Listen to the podcast:
Most U.S. employers now have incentive programs designed to nudge workers toward healthier choices. But according to David Asch, who teaches both at Wharton and at Penn’s Perelman School of Medicine, most of those programs aren’t designed properly to deal with the human element. And perhaps it shouldn’t surprise us that, given how irrational people really are, systems based on the apparent logic of classical economics sometimes miss the mark in getting us to do the right thing.
But those problems can be fixed. Asch spoke with Knowledge@Wharton about his research on behavioral economics in health care, how it could help companies develop better health incentive programs for their employees, and why the solutions are both more and less simple than you might expect. His recent paper on this topic is titled, “Behavioral Economics Holds Potential to Deliver Better Results for Patients, Insurers, and Employers.”
You can listen to the interview using the player above. An edited transcript of the conversation appears below.
Knowledge@Wharton: Can you give us a brief overview of your research?
David Asch: For many American companies and American employers, even if they’re not in the health care business, health is central to what they do, because it keeps their workforce healthy — and because frankly, health care costs have an incredible impact on their bottom line. Everybody knows that health is determined so much by individual human behavior; we now have much better ways of motivating that behavior from principles of behavioral economics — principles that, frankly, employers have been a little slow to adopt.
Knowledge@Wharton: What are some of the key takeaways from your research in terms of how employers could use these principles to design better health incentive programs?
Asch: First, let’s back up a little bit and ask: “Well, what are they doing?” I think so much of our approach to health — whether it’s from a doctor approaching health care, or an employer or an HR department thinking about wellness for their employees — is based on the principle of education. We should just “educate employees” that smoking is bad for them and that fitness is good for them, and they should be adhering to their prescriptions. It’s a very rational model, but it assumes that if our employees only knew what to do, they would do that.
As soon as you say those words, you realize how limited that approach really is. Because of course people do know that smoking is bad for them, but it’s very hard to quit. Or they start smoking in the first place. They know they should take their medications. So giving them education rarely is going to be enough to move the needle, yet so many programs that employers use are fundamentally based on that.
The next level that employers use is thinking about economic incentives: “Why don’t we give a reduction in health insurance premium or some kind of pay for performance approach to get employees to move forward?” And there is some evidence that that works. But the work that we do has been largely focused on different kinds of arrangements to help motivate either employees on the employer side, or patients in the clinical setting, to try to get people to take better care of their health — recognizing the psychological foibles and pitfalls that we all fall into.
“We see losses as more potent than gains. So let’s see how we can frame things as losses.”
Knowledge@Wharton: So if I’m an employer and I’m trying to design an effective health incentive program, what are some conclusions from your research that might be a surprise to me as I’m sitting down trying to design this plan?
Asch: A lot of the results are surprising, and they’re surprising because they’re sort of irrational. In fact, they take advantage of the fact that many of us don’t think in rational ways. Here’s an example of a recent study that we did; it was led by a colleague of mine, Mitish Patel. He, our colleague Kevin Volpp and I — we did this with a group of employees to try to get them to walk more. We did a randomized controlled trial, real science, thinking, how can we get employees to walk at least 7,000 steps a day?
One group was a control group. They just were given feedback about how far they walked. Some groups were paid. In this case we paid them $1.40 a day for every day in which they walked 7,000 steps. A third group got what we would call a loss framed incentive — exactly the same amount of money, $1.40 a day, which is $42 a month. What we did is we gave them $42 a month in a virtual account before it started, and we took $1.40 back for every day they didn’t walk 7,000 steps.
Now mathematically, those two incentive arms are the same, right? It’s $1.40 gained if you walk, or the failure to lose $1.40 if you walk. The same thing. It turns out those who got the gain incentive – those who were promised $1.40 for walking — didn’t walk any more than the control group. The group that got the same $1.40 framed as a loss incentive walked 50% more than the control group. It’s irrational. It doesn’t make any sense. It’s the truth. Real data shows that. And it’s consistent with theories of behavioral economics, but inconsistent with how a lot of employers or doctors think about the best way to take care of their patients.
Knowledge@Wharton: What are some of the underlying forces at work here? If I’m doing any sort of incentive program or employee health program, how can I think about this using behavioral economics to design something that will actually get the result that I want, which is healthier employees? What are some of these irrational ways of thinking that would help me out? What is the best way to make people do the things that we all know we should do, but that we don’t necessarily actually do?
Asch: I think you just phrased it perfectly: How do we get people to do the things we all know we should do, but don’t do them anyway? As soon as you say that, you realize what strategies you have to use, because they’re not going to be the rational ones. Because you just gave a very rational argument: Of course I should do this. And the rational view is, well, then I’d do it. But of course, that’s not what really happens. So you have to move beyond the typical education-based solutions. I’m not against education; I’m a professor. I’m in the business of education. But it can be a bit of a decoy when something more effective might be in reach.
I think the answer to your question, the more direct answer to your question is to think, “What are the pitfalls we normally fall into?” One of them is what I just mentioned — we see losses as more potent than gains. So let’s see how we can frame things as losses. Another pitfall is that objects right in front of us or goals right in front of us are so much more salient, so much more meaningful than things far in the distance. This is why we don’t save enough for retirement. It’s why we eat that piece of chocolate cake right in front of us, even though we’re on a diet. It’s why we don’t take our blood pressure medicine even though we desperately want to avoid a stroke that might happen years later. All of these things reflect errors in which we focus too much on the present and not enough on the future.
How do we fix that? Well, can we accelerate the benefits? Can we create incentives today for taking my antihypertensive drug? Incentives today for not eating the chocolate cake? Incentives today for going to the gym? Not the incentive we would normally get, which is being a thinner, healthier person 10 years from now, or several months from now. Financial incentives are a way to do that. But there may be other incentives too — social incentives. If I set up a system to go to the gym with a buddy — we don’t need research to tell us this: People are more likely to go to the gym if they’ve made a little pact with a buddy to show up, because they feel guilty if they don’t show up. Those are social incentives, and those are very powerful also. Yet in health care, we often think of health care as so private that we don’t want to engage two people together with it. But we do that with so many other aspects of our life, and going to the gym or taking your medicines — it’s really not that private.
Knowledge@Wharton: One of the misperceptions that the research addressed that I thought was very interesting was the idea that, while education is obviously important and it’s obviously important to be transparent about what you’re offering someone and what you’re doing, sometimes it is possible to give people too much information, to throw so much at them that they actually miss what you’re actually trying to tell them or get them to do.
Asch: Absolutely. There is a fine line between “enough information” and “too much information.” I think the best example for that is the health insurance plans that employers give to their employees. It’s April right now. The University of Pennsylvania has its benefits open season…. And I have to select the best health care plan for me and my family. Well, at Penn I spent 15 years as the director of the Leonard Davis Institute of Health Economics. I have to think of myself as someone pretty knowledgeable about health insurance and health benefits, and I can’t figure out what the best health insurance plan is for me and my family. And the reason is, there are these complicated tables of deductibles and co-payments and out-of-pocket maximums and HSAs and all of these other kinds of things. Each one of those is a financial incentive designed to try to help me move my health care in a particular direction. But I would submit that a set of financial incentives you don’t understand is a set of financial incentives that can’t be effective. It’s just too complicated. I think we need to greatly simplify our approach, which is designed largely by economists and actuaries, and turn it into an approach that’s understandable by humans.
“Lots of our employees are smart and they can handle hard things. But why should we rely on that? We should be making it easy.”
Knowledge@Wharton: I feel like sometimes companies don’t do this because they feel like they’re going to insult people’s intelligence — like, of course you can figure this out. Of course you want all this information. But maybe people really don’t. They just want to know what’s best for them really in the end. That’s all they want to know.
Asch: Yes. I don’t think it would be insulting. I think I know a lot about health insurance, and yet it’s just complicated. And I’m busy. It’s not about cognitive demand or cognitive ability, it’s just more about the idea that we should be making it easy for people instead of relying on their ability to handle hard things. Lots of our employees are smart and they can handle hard things. But why should we rely on that? We should be making it easy.
So much of behavioral economics is about that. How do we make it so that it’s easier to walk? How do we make it so that it’s easier to take your antihypertensive medications to lower your blood pressure? Why do we have co-payments for antihypertensives? They’re not drugs of abuse. No one’s going to go and take lots of blood pressure medicine for the fun of it. We want people to take their blood pressure medicine. Why do we make it so hard for them to get that medication?
I think that once we take a step back, we can find lots of things that we’re doing that don’t have a good reason psychologically. We should be defaulting patients into automatic-refill plans so they don’t run out of their medications and fall off the wagon. There are lots of approaches. We just have to recognize: What are the steps required for certain goals? And, what are the psychological pitfalls that we all have as we approach those goals?
Knowledge@Wharton: Health incentive programs have entered the cultural lexicon. It seems like every company has them. And if they don’t, they’re probably going to have one next year or the year after that. So, are the companies being thoughtful enough when they’re adopting these programs, or are they being adopted in a vacuum without anyone thinking about the particulars of that company, of its employee population, or how to design a program that doesn’t just work for the short term to get the company what it wants, but actually makes for people who are healthier in the long term?
Asch: More than 80% of U.S. companies are using some form of financial incentive to promote the health of their employees and/or their spouses, dependents and the like. So it is widespread, and it’s probably going to grow more than that. What I think is interesting — and I don’t want to be too insulting, but I think that a lot of employers are not using these techniques as efficiently as they could. One of the reasons is, as I’ve alluded to, they’ve got too rational a perspective. The assumption is, “Well, I’m using economics to improve behavior, therefore I’m doing behavioral economics.” That’s not behavioral economics, that’s plain economics. Paying someone to do something is a highly transactional approach.
It becomes behavioral economics — and much more potent — when you use psychological tricks that require these little clever design elements. Actually, that’s not widely understood. Those little tricks take some experience, and it turns out that subtle differences — as I mentioned the framing of a reward as a gain or a loss, for example — a very subtle difference can have a huge impact. So unless you know what you’re doing, it’s actually hard to make use of behavioral economics in the workplace.
Knowledge@Wharton: What would you say sets this research apart from other research that’s being conducted on this topic?
Asch: You know, there isn’t a lot of work that is currently being done in behavioral economics in health that combines work in the wellness space — How do I get people to walk more or to lose weight? — with our insights from clinical medicine — How do I take someone who already has an illness, let’s say, congestive heart failure, and get them to stay out of the hospital? We acknowledge that we can learn something from both of those settings, and combine them.
“We don’t need research to tell us this: People are more likely to go to the gym if they’ve made a little pact with a buddy to show up, because they feel guilty if they don’t show up.”
If you’re an employer, for example, it’s easy and comfortable to think about wellness, but actually, wellness is not going to save you a lot of money in your employee group. That’s not where the money is. The money is in the very few employees that you have who have some serious illnesses. The ability to improve the health and health care of those employees is going to have a much larger effect on your bottom line than the important — but I would say highly diluted — effects of the wellness activities for your employees.
Knowledge@Wharton: What’s next for your research? Where are you going from here?
Asch: We’ve done a lot of work on the use of financial incentives to motivate people, and again, financial incentives that are designed with psychology in mind, not just transactionally. There’s been tremendous success in identifying ways to manipulate financial incentives to motivate employees. I think we’re increasingly interested in using social incentives. People sometimes freak out about financial incentives: Are you paying people to lose weight or to stop smoking? It can seem a little creepy or like a little bit of a “nanny state.” But actually, I think people are much more motivated by the social interactions that they have. They want to feel good among their colleagues. They don’t want to be a deadweight. They care what other people think about them. And to the extent that we can harness those very legitimate and very pro-social emotions and apply them to health settings, I think we can advance individual and population health forward a lot.
One of the barriers to that has been, as I mentioned, that people have often thought of health and health care as very, very private. But if you were to watch what people say about themselves on Twitter or on Facebook, you would realize in an instant that actually, people are quite open about their health. I think that privacy is important, but honestly, it’s probably not as important as many people think it is.