Bloomberg Audio Studios, Podcasts, Radio News.
Hello and welcome to another episode of The Odd Laws podcast. I'm Joe Wisenthal and I'm Tracy Alloway. Tracy, what do you think about DOGE.
The coin or the new Department of Government Efficiency the latter. Here's what I will say. Part of me hates that government efficiency is being politicized in this way, because if you think that government services are a desirable thing to have, then you should definitely be against waste and inefficiency and fraud in that market, because at a minimum, if you're wasting money, you could be using that money to do
even more. And then, obviously, if you think that government is just bad in general, then I imagine that you also think that the government wasting money is also bad.
There's no constituency there probably shouldn't be an ideological constituency or political constituency for waste, right.
Yeah, exactly, Like I feel like we should all agree on this, But also I hate the way it's kind of unrolling. Let's put it that way.
I would say I broadly agree waste is bad. Whether the existing makeup of this sort of quasi blue ribbon commission to get rid of waste whatever that means is actually going to do it. Look, I'll be open minded, but I like the premise of cracking down on waste. I will just say that, you know what the other thing is. So first of all, part of the issue here is that waste is probably difficult to define. Fraud
in some cases is probably difficult to define. There will probably be political fights over certain types of spending that it's like, you call this waste, I call this a good allocation of whatever. I imagine many of the political fights around this will sort of revolve around some of these questions. There's a lot there. But I'm glad we generally agree that fraud is bad.
Fraud is bad, waste is bad. But I think you're absolutely right that the definitions are going to be crucial, right, And this is where I worry about the politicization, because you can just come in and say like, oh, well, I don't like this, so I'm going to call this a waste and go on from there. But I think we should talk about it because you know, this has come up in a number of episodes, specifically on the PPP post pandemic, and so it's clearly something that is on people's minds totally.
Right now, we are recording this Thursday, November fourteenth. We are in a moment where resources in the economy are constrained. Right, the unemployment rate is low, inflation continue used to be by some measures above the Fed's goal. It is not crazy from like a macro standpoint to think we need to make things more efficient and better to have a better use of our real resources right now.
And one, this is your other middle aged man thing. You know, you've decided to become like a real resource constraint guy.
Right yeah, I'm like, oh, we got we got to crack down. You have to make tough decisions, to make tough decisions. We're taking the candy away from the kids. I'm going to be one of those.
Here's a tough decision. If you're going to talk to someone about government waste, would you talk to Elon or Vivek. There's two.
Well, Elon would get more downloads if we had a podcast, so I would talk to Elon. But we actually have a better guest. I think we have the perfect guest to talk about government waste fraud where we might be able to move the dial in a substantive way on this kind of stuff. We're going to be speaking with Jetson Leader Luis. He is an assistant professor at the
Questionroom School of Business at Boston University. He's also a faculty research fellow at the nb ARE and this is what he studies, particularly areas around fraud and how companies defraud the government wasting millions, probably billions of dollars. So, Jetson, thank you so much for coming.
On odd Lots, Thanks so much for having me.
Why do you describe your research? I kind of described it, but why did you describe what you do in your background?
Yeah? That's great. So I'm an assistant professor. I study economics, and in particular I'm interested in questions about fraud in government spending. I receive my PhD from MIT Economics in twenty twenty, and I've written a number of papers trying to explore the mechanisms the government can and does use to cut out fraud in it in public expenditure.
Was there a particular moment or reason that drew you to this particular field.
I think fraud in government spending has historically been underanalyzed, and in particular in the healthcare system, where I've done a lot of work. There are really big and impactful policies that are being used to try to eliminate fraud, but historically we didn't really understand what worked and what didn't work and why it worked, and so I think, you know, there was a big opportunity in the research there. And I've always been interested in questions of bad behavior.
I have a paper that I started early in my career on fraud in the World Bank that we just got accepted at the journal World Development. And so overall, I'm really interested in this question about, you know, where's the money going and how can we you know, fix that to make sure that the federal funds are being used for people who need them.
I definitely want to talk a little bit about fixing the problem and identifying the problem and so forth, But actually I really want to start on how to defraud the government because no, for real, because I sometimes we'll see headlines it's like so and so arrested for insurance fraud of some sort, and you know that's bad, But I also like have there's like weird thing that happens in my head where I'm like, I wouldn't even know how to defraud.
Tell us what's the most accessible way to defraud the government.
So, like, give us an example, or when we talk about, okay, fraud in the medical system, what are people doing? What is a classical form of defrauding to government in the world of healthcare.
So, in the world of healthcare, there are some really obvious frauds that have persisted for years and that I think we're finally maybe starting to wrap our hands around. And one that comes to mind immediately is the ambulance market.
Okay, say more so.
This was actually described to me when I first heard it by a colleague friend who works for the federal government as the perfect healthcare fraud. And so ambulance services are paid for by Medicare. Medicare is the old age health insurance program for Americans. We spend more than eight hundred billion dollars a year on this program. We spend another seven hundred billion dollars a year on Medicaid. So we're talking about one point five trillion dollars of outlays
to these programs. It's very hard for the government to ensure that every dollar that's going out is legitimate, right, It's a volume problem. Ambulance services are highly reimbursed and low overhead. If you want to start an ambulance company, you need to buy an ambulance. That's like thirty grand used online. You can actually go and google it yourself. You can go buy an ambulance, and then you need
a couple of employees. It's actually a super low overhead business, which means it's easy for people to start starting around. We think two thousand and three, the market for ambulance services, and in particular repetitive non emergency ambulance services, started getting saturated by intentional fraudulent actors.
Wait, what's a non emergency ambulance?
Non emergency ambulance is a patient who needs to go to a service because they are sick enough that they can't ride in a taxi or take the train, and they the only safe way for them to get to a service is in an ambulance. And in particular, this really blew up in the dialysis industry. Dialysis patients, there are about half a million of them. We actually spend I think you know this, one percent of the federal budget on the dialysis program.
Incredible, set not.
One percent of medicare. One percent of the federal budget is the dialysis program. We do not in general pay for ambulance rides or taxi rides for these people to go to and from the visits. They are responsible for getting themselves to the clinic every day, three times a week, generally for a few hours, and that's in perpetuity. It's very challenging to get a kidney and therefore to get off of diale. So we had this system, and this
is sort of the canonical Medicare fraud. We build in a little thing for the few people who need it, and that turns into a loophole through which bad actors drive a truck. So we built in this provision, which is if the only safe way that you can get to the dialysis clinic is in an ambulance, Medicare will pay for an ambulance, and they pay for it at a competitive rate for the ambulance companies, at say two
hundred and fifty dollars for a one way ride. Now that's not that much money for a real ambulance, but it's a heck of a lot of money for a taxi. And what happened is thousands of firms around the country opened with the express intention not of giving people serious medical care, but of becoming an expensive ambulance taxi and build the government. We have one hundred percent data from
the dialysis system. We can see all of these payments more than seven billion dollars for non emergency ambulance transportation over the following ten years, seven point seven billion dollars, and a lot of it was fraud, and the government cracked down. The government really tried to crack down, and in particular, the used a number of tools. The first one is they started throwing people in prison. This is
what you say, you see in the headlines. Yeah, but it's so easy to start an ambulance company that we'd see these stories where, you know, someone gets busted and their family member goes and opens a company next door the next day. And this persisted for years, with thousands of companies and billions of dollars of spending just down the drain.
I imagine some of the difficulty is also deciding who genuinely needs an ambulance ride and who doesn't. Right, So this is something that I never quite understand about US healthcare in general. The first ten years of my adulthood were in the UK, and there's a national health service there, and if the doctor told you you needed something. You know,
you got that something. Maybe it would take a while, but eventually you would get it, whereas in the US you seem to have all these decision makers in the process, and yet we're talking about medical care, which you would think would need to be dictated by like highly trained medical doctors.
This is super interesting that that's your intuition, because that's exactly how they fixed it. So the way that Medicare closed this loophole was by requiring what's called prior authorization. Now if instead of going to an ambulance company and saying, you know, please give me this ride, or the even worse, the ambulance company coming to the patient and saying do you want a taxi ride, which is actually what was happening, instead they required that a physicians signed off and say, look,
this patient is actually sick, they're bedridden. There's no other way that they can get to dialysis this week. And if you didn't have the doctor's note, Medicare didn't pay. And we estimate that around the timing of them implementing this prior authorization requirement, which they rolled out in different states at different times. This is like what economists love in our research and actual test a difference, and difference exactly when they rolled this out in different places, in
different times. We see a sixty seven percent drop in spending the next month, persistent, and not only that, we can then trace the patients and say were these patients harmed? Did they actually miss their dialysis visits and end up in the hospital, And we find no evidence at all of negative patient health effects, And so we actually saved billions of dollars by putting in something that was so basic, which is this the doctor sign?
Why didn't they do it before?
The structure of Medicare is largely disaggregated, where individuals are able to go to different services as long as they qualify for them, and some of those are doctor's visits, hospitals, medical equipment, pharmaceuticals. We pay for a lot of things, and there is a real worry that requiring too much paperwork can burden the system, and we don't want to turn the Medicare system into an even more heavily administrative
burden system. So there are different qualification rules, but largely what happens is the qualification rules are not always enforced upfront. We do a lot of requiring people to follow the rules, maybe we do some audits, maybe we chase after them with criminal lawsuits afterwards. But largely there are circumstances where nefarious actors, it's a relatively high trusystem. Nefarious actors will find these loopholes and drive a truck through them. And I can talk about a million examples of this.
Well, let's talk about so Okay, it sounds like the ambulance from taken care of more or less.
So the ambulance fraud is less than it used to be. And that's the dialysis ambulance fraud. Oh yeah, this paper was just accepted at the General Political Economy, So we're super happy. Shout out to my.
Congratulations who did a great job there.
The ambulance market itself has other frauds, right, we're talking about one type, which is this repetitive dialysis fraud. There are still lots of unnecessary ambulance rides, ghost ambulance rides where patients don't even get in the ambulance and bills are sent. One question is what can the government do? And part of it is I think that there needs to be a better focus on using data to detect and stop the fraud.
Yeah, talk to us about the data because I imagine, like you're talking about government spending and programs, plus in some instances the medical industry, there must be interesting data available to you.
So I have fantastic access to data. I use one hundred percent sample Medicare claims data from nineteen ninety nine through twenty nineteen for all inpatient and outpatient services, durbal medical equipments. I can see twenty percent of physician office visits and party pharmaceuticals. So it's an ridiculous volume. And you know, I'm very equipped with data. I teach data analysis and I have you know, PhD students and other professors who work with me. And even for us, it's
a big problem. How do we actually wrap our hands around this? Yeah, the government has not historically invested very well in its data analysis for anti fraud. Part of the reason is that the organizations that are responsible for this, which are the Department of Justice and the Office of the Inspector General. Those career lawyers are fantastic. I cannot say enough positive things about my colleagues at the Department of Justice and the Office of the Inspector General, but
there are too few of them. We do not pay them very well, and they are not data analysts. They are lawyers.
Okay, so you mentioned there's still some ambulance the dialysis specifically, sounds like that was mostly taken care of. There's other ambulance fraud out there. You mentioned that some people that there's billing ghost ambulances or people never even ride the ambulance. What's hot right now, what's the new ambulance fraud?
So what's amazing here is that it seems like it's a constantly evolving marketplace. Right We have to think fraud is a technology where people figure out a loophole and then they tell their friends and these things spread and eventually the government catches up, and so we're playing cat and mouse every year. Right now. I think it's wound care.
Okay, say more about wound care.
There's been a rise in these expensive treatments for patients with non healing wounds. So, if you're diabetic, you're likely to have neuropathy, and one of the things that comes with diabetic neuropathy is that you often have wounds, particularly on their feet, where the patient doesn't heal. There are some modern technology skin substitute things you can graft onto these wounds that seem like they pay pretty well and potentially even work, and then a few doctors have just
started spending millions of dollars on that. But that's a flash in the pan. Historically, we see fraud really rife in the durable medical equipment industry. There's been fraud in basically everything that healthcare touches. Another thing right now that's really popular. We're seeing a lot of fraud in vascular care that is helping patients who have collapsed vain be able to receive intravenous treatments. But again, it's just like just to.
Drill into specifics, let's just durable medical equipment fraud. Yes, I want to get into it. What am I doing so?
Or you want to get into durable medical equipment fraud, Well you're in luck because it's you know, durable medical equipment has been the wild West of the healthcare system for twenty years with billions of dollars of fraud. I have to admit I'm writing a paper on it right now. That's I'm excited. I know a lot about it. So you remember the scooter store they used to like advertise on late night TV. Are you an old person?
Number.
Would you like a free wheelchair? Yeah? So there's been some excellent kind of investigative journalism on this so durable medical equipment. I want people to think walkers, wheelchairs, oxygen pumps, hospital beds in their home. Things that people need that are supposed to be permanent. You know, seatpat machines, yep. And these are largely given by suppliers that are sometimes big national firms and sometimes small mom and pop shops. I want you to think about Flora. This is sort
of a Florida story. If you're interested in selling someone a you know, a fraudulent walker or pomp or something like that, it's actually.
Wait, what does it mean? A fraudulent walker one that doesn't work well.
That you don't need.
I guess this is super interesting, right, So what do we mean when we say fraud? Healthcare fraud has different types, right, and I can break it. There are three types of health care fraud. There's upcoding. That's where I sell you a little push wheelchair, but I go build a government for a super lux automatic wheelchair. We call that upcoding. There's medical necessity fraud. That's where we say that a patient needs something and they don't. And then there's substandard care.
That's where we have a patient who actually does need something and we give them junk. And all of them happen in all forms of medicine, but in particular and durable medical equipment. I think it's a lot of medical necessity fraud. I think we have patients who are getting a knock on the door, high do you want this fancy new device free to you? Now, what's really interesting is we're actually supposed to collect a twenty percent copay
for the durable medical equipment products. But the fraud and that's designed by medicare to make sure that patients aren't getting stuff that they don't need, so they.
Have to have some skin in the game.
They're supposed to do. But if you're a fraudulent firm, you just don't collect it. You're very happy to have the government's eighty percent, and the patient wouldn't take it if they had to pay.
Yeah, what's been the I'm trying to think how to frame this, but I guess what's been the cultural or incentive approach in government to stamping out fraud. If I am a government official and I design a poor social service program of some sort that has a bunch of loopholes that ends up costing lots of money. Do I get in trouble or do I get rewarded if I managed to tweak the program so that it doesn't have a lot of fraud in it.
The government generally underinvests and miss prices. It's anti fraud investments, by which I mean when we consider how we are measuring what the government is doing to stop fraud. You're talking about these career civil servants and how we reward them. Historically, the focus has been on how much are you getting back? And I've made this point in a bunch of research, and I recently released a white paper through the Center for Medicare and Medicare Medicaid Services saying how much money
you get back is irrelevant. That is the wrong number. That is really the number that everyone in government is focused on. When you say anti fraud recovery. We caught this many people, we put this many people in jail, and we got a billion dollars back this year. And the point that I've made is the money you're getting back is just a small share of the effect of your anti fraud efforts. What you should really care about
is your deterrence effect. And I've shown in research deterrence effects are in many cases like ten times larger than these recovery dollars. So you go to the Partner of Justice and you talk to their healthcare fraud people. They write a report to Congress every year. It's called the Healthcare Fraud and Abuse Report, and they put a number there for return on investment, and Congress asks them, tell us how much money did you spend and how what
was your return on investment? And they say the number is four, And okay, first of all, four ex return on investment already very good. That immediately means that we should spend more resources there. But I think that's actually the wrong number. I think the number is forty because the four is only counting money that they're getting written to them in terms of checks back. But if you count to terrence, and you have to count to terrence, the value of these anti fraud efforts is huge. So
do we reward people in terms of the value. They bring this to Tracy's question in some sense, yes, right, Look, there's a press release they trot out the attorney general, maybe the civil servant gets exact and.
A minimum I could go to my boss and be like, I saved US twenty million dollars.
But largely no, we don't pay these people. Well, we don't retain them. Well, if you look at your average assistant US attorney, they go to the government for a few years, do it fantastic work, and then realize that private industry pays three times as much.
In the league, when we were talking about the original dialysis ambulance fraud, it sounded like there was a very
elegant solution, just get a doctor sign off. When you look at some of these other emerging frauds, and you mentioned wound care for example, or you mentioned someone gets a really nice scooter that they didn't really need, or something like that, because and you mentioned that the requirement of putting up twenty percent copay is not that effective because the seller is just like, you know what, We'll take the hit and only take the eighty percent. Is
no big deal. How much of the challenge is on the data and identification side, which we should talk a little bit more, versus the versus the mechanism which you use to crack down on it.
So I'm actually really in favor of this program. We use the big whistleblower program called the False Claims Act. Okay, and I've written extensively. Do you guys know how this works?
No?
No, it's the weirdest, coolest thing. I mean for an economist. Okay, So there is a private market for anti fraud in the US. If you know about a company or person that is defrauding any public program, not specific to healthcare, you can hire your own attorney and there are firms that specialize in this. You can sue that person in federal civil court, and the whistleblower gets a share of
the money they bring back to the government. And this is a super effective program because it means that every nurse, every billing agent, every doctor in a hospital, if that organization.
This is what we're going to get into. Tracy, I'm not gonna do fraud because even though I want to know how it works, but I do want to make that.
I'm going to become a fraud bounty hunter.
It's bounty hunting. But bounty hunting is great because of two reasons. The first is there's this private information component. Rather than waiting for some analyst in Washington to figure out today's fraud, just make a huge reward available for the individuals who know about it, because there's lots of information. Right from an economist perspective, the problem here is the problem we have in a lot of healthcare. It's information asymmetry.
The doctors and the hospitals and the nursing homes know so much more about the patient than the insurance company. And here the insurance company is the government. And so rather than trying to make just top down solutions, and there are some top down solutions, don't get me wrong, it's really important that we also allow the individuals who have the information, the valuable information, to be rewarded for that.
So the whistleblower program private information. There's also the private cause of action, the idea that every person can become Literally the legal term sometimes use is private attorney general. Right, you can go and hire a lawyer and sue, and that lawsuit is on behalf of the United States of America. And this has been an extremely effective program historically.
Say more about that.
Yeah, So I ran a Freedom of Information Act request on the Department of Justice for data on every whistleblower lawsuit from nineteen eighty seven through about twenty seventeen. When I first filed this. There are thousands of these cases. They've brought in billions of dollars for the government. Fifty five percent of them are in healthcare, but they're also
used all over the government. We see cases related to the Transportation Department, the Department of Education, the Department of Defense, and the ideas the same. Instead of trying to have the government figure out how to run anti fraud, anti fraud can pay for itself. There are lots of people who would love to earn a million dollars being a whistleblower, and these whistleblowers get paid pretty well.
Since you mentioned transportation and education, just then, can you talk about some examples of fraud outside of the medical sector.
Absolutely. So. I have a recent paper about the unemployment insurance market during COVID, so I'd be happy to talk about that.
Oh, that's great.
So during COVID we had the biggest expansion of unemployment insurance in history. I think you guys know about this. You've talked a little bit about PPP, and there's some great research there on fraud. PPP was for companies. Unemployment insurance was for individuals who lost their job, and we expanded it to also include gig economy workers through a program called PUA. You guys probably know about this.
So this is starting to feel like history imade so crazy because this is like yesterday, but also it's it feels so long. Yeah, the time distortion is real.
So unemployment insurance rose heavily during the pandemic, and with it came a lot of fraud. Now why is there fraud in the unemployment insurance sector. Well, the government's cutting checks,
and the government's cutting checks really fast. So you might remember at the beginning of the pandemic, there was this immediate recession, and there was fear of the big macroeconomic consequences of everybody being out of a job on many people, I should say being out of a job, and so the government really loosened and expanded its use of unemployment insurance. But whenever the government says, hey, we're going to write eight hundred billion dollars in checks, people get creative about
ways to builk the government. And in particular, this type
of fraud was really identity theft. It is super easy, super easy to go on the dark web and buy a social Security number, and so that's what a lot of criminals and organized criminal groups did during the pandemic, there was a widespread fraud where individuals would go purchase identities, apply en mass to state unemployment insurance programs, collect them, take it out of the system, and then we think in many cases it was either offshore or rerouted to
criminal organizations. So this was so incredibly wides But actually my wife got a prepaid debit card in the mail from the unemployment agency and she didn't lose her job. I've talked to just dozens of people across the country and in all sorts of sectors, who said, yeah, this actually happened to me.
Wait, this leads to something that I want to ask as well, which is how is fraud propagated? Because this kind of gets to Joe's question earlier, how do people actually do this? How do I learn to do fraud? Is it like I just look it up on the internet, or is it like someone I'm associated with tells me.
So different frauds have different mechanisms by which people learn them, but in general, there is a social learning component to this. Absolutely. So with some of the institutional frauds that we've been talking about, a hospital that decides that they're going to suddenly charge a lot of money through some loophole. Often
there are business decisions being made by executives. Sometimes there are consultants involved, and big national chains often are the ones that's read these because they have kind of centralized management.
The hospital administrators from different hospitals look at this. I'm thinking, for example, of the tenant hospitals which paid nine hundred million dollars back to the government for a small little loophole that was supposed to be for an outlier payments program, and they just drove a truck through that loophole they ended up stealing. So if you go in patient in a hospital, the hospital gets paid fixed themount under Medicare through what's called a prospective payment system. They don't pay
per cost. They just say, you know you have a pneumonia, we're going to pay this much. The government was worried when they set this program up that some very expensive patients wouldn't get treatment because if the hospital knows that and they know that, they're going to lose money on you,
so they made this asterisk. A lot of these things are asterisks to have an outlier payment system where if a patient is super expensive, then they pay extra and the tenant hospitals figured out how to make every patient look super expensive by manipulating some of their balance sheets, and they ended up spending nine hundred million dollars to excuse me, they end up to settle claims, so that Tenant never admitted fault, I should say, but the government
received nine hundred million dollars back from Tenant because you know, these allegations were I think true that that Tenant had done this, and I estimate actually that the government lost billions of dollars to that. Okay, So but I want to go back to this question, So how did that one happen? Well, there was a consultancy in New Jersey that was going around telling people, hey, do you know
about this outlier payment system? And that's how we think that fraud spread, and it's spread all over the country. So there's some of this corporate learning from other companies, and there's also a lot of social learning. So in the case of unemployment insurance fraud or PPP fraud, you can go and you can find Telegram groups and Facebook groups of people that are like, here's how you apply
for a PPP loan. There's a great new paper by John Griffin at the University of Texas on these Facebook groups and like they're called like fraud kings, like they know what there were maybe PPP loan kings. There's something that it's like really obvious what they're doing and everyone knows.
And so in the case of health care fraud, in the case of non healthcare fraud, often it's you're surrounded by people who know how to do this, or you meet them through digital platforms, and then people learn and so a lot of the fraud we see propagates through communities.
In the case of the ambulance market, we saw that there were certain Eastern European groups that were responsible for this in different parts of the country and often from the same from original areas, and so you know, generally the understanding, at least from the government is that there must have been some social learning going out there.
It's very hard to prove, of course, right, someone figures it out and then tell family that's right, that makes sense. Let's talk about detection via data and more, and you talked about how you have access to all of this data and so forth. Obviously you can describe fraud qualitatively by saying this is how an ambulance company cheats the government,
et cetera. What do the fingerprints of fraud look like when you look at it on the macro scale, What pops up in the data that would at least be a yellow flag and say this is something we need to look at more.
So there are huge run ups in spending in every type of fraud I've ever seen, and every type of I mean, the whole point is if you're not making money, it's not a good fraud, right, And so if you just make a plot of bending against you know, time, you can often just see these really big exponential growth. Now, some of those are legitimate, because if there's a new great medical technology and people start using it, that also looks like a technological adoption curve, right of a kind
of big upper rank. But when it's fraud, it you know, first of all, it looks like massive year over year increases. The second is that it'll often be way too much, to the point where it's obvious that nobody's getting this. For example, sometimes you'll see a doctor who's just billing from too many home care visits, but like the homecare visits are sixty minutes, and the doctor's billing for five thousand of them a year. It's like, well, that's five
thousand hours. There aren't five thousand work hours in the year. The government should just be able to detect that. You should just not pay those So where the failure is is not in how hard it is to detect in data such and not that hard to detect in data. It's on the incentives for the enforcers to look at that data and use it appropriately. And that's where we get back to this limited enforcement capacity.
How do you measure benefits on the other side, because that's it seems again like a potential avenue where there could be some disagreement.
So I think it's super important that we preserve access to the public programs. And my research is not focused at all on, you know, how do we take things away from people? And in particular, I always try to measure very hard whether there are health effects associated with this. So in the case of the ambulances, we're able to show pretty definitively that there are no negative health effets associated with cutting this, and this is something I do
in all my papers. You really have to ask the question, were people losing care that they needed? And so super great question in the context of some of these very obvious frauds. Sometimes people aren't even getting the service. So if the government's paying for something and nobody ever got it, taking it away is costless. So that's the best efficient thing that we could do, is just stop paying for things that aren't even happening. Right, Let's not even talk
about waste. Let's just talk about these outright frauds. And so how do you measure it? I mean, if you look at very large scale claims data, as I do, and the government can, you can see, Okay, we cut out this provider, let's look at their patients. Did they go to the hospital more? That's an empirical question, right, And so there's no reason that it just has to be a gas, right. This is something that we should be measuring as part of our data analysis associated with handy front.
So it's very obvious why, whether we're talking about the government or private insurance, that it's just rules on top of rules on top of rules and asterisks and so forth, because this is very complicated stuff. Rules also create problems and compliance by the rules. Can you know strict adherence
to the rules can also have negative effects. I have to imagine, for example, that there are say, many people in the US currently on some sort of GLP one drug, but maybe technically they don't have the thing, but you know a lot of it. There seem to be a lot of benefits from weight loss and so forth. Have you found examples in your research in which there is some sort of positive externality from deviation from the rules?
Absolutely? Absolutely, Thanks for cueuing this up. I have a great story. So let's talk about the hospice industry. Hospice care is an end of life benefit for patients who have a prognosis of six months or less. So if you're dying and a physician certifies that you're dying within six months, you qualify for hospice. What is hospice? You give up the curative care, and you stop taking all of these meds with the horrible side effects, and you stop going to the hospital as often, and you can
die peacefully at home with pain medication. And I think that this is a great program. I think it's really important. We spend twenty billion dollars a year on the federal hospice program, and more than fifty percent of Medicare patients who die every year will have had a hospice claim, Okay, this is important. This is the way that we're treating people at the end of life. It's really hard to know who's dying within six months. That is not a
trivial estimate, and historically there was subjectivity in this. So over the twenty year period nineteen ninety nine through twenty nineteen, there was a quadrupling of for profit hospices, and many of them increasingly took Alzheimer's and dementia patients. Why because they stay for a long time on these hospice programs, and the hospice programs are paid about two hundred dollars
a day. And the federal government cried fraud, and we saw one hundred and sixty three federal whistleblower lawsuits against for profit hospice companies more than three hundred million dollars in settlements, saying this is fraud. You shouldn't have taken the patients. You should have known that they weren't dying
fast enough that they did not qualify for hospice. So I wrote a paper with John Gruber at MIT, as well as one of our PhD students and David Howard at Emery, some really top notch economists and we looked at this program. We said, Okay, this is something people are really concerned about. And what we found really knocked our socks off. The fraud actually was not as bad as people said, in particular because did some patients go to hospice who may otherwise not have because they were
not dying fast enough. Sure, but it's still a heck of a lot cheaper to go to hospice for six months than it is to go to the hospital and the nursing home and the homecare agency and the Durbal medical equipment and the pharmaceuticals. And we estimate that these patients saved tens of thousands of dollars. And the patients liked it. They and their families are picking hospice. This
is something that they want. They don't want the hospital, so they're getting a service they want and the government is saving money. And yet we've decided that this is a fraud because there's this rule. Now, when you think about it from that way, six months is totally arbitrary. Why is it six months and not eight months? And so this paper, it's called Dying or Lying, It just
was accepted last week at the AER. It shows pretty conclusively that the policies aimed at limiting fraud threw the baby out with the bathloar.
You know, we started this conversation mentioning DOJE the new Department of Government Efficiency. I guess one obvious question to ask you would be, you know, if you were in Elon or Vivek's shoes, what would be your you know, the first thing you would start with when it comes to rooting out fraud or I don't know if you want to get into wider types of inefficiencies, but what's the first thing you would do?
Yeah, they make you the real point person on this. How do you start?
So first, I'm optimistic about the Department of Government Efficiency because I think that there are some clear evidence based, obvious wins that we have left on the table in terms of saving government money. The first is, we know that anti fraud efforts, particularly in the Medicare and Medicaid programs,
have huge return on investments, and we underfund them. So we could staff up those offices at the Department of Justice, at the Department of Justice Districts, at the Office of the Inspector General and literally just chase down the leads that we already know about we don't even have to go after other leads. If you look at your average civil assistant US attorney, they've got thirty good cases on their desk and they get to pick two of them sore.
And many of these cases allege millions of dollars of fraud and they say, I'm sorry, we have a number. We can't go under anything right now less than ten million, and they drop the case. So even if we just doubled the number of people or at the Department of Justice who focus on fraud against the government, that would pay for itself many times over. Okay, So that's an obvious one. The second is to start getting serious about data.
If we do not hire and recruit excellent analysts to look at government data from the government, we are going to miss obvious frauds. Everything I've talked about so far is not rocket science. We're talking about huge amounts of billing for patients who obviously don't need it in some of these cases, and the government pays it and they miss it. They have access to that data in real time, they're writ in the checks. Why are we not screening that?
And the answer is not a lot of data analysts work at the Department of Justice, and there are some. There are some that work at the Office and the Inspector General of Health and Human Services, but not that many. Why because they pay like seventy thousand dollars a year. And if you're a good data analyst, you don't go to work for the government for seventy thousand dollars a year. If the government wants top talent, it's got to be
willing to recruit, and those are competitive positions. And so if we get serious about data, we get serious about machine learning, we get serious about investing in the attorneys who are doing the good work, and just the programs work really well. The second is rigorous evaluation, right, we need to know when we try policies, do they work
or do they not work. When we put in a prior authorization program, when we put in a new screening for people to take a selfie on their phone before we pay their unemployment insurance claim, that's actually how we fix the unemployment insurance problem. It was an identity theft problem. You just got to take a selfie on your phone. These ideas work, but we need to evaluate them as serious policy.
As I just have one more question, and I don't even know whether this is capable of being ascertained in a substantive way. But when you think about the stuff that you research and we're talking about, like, you know, big question is can we meaningfully move the dial on fraudulent spending or maybe wasteful spending? And do we have a number it exists? Do we know how much could potentially be saved? So?
I think healthcare fraud alone in the United States is something like one hundred billion dollars a year. Now that's across private and public. But we spend one point five trillion dollars on Medicare and Medicaid. So the idea that it's a big chunks, now, it's not that big of a chunk. We spent you know, two point three trillion dollars overall, and so to say fifty to one hundred billion dollars a year of fraud, yeah, I think that number is reasonable. Yeah, So can we move the needle
on it? Absolutely? There are things that we know work really well, huge treatment effects at very low cost. Now, government spending overall, we spend what six seven trillion dollars a year in the government. When we look at these other public programs there are some that have obvious frauds going on, Like I mentioned the unemployment insurance system that had tens or possibly one hundred billion dollars of fraud
there as well the PPP program. We spent one hundred billion dollars or at least on fraud in that program. And so not every program is rife with fraud. There are programs that are very hard to defraud. It's hard to defraud social Security why because they have your full earnings record and they pick the number and they send you a check. Very limited fraud in the social security system.
I think that there's probably a lot of fraud in some of the infrastructure and defense, and so our framework has to be like where does the government know the least, the government doesn't really know what's going into every element of a defense spend or every element of you know, how is the road being built or what is the hospital doing? And when you have those big information, these symmetries,
that's where the big frauds. So if I think that there's one hundred billion dollars just for Medicare and medicaid spending, or maybe you know, throw in the advantage spending and the you know, other the federal employees health benefits and the VA. To get to the hundred billion number of that healthcare fraud, I think that it very easily that hundred billion dollars also occurs. Yet again in some of these other like defense, I'm sure.
That there's plan Jensen leader luis amazing. Thank you so much for coming on them.
Park, Thanks for having me.
Tracy. I am pro getting rid of fraud. No, I've for real, I don't take a lot of it.
It seems like a low bar. Although you did ask repeatedly how you could commit it, but I know that was for informational purposes only.
No, I'm going to actually I'm not going to get into the business of medical fraud. I'm going to get into the business of being one of those independent whistleblowers to bounty hunters. Yeah, that sounds great, but I feel like I need to know how it works so I can identify it. But I really do think, you know, setting aside, setting aside everything fraud is bad.
I do think, and I kind of said this in the beginning, that we can all agree that fraud and waste is bad, like no matter where you fall on the political spectrum, because if you free up money that's not doing anything, then you could, in theory, put it to a different use and get more bang for your bucks,
so to speak. I think obviously, and we touched on this, like a lot of these decisions over what's fraudulent or especially what's wasteful maybe not necessarily fraud come down to specific judgments and they can be subjective, and I think that's where a lot of the disagreement is going to be going forward. But I do think the point about using the data better. The government must have some amazing data. We kind of talked about it, especially in these specific sectors like healthcare.
Two things on the data that were really interesting. So one is just this idea that if you're a talented data scientist.
Go to the doge.
And this has come up in some of our past episodes that we've done with a few other guests. There does seem to be this structural issue, right of how government pays and whether the how compelling a job in government is, et cetera. And so there does seem to be an issue with how do you staff up a big team of data scientists that are incentive and have the agency and capacity to use that data. Do something
when they discover it. And then you know, very interesting comment that last one about it's very hard to defraud social security, and so this idea of like where is the fraud most likely to exist? Areas in which there is some limited asymmetrical as economists like to use information, which is, you know, the government doesn't know what happens when you are not you and I go into a doctor, right,
there's some level of information asymmetry there. It doesn't really know what kind of walker you or I are going to need in thirty or forty years to get around, et cetera. And then furthermore, we didn't touch on it, but I do think or absolutely as a podcast, are going to need to do more, should do way more on defense spending. There's a million angles that we have
to do on that. But you could see like we don't really know what went into the assembly of this and the cost of this program and is this part really worth this much money or maybe as a competition thing. There's a lot of fruit there for future episodes.
But I do think one of the big tensions here is the sort of government generalists versus like the experts that they're listening to what I mean is, for instance, if you're a defense contractor and you're building I don't know, like a submarine launch pad or whatever, like the government official isn't necessarily going to know all the nuts and bolts that need to go into that. And so yeah, I always wonder how you sort of overcome that informational gap.
There's a lot there. Let's do more on this topic.
All right, shall we leave it there for now?
Let's leave it there.
This has been another episode of the All Thoughts podcast. I'm Tracy Alloway. You can follow me at Tracy Alloway.
And I'm Joe Wisenthal. You can follow me at the Stalwart. Follow our guest Jetson leader Luis. He's at jetson Econ. Also, he has a number of papers on his website that you can just click on and go read. They're all really fascinating. Follow our producers Carmen Rodriguez at Carman Armann dash Ol Bennett at Dashbot and kill Brooks at Kilbrooks. Thank you to our producer Moses Onam. For more Odd
Laws content, go to Bloomberg dot com slash oddlog. We have transcripts, a blog, and a newsletter, and you can chet about all of these topics in our discord with fellow listeners Discord, dot gg, slash up, lots.
And if you enjoy all lots, if you like it when we dive deep into the data around government fraud, then please leave us a positive review on your favorite podcast platform. And remember, if you are a Bloomberg subscriber, in addition to getting our new daily newsletter, you can listen to all of our episodes absolutely ad free. All you need to do is find the Bloomberg channel on Apple Podcasts and follow the instructions there. Thanks for listening. Eight