Image of Business Instructional Facility

Jan 12, 2026 Faculty Finance Research in Education

Ambulance Taxis: Curbing Medicare fraud without hurting patient care

By John Moist

When an ambulance makes three trips a week to transport a dialysis patient without medical necessity – with no need for onboard medical care or monitoring, for example – it’s acting less like an ambulance and more like a taxi funded by Medicare. Between 2003 and 2017, Medicare spent $7.7 billion on 37.5 million nonemergency ambulance rides for patients traveling to and from dialysis treatment. Many of those rides didn’t meet Medicare’s rules for reimbursement, which require that a patient have no other safe way to get to their appointment.

A new paper by Gies College of Business economist Riley League and coauthors Paul Eliason, Jetson Leder-Luis, Ryan C. McDevitt, and James W. Roberts poses a deceptively simple research question. What actually works to stop that kind of fraud? Should the government pay for the ride and pursue lawsuits afterward, or tighten the rules up front?

“We Could Answer a Pretty Cool Question”

League is an assistant professor of finance, an expert in studying waste and fraud in healthcare, and part of a research team that exposed growing antitrust concerns in the dialysis industry. As he recalls it, this project formed out of complementary expertise.

“This came together because each of the different people on the project brought a different angle,” said League. “It felt like the perfect opportunity to ask how approaches to fighting these fraud schemes work. We’ve got awesome data on this population, so we could hopefully answer a pretty cool question.”

In their Journal of Political Economy article “Ambulance Taxis: The Impact of Regulation and Litigation on Health-Care Fraud,” the group focused on a specific slice of Medicare recipients: people with end-stage renal disease who require dialysis, typically three times a week. Most patients travel to clinics on their own, but Medicare will pay for transit in an ambulance if a patient needs it. That distinction, it turns out, can be profitable to blur.

Between 2003 and 2014, nonemergency ambulance usage more than tripled, while the number of dialysis patients grew by 54%. In 2011 alone, ambulance rides to and from dialysis facilities made up $700 million in Medicare spending. But how does the fraud happen? As it turns out, it’s often a company affair.

“When we think about who is actually committing fraud in this setting, for our analysis, it’s a firm,” said League. “There are fraudulent ambulance companies that commit ambulance fraud, often very shoestring operations with one mastermind and a few people who drive the ambulances. What we’re really focused on is the fraudulent transportation of people with those ambulances, which in practice is an operation performed by a specific company.”

League and his coauthors compared two very different methods for combating fraud:

  • Pay-and-chase litigation: The government “pays” claims more or less automatically, then later investigates (“chases”) suspicious activity and sues or prosecutes them to recover funds.
  • Up-front regulations, including prior authorization: Providers of services must secure approval before delivering and billing for a service.

For years, Medicare relied on the “pay-and-chase” approach: investigations, litigation. Meanwhile, spending on dialysis ambulance rides kept rising. In 2014, Medicare changed course. It ran a pilot program requiring prior authorization for nonemergency rides to dialysis facilities in New Jersey, South Carolina, and Pennsylvania – three states with high rates of nonemergency ambulance claims.

Rates of nonemergency rides for dialysis patients fell sharply. In 2016, the program expanded to five more states and Washington, D.C. By 2022, prior authorization requirements stretched coast to coast. For League and his team, that staggered deployment made the policy rollout a natural experiment – a way to compare those two very different methods for fighting fraud.

“Good Shoe Leather Data Work”

To see the full picture, League and company compiled a unique data set: one that combined Medicare claims information from the United States Renal Data System with another set of old-fashioned research.

“The claims data let us see all of the ambulance rides these patients take, the other services they use, and their health outcomes,” said League. “The more unique data set is the one we collected by hand – a universe of these fraud cases against ambulance companies accused of transporting dialysis patients.”

That’s no small feat. The kind of granular information they needed wasn’t already available in a neat bundle.

“This involved a lot of reading press releases from the Department of Justice, combing the internet, talking to people at the DOJ, all of that good shoe leather data work,” League explained. “Then we basically compare places where lawsuits or prior authorization were put into place to where they weren’t. What happens to places where prior authorization goes into effect?”

It’s a textbook difference-in-difference analysis. By leveraging the staggered rollout of a policy change, the researchers could test the effectiveness of the policy itself.

In areas that shifted to prior authorization requirements, the numbers were definitive. Medicare spending on nonemergency dialysis ambulance rides dropped 68% immediately. Litigation didn’t have the same effect.

“We find prior authorization was extremely effective at reducing this healthcare fraud,” League said. “It’s about a two-thirds reduction in overall spending. A lot of these rides just go away and stop happening. That coincided with a lot of exit among firms. Many companies that were previously giving these ambulance rides stopped doing it altogether. It makes sense – if a firm’s business model was to provide a lot of nonemergency rides to dialysis patients fraudulently, when that revenue stream dries up they’re going to close up shop.”

But the market undergoes an adjustment. While some firms depart, the share of companies focused on dialysis rides grew, suggesting a movement toward providers who could meet the stricter rules.

“Importantly, there are patients who need this service, and firms that focus on that market segment after the regulation are not necessarily fraudulent,” cautioned League. “In fact, I think they’re quite likely not to be. Specializing around a regulation is something we see in other settings…we think a lot of the exit is because firms can’t get patients who should not be getting ambulance rides through the system.”

What Prior Authorization Means for Patients

A fall by two-thirds in nonemergency rides might raise a few eyebrows. Was the change a good deal for patients who simply need to make their appointment?

“When we first saw that prior authorization regulations dramatically reduced the rides that were given to these patients, we thought ‘oh well, we know what’s going to happen,’” League remembered. “This could be really good at reducing spending but have unintended consequences. So we checked. Were patients still able to get to their dialysis sessions? Indeed they were. We can rule out even a 1% change in the share of patients getting to their sessions.”

The downstream effects were also heartening. There were no effects on health outcomes such as hospitalizations or mortality.

“That was really surprising to us,” League said. “But it also validated that the reduction in utilization was, by Medicare’s rules, fraud. Firms pretty quickly internalize that they’re not going to be able to commit this fraud, and they stop trying. Some of that phenomenon where the fraudsters just sort of give up and stop trying to commit the fraud happens here, which lowers the cost of administering the regulatory apparatus.”

The researchers estimate that if Medicare had implemented prior authorization for rides back in 2003, at the outset of their data’s timeline, the program would have saved about $4.8 billion – without any harm to patients.

What makes the distinction so clear in this case between regulation by prior authorization and “pay-and-chase” litigation? League points to two key factors:

  • Limited Liability: Many firms committing fraud are small operations. For companies that can close shop or go bankrupt before litigation concludes, the threat of damages isn’t a meaningful deterrent.
  • Low Likelihood of Detection: With thousands of small providers and millions of claims, it’s hard – and costly – to find a fraudulent needle in a very big haystack.

“This is especially true for these small, fly-by-night-type firms,” League explained. “They may steal a few million dollars, but it’s not as if when you go after them civilly, they’re going to say ‘here’s the money back with triple damages.’ Instead they basically say ‘Sorry, the company doesn’t exist anymore,’ and there’s nothing to recover.”

That logic extends beyond dialysis to a wide range of policy debates. From COVID-19 relief to disaster aid, many programs rely on paying first and chasing fraud later, often with similarly disappointing outcomes. This research suggests that in some cases, carefully designed prepayment checks could be more useful – and come at lower costs – than aggressive litigation.

More research from Gies Business:

More Articles