rate for unemployment insurance from about ten percent down to under three percent in a couple of years. That shows it's possible — but it takes investment and political will. Luna: So why don't more states do that? Is it a cost issue or something else? Lucas: It's a mix. Some states have outdated IT systems that are decades old, and upgrading them is expensive. But there's also a lack of urgency — the federal government covers the cost of the improper payments, so states don't feel the financial pain as directly. That's a classic principle-agent problem. Luna: That's a good way to put it. And it's not just states — the federal agencies themselves have the same incentive issue. If Medicare overpays, it's the taxpayer who eats the cost, not the agency. Lucas: Exactly. And that's why some reformers have floated the idea of a 'pay for integrity' model, where part of the savings from reduced improper payments goes back to the agency. That would align incentives. But it hasn't gained much traction in Congress. Luna: So we've got a structural problem that's going to be hard to fix. But is there any low-hanging fruit that could make a dent? Lucas: A few things. One is expanding the use of prior authorization and prepayment review in Medicare — catching claims before they're paid rather than trying to claw back money after the fact. The government currently recovers only about three percent of improper payments, so prevention is far more effective. Another is increasing funding for the Inspectors General — they're the ones doing the forensic audits, and they have a huge return on investment, often identifying dozens of dollars in savings for every dollar of their budget. Luna: That seems like an easy win. Why wouldn't you fund the people who are literally finding the waste? Lucas: You'd think. But it's a perennial battle. In the last budget round, the administration actually proposed cutting the OIG budget by about five percent. There's a political calculus there — some members of Congress see oversight as an attack on the programs they support, so they resist funding increases. Luna: So the politics get in the way of saving money. That sounds about right. Lucas: It does, and that's the frustrating part. The numbers are so large that even small improvements would save billions. But because the problem is decentralized and the incentives aren't aligned, it's easy to talk about and hard to act on. Luna: I keep thinking about that two hundred billion figure. If you could wave a magic wand and eliminate all improper payments, what would that do to the deficit? Lucas: It would cut the federal deficit by roughly ten percent. That's significant — it's about the size of the entire Department of Transportation budget. And it wouldn't require raising taxes or cutting programs, just spending the money correctly. That's why this is such a compelling issue for fiscal hawks. Luna: That's a good way to frame it. So do you think we'll ever get there? Or is this just an accepted cost of doing business? Lucas: I'm cautiously optimistic. The pandemic forced a lot of modernization, and there's now a new office at the Department of the Treasury focused on payment integrity and analytics. They're trying to create a whole of government approach, but it's still early days. The real test will be whether the next administration continues that push or lets it fade. Luna: So it's a matter of sustained attention, not a one-time fix. That's a good note to end on. Lucas: I think that's the takeaway. These are problems that have been built up over decades, and they won't be solved overnight. But every percentage point of improvement is billions of dollars that could be spent on something else — or not spent at all.