Insurance Choices with Aaron Yelowitz: Part 2

In Part 2 of my conversation with Dr. Aaron Yelowitz, we move beyond asking who can best help consumers with complex health insurance choices and ask instead, “Why are they so darn complex in the first place?!”

We discuss how this complexity attempts to fight inherent market failures in insurance offerings, but confusion may also be a strategic financial move by insurers. Would standardization of plan choices help consumers? What other markets have faced the same pressures? How might this hold back consumer-focused innovation in the long run?

Dr. Aaron Yelowitz is an Economics professor University of Kentucky, a joint faculty member in the Martin School of Public Policy and Administration, a senior fellow with the Cato Institute, and a research fellow with the Institute of Labor Economics (IZA).

Be sure to check out Part I of our conversation, answering the notorious question: “Should we trust consumers or bureaucrats with insurance choices?”

A Conversation about Consumer Choice and Health Insurance: Part 2

Tina Marsh Dalton

I would love to know your opinion on this idea that’s been running in my head about insurance markets. Obviously, enrollees choosing plans that waste their money is bad. But I’ve wondered — why is our question, who’s the best to make this really complex, confusing choice? What if the question was, why is this choice so confusing? Why are plans so complex? I almost wrote a paper on this in grad school. Is it a way to stop adverse selection?

Aaron Yelowitz

It could be adverse selection, it could be moral hazard. Under the simplest theory of insurance, we would want to fully cover you against bad things happening to you — you pay a premium up front, no coinsurance, no deductibles, no maximum out-of-pockets. But if the bad thing happens to you, it’s like it never even happened — that’s Allstate’s motto. Of course, whenever something bad happens to you and we compensate you for it, then the moral hazard problem — the idea that you would go to the doctor more often for relatively minor things — comes into play. So the idea is maybe cost-sharing, skin in the game, puts discipline on consumers. It’s fundamentally the trade-off between: we don’t want really bad things to happen to you, but we also know that if we make something free or near-free, people will wastefully use it. But try explaining that to consumers. The design of those plans can attract different kinds of people — imagine that I’m old and sick, and you’re young and healthy. I might be attracted to a low-deductible plan. You might prefer greater protection too, but because you don’t go to the doctor that much, you’re fine with a high deductible. You’d still like the protection, but if the pool of people in the more generous plan is more adversely selected, you don’t want to hang out in that neighborhood if you’re a healthy person.

Tina Marsh Dalton

I view this from the insurer provider side — I want to have really complex plans, because that’s going to make it hard for the old person to know which plan to choose and hard for the young person too, and we’re going to have things smoothed out. I don’t see much policy discussion about, rather than helping people make the complex choices, how can we make the choices less complex?

Aaron Yelowitz

You know what’s interesting? Amy Finkelstein, of course, has a paper on everything, so it’s no surprise. Medigap plans used to be the Wild West. Prior to Medicare Part D — which provided prescription drug coverage, passed in 2003 and implemented in 2006 — drugs weren’t as big of a deal in the 1960s, but they became a much bigger deal over time. Part of the way seniors would get drug coverage, because Part A and Part B of Medicare didn’t cover it, was through what was called a Medigap plan. These still exist today. And it used to be there was almost zero standardization — so call it too much choice, but more just there was no easy way to comparison shop. At some point, the government came in and put some standardization on, here are the things a plan must have. And if that sounds familiar — when you signed mortgage documents at some point, there was surely a standard bureaucratic form you filled out. Or if you apply for a credit card, they’re required to show APRs, that kind of stuff. We see this kind of standardization pop up in all sorts of areas.

Tina Marsh Dalton

Those are both complex financial areas, right? And sometimes you have a lawyer helping you in the mortgage case, whereas you don’t have that in the healthcare case.

Aaron Yelowitz

We could ask, though — how much good does that do? When you put some things as prominent — you’re putting your thumb on the scale and saying the APR matters, or this matters, or that matters. Directionally, that’s probably true. Is it true for everyone all the time? The example I like to give is not a financial product, but imagine you were deciding which cereal to eat this morning. There is standardization that exists today that didn’t always exist — the side of the cereal box with calories per serving, broken out into macronutrients. Those are there so you can be a better shopper. But of course, you can’t adjust what a serving means, and we might be targeting the wrong thing. Those in the longevity space, for example, might say that the view of fat versus carbs in yesteryear versus today is vastly different. Imagine you put your thumb on the scale that carbs are okay and fat is bad, back in the day, and now perhaps we don’t think it’s nearly as obvious. So that’s exactly the idea — the bureaucrat comes in, standardizes something, we might think markets work a little more efficiently because of that. But you are putting your thumb on the scale saying what is important.

Tina Marsh Dalton

Yeah, it’s true. So we’re still stuck in the same dilemma of somebody’s making a choice. I think it has interesting distributional effects too, because we’re talking about this distribution of consumers, some of whom might be quite sophisticated. So those sophisticated consumers might be very well off in a system where they get to make very individualized choices. What they find in the Medicare literature is that poor plan choice happens more in the 75-to-80 age range, whereas 90-year-olds make great choices — because it’s like their kids or caregivers are making it for them. But then, when you standardize, you’re taking away the benefits for the top of the distribution, while you might be bringing up the bottom of the distribution.

Aaron Yelowitz

Right, and you alter behavior. Imagine we say carbs, protein, and fat matter, but other things don’t matter to the same degree. People today might say, for example, food dyes matter. You’re putting your thumb on the scale, saying these things matter and these things don’t, and at least for some people — as current conversation would suggest — there are other things that really do matter. And in some ways, it feels like it stifles innovation. Imagine you’re the cereal maker that says I’m not going to use food dyes, but that’s hardly rewarded in this context. Imagine you’re the professor who says, I want to do podcasts as part of my research dissemination. But if the profession rewards boring extensions in peer-reviewed journals rather than anything else, sometimes you get a disruptor through outside funding that can help move things, but not always. So the problem with standards is someone defined them somewhere — directionally, they’re probably not crazy, but could they stifle innovation along other margins that really are important, and perhaps margins we haven’t thought of?

Tina Marsh Dalton

That is really interesting when I’m thinking about how to make progress in this space. I think in the US, that’s especially important in healthcare, because we’ve hit this rough patch — how do you create a space for innovation? I’m thinking of mortgages now. I’ve had chats with my good friend who’s a realtor. His family actually has a plan that’s outside of the insurance space — it’s a religious-based plan, an international religious-based plan, and I’ve been fascinated by how he found it and how it’s existing outside of our pretty strict structure. How do we get innovative spaces while maintaining the safety net — the health insurance exchanges are still there, Medicaid is still there?

Aaron Yelowitz

I do think there’s a fundamental tension there. I’m a little bit aware of these collective plans where essentially a group of people, often incorporating religious values, get together. And one-size-fits-all, as an administration changes one way or another, has all the culture wars that exist today feeling like they are, to some extent, at the margin of healthcare decisions, autonomy, and fundamental values — do I have to pay for what you’re doing, and do you have to pay for what I’m doing? There’s a lot of that going on right now, and it’s not new to today — it’s been there all the time. And fundamentally, when I’ve quickly looked over these plans, what I’ve worried about is that they often have a maximum amount they can pay out. The thing that has concerned me is the undermining of coverage for some catastrophic event — the whole idea behind insurance is the small-probability catastrophic event. These plans are relatively small, and imagine they’re kind of under-financed in some ways. If you happen to be on a plan and had a $5 million expense, that is backbreaking for them, so they cap it off.

Tina Marsh Dalton

They need to get their size, yeah.

Aaron Yelowitz

Yeah, so you could imagine that this is just life — people make choices, and basically, do people want to buy plans that accord with their fundamental values, even if they potentially have existential risk to them? Free market people would say, that’s life, people will do that.

Tina Marsh Dalton

Well, even more — looking into this particular plan, it seems like it’s also an innovative, progressive dimension in that it seems to go back to the basic idea of insurance. In the US, it’s like these prepaid discount plans rather than what insurance should be at its core: there’s something big and unpredictable, and if it happened, you’d be wiped out, so I’d like to take care of just the risk part. That as an innovation — we’ve gotten off track, and getting back to this basic principle is how insurance started. I have some posts where I was reading Black Beauty to my children, right? And this is a horse’s life back in the 1800s, but there’s this one scene where the horse’s cabbie gets sick, and he’s talking about his trade organization, which would reimburse him while he is sick — this event where everyone pays in, and then when you get the bad draw, you get paid out.

Aaron Yelowitz

Right, and you could imagine also that in smaller groups, the adverse selection problem isn’t the same, because if I try to fraudulently use insurance and we’re in the same plan — these are my friends, my neighbors. There’s also some vetting. If you have a pastor who will sign off that you show up every week to church, that’s something — you’re able to get somewhere on time every week. The other thing is one should wonder about things that are correlated with signing up for that plan, like lifestyle behaviors that are actually advantageous toward not having big expenses. Imagine that those who sign up for such a plan abstain relatively from drinking and drugs and smoking. All of that will do so much more to move health expenses than whatever fancy plan you have. Lifestyle behavior has got to be, by far, the most important factor in explaining the variation in expenses across people. Diet, exercise, those behaviors — they’ll explain just a bunch of the variation in costs.

Tina Marsh Dalton

Yeah, so if we can move to one section of my paper — talking about the prevention problem. My paper is laying out some arguments that the government should be involved because there are fundamental problems. But one of the problems I’m laying out with insurance markets is that all these insurance markets we just talked about, in terms of choice, are focusing on a one-year or two-year horizon, even with inertia. They’re not focusing on the 40 years. Do we have ways that markets or governments can solve this prevention problem?

Aaron Yelowitz

That’s a fascinating issue. Among things that relate to what you said — the first is the way that we reimburse providers. When you’re sick and use healthcare, people get reimbursed. When you’ve done everything right so you don’t use healthcare, they don’t get reimbursed. A few years ago when I was being very vigilant about my health, I actually learned I had sleep apnea. Since it was at a time when I was in sort of peak health for my age range, it wasn’t because of some of the other comorbidities. It turns out that, sleeping on my back — big sleep apnea; sleeping on my side — not so much. But of course, they were all recommending CPAP for me rather than a behavioral change, because you get reimbursed from one and you don’t from another.

Tina Marsh Dalton

It’s the subscription economy. CPAP is like a recurring expense for your practice’s income.

Aaron Yelowitz

Exactly. And then the other pressure — this is a big issue — is not only how do our lifestyle behaviors in our middle ages then percolate over to when we’re old, but imagine that you stop taking care of yourself, and you just switch to a more generous plan. And along the same lines, we might imagine that medicine in the future gets more personalized. With gene editing and gene therapy, there are things we can just fix that we couldn’t fix before. Those are super expensive, but they have long-run benefits — a large upfront cost, with benefits that accrue over a long period of time. As an example, think about Sovaldi, which is a permanent cure for hepatitis C. It’s a 90-day regimen. Those who get hepatitis C tend to be politically unsympathetic because it’s associated with IV drug use. But imagine that everyone who had Hep C took the regimen at the same time — you’d get rid of Hep C, because the contagion is through needles. But here’s the deal: it costs $100,000. What would you do? You would enroll in the fancy plan that covers the $100,000 this year, and then enroll in the same plan without Sovaldi coverage next year, paying a massively lower premium. So this is emblematic of the issue that is percolating up — personalized medicine, breakthrough treatments with large upfront costs but long-lived benefits, and how does one pay for those, which will become more common because of gene editing and gene therapy.

Tina Marsh Dalton

I think one thing I’m always searching for is that the problem in the current system is, who benefits from good health? It’s no one, or it’s me, and I’m not the direct payer in my healthcare relationship. I pay the insurer, and the insurer pays the provider, but we don’t have any direct financial relationship.

Aaron Yelowitz

Yeah. I’m always a little surprised. A few years ago, I was giving a talk up near Indianapolis — Ball State was hosting one of their conferences — and I was talking about the healthcare system. This was literally the time when I was investing very heavily in eating right and exercising. Today, I would just buy my way out of some of those problems. It is more expensive to eat healthy than to eat badly. Exercise has parts of it that are slightly rewarding, but it’s a big time sink — real effort. And there can be side effects: if you walk too much, you can get plantar fasciitis, that kind of thing. But was my health insurance lower cost? No. Did even an unnamed employer offer free gym memberships? No. Which — if you want to get rid of frictions, then for employees who are there for life, directionally that has got to pass the cost-benefit test, because most people still won’t go to the gym even if it were completely free.

Tina Marsh Dalton

Well, employers maybe have some skin in the game too, right? If it shows you’re more productive — so in principle.

Aaron Yelowitz

They have skin in the game. And maybe this is just a bureaucrat who doesn’t want to shake things up, but it seems to me that every once in a while they’ll send you a free Fitbit or offer a $100 gift certificate to the person who gets the most steps, but it feels like it is massively not optimized for how lifestyle behavior really matters. And if you think about it — yes, insurance companies can switch, and they worry about you switching from one plan to another. Should your employer care about whether you smoke? The answer feels like yes, because many people are with their employers for long amounts of time. If you smoke today, it matters ten years from now, probably. There is this mystery of — many people get employer insurance, employers take it out of wages, and there is not much push for lifestyle incentives. Maybe it’s just because it’s hard to change these behaviors. A few years ago, I would joke about — well, if there was some magical pill that could just make people not obese. And we are getting there. But generally, the issue is lifestyle matters a lot, but it’s also super hard to change, especially in the long run.

Tina Marsh Dalton

Maybe we can bring it back to consumer choice and go full circle. The idea you’re describing is the employer should care, but the employer has to care over all of its thousands of employees, and the individual benefit of one employee not smoking isn’t going to come back to them in the same way as it would directly. So some idea about the individual in choice being important, even in creating this prevention system.

Aaron Yelowitz

Right, and that’s where the skin in the game would come in — that’s where consumer choice and health savings accounts come in. Imagine you’re allocated money, and if you make smart choices with respect to your health, you could use that to buy healthy food instead of paying some premium, or something like that. It feels like there could be more incentives in that direction.

Tina Marsh Dalton

Yeah. I do a bunch of analogies in the paper about the idea of auto insurance versus health insurance — insuring our car versus ourselves. What are the differences? And there is something about why isn’t there a good driver discount? If you brought individual choice into insurance, you can.

Aaron Yelowitz

Right. Often when I think about car insurance, back in the day — and this is how I try to relate it to my young adult students — there were good driver discounts. Do you have a GPA over 3.0? Insurance prices are very high at 16 and kind of fall off as you get a bit more experience. Today, the way I think of insurance companies is as seeking out information to give you as personalized a quote as possible. So the difference today versus back in the day is — for a 16-year-old, or an 18-year-old with a few years of driving, your app on your cell phone can show the time at which you’re driving. Driving after 11 PM just turns out to be more dangerous than driving in the afternoon. More truck drivers, et cetera. Hard stops versus decelerating quickly, driving over 80 miles an hour — these are things that apps now can track. What they’re trying to do is like a hotel with their list price rack rate. Everyone qualifies for that rate, but then they subtract a discount if you, A, opt in. Much like the supermarket card you scan every time you go shopping — you might say, why should I have to scan that to get a discount? Because what they’re getting is your information on whether you’re buying Frosted Flakes or Cheerios. Here, the idea is there’s some high list price, and if you choose to opt in and have them monitor — your phone can tell when you’re driving, or when you’re in a car — then they can tell whether you’re going over 80 miles an hour or driving late at night. And it turns out those things are probably far better predictors than just getting a speeding ticket. So they’re searching for that information. If they use it and others don’t, they’ll have a competitive advantage because they can keep the good customers.

Tina Marsh Dalton

Mm-hmm.

Aaron Yelowitz

So you could imagine — we often don’t talk about auto insurance as having the same crises in markets that we often talk about with health insurance. We often don’t talk about life insurance — something I’ve written quite a bit about — as having these kinds of issues. With life insurance, the reason why the market actually works pretty well is we price based on risk. If you purchase a policy, they come to your house, take your blood, do a blood panel on risk factors, ask you a handful of questions — like, have you ever been skydiving? Very few people have, but it would be a leading indicator of a taste for risk. And you sign the right of way for them to gather your healthcare information. Sleep apnea would matter, for instance, depending on whether it actually affects mortality from age 40 to 50. Some of these things might drive up health expenses, but if life insurance just cares about whether you die or not in those ten years, the degree to which it will matter isn’t clear. But those markets tend to work pretty well, and you don’t hear about the government wanting to take over the life insurance market. You sometimes hear this about car insurance, or after wildfires in the Palisades, about centralized home insurance. The problems with home insurance often come because they want to price risk inappropriately — think about hurricane insurance in Florida, the coast versus inland, or wildfires in hilly parts of Los Angeles with $5 million homes versus flat inland homes worth $500,000. The problem is cross-subsidies. That’s the government actually getting involved rather than letting the market work.

Tina Marsh Dalton

Well, I don’t know if we’re done with prevention, but I hope the innovators listening to this podcast have gotten some ideas about what we could do. I’m going to ask you a question I’d like to end with: hopefully this podcast moves the needle on some dimension of improving healthcare, but if we’re going to improve systems, we should improve our questions. So my last question for you is, what is the next better question we should be asking?

Aaron Yelowitz

I would say — in principle, what we care about is health: not only longer lives, but more vibrant lives. Yes, medicine can matter for that, and can be expensive. One does not want to dismiss that — imagine if there were a cure for sickle cell anemia, there are vaccines we should still be getting, there are breakthroughs that very well could be significant. By the time we’re the 90-year-olds you were talking about, who knows what treatments will look like and how less invasive and less painful they will be.

Tina Marsh Dalton

There are vaccines that we should still be getting. Yeah, there are breakthroughs that very well could be significant.

Aaron Yelowitz

But to me, I’m still pretty attached to the idea from a few years ago when I gave that keynote address at Ball State, which is: lifestyle behavior matters. Incentivizing people to make good choices goes a little bit against the crux of what insurance is, which is essentially that when bad things happen to you, we bail you out in some ways. In a world with Ozempic, or a world with statins, high cholesterol or obesity is less of a concern, because you can almost buy your way out of it with drugs. But I think the hard part is moving the needle with respect to these hard-to-move behaviors — the reason why people don’t want to eat broccoli and prefer to eat pizza is because pizza tastes good. The real question is, can there be innovations? Could there be a potato chip that’s actually healthy? Could there be food that people actually want to snack on that also does no harm? If people are clever enough to do gene therapy, isn’t there a way of creating food people actually want to eat rather than just pretending to like the healthy stuff? Innovations to make habits stick better. Could exercise be made more fun?

Tina Marsh Dalton

We’re innovating on the wrong side in insurance. What about the other side?

Aaron Yelowitz

Right — you could almost say vaping is a huge innovation relative to combustible cigarettes, conditional on whatever the verdict is on long-term health effects. It’s moving the needle in the other direction, at least. And so, are there better substitutes for pizza and for streaming Netflix? You’d run into the problem of the genius who comes up with that innovation — much like Steve Jobs with the iPhone, or the Google engineers who came up with the seminal paper that launched generative AI. You worry about whether you would capture enough of the returns to actually want to do it. But it seems to me my key takeaway is that’s the hard, hard problem. And I am more dismissive of the idea that, well, people love smoking and love taking drugs, so that’s just how it is. If you ask most people, do you wish you had just not gone in that direction? Many would say, I wish I hadn’t gone in that direction. I wish I had lived a lifestyle that was more robust in various ways. Somehow incentivizing that — I don’t have the solution, but I’d love there to be more attention to it. And to me, at least, that would also make a topic like Make America Healthy Again feel less controversial if it were actually focused on lifestyle behavior. What politician is going to say Americans shouldn’t eat 10% fewer calories and exercise 10% more? That is almost universally agreeable.

Tina Marsh Dalton

Well, and it also feels like flourishing — we could all get behind that. Could we innovate so that we would be creating better habits and being happier?

Aaron Yelowitz

Yeah, exactly. And there are also modern ills that come into this — doom scrolling and things like that, right? Things that maybe we don’t define as physical health, but probably do affect thriving in various ways. You run into the issue of personal freedom — if someone wants to be in the basement and play video games 24 hours a day, that’s between that person and their parents. But generally speaking, would you say, that’s what I wish I were doing? And obviously, I will push back against anyone saying, don’t tell me how to use my apps and social media. However, could there be some agreement that generally speaking, you wouldn’t wish it on your own children — so how could we present incentives to improve health broadly speaking, mental and physical?

Tina Marsh Dalton

I love it. Okay, I’m going to go search for some people to speak to that, or we’re going to put the call out right now to the innovators listening to this podcast to start weighing in.

Aaron Yelowitz

Yes, and each of us will collect just a very small royalty.

Tina Marsh Dalton

Yeah, the consulting fee.

Aaron Yelowitz

Yes, exactly.

Tina Marsh Dalton

Okay, well, thank you for coming today.

Aaron Yelowitz

My pleasure. Thank you so much for having me.