Become a Creator today!Start creating today - Share your story with the world!
Start for free
00:00:00
00:00:01
Insurance – The Largest Health Policy Study Ever | Harvard Researcher Joseph Newhouse
16 Plays2 months ago
Today's guest, Dr. Joseph Newhouse, is a longtime Harvard professor whose research has shaped U.S. health policy for decades. He is best known for leading the RAND Health Insurance Experiment, the largest study ever conducted on health insurance and cost-sharing.
In this episode, we explore the story behind the RAND experiment - how it came to be, the challenges it faced, and what its findings revealed about how patients respond to cost-sharing. Dr. Newhouse explains why more care doesn’t always mean better health, and why outcomes varied so sharply for low-income patients with chronic conditions. We also discuss how RAND’s lessons still influence today’s debates over Medicaid, Medicare, and employer insurance, and where he thinks the study’s findings hold true. Finally, Dr. Newhouse reflects on what he might change if the experiment were run today, including how Medicare and value-based insurance design could be studied in new ways.
Recommended
Transcript
Introduction to Healthcare Theory Podcast
00:00:00
Speaker
Welcome to the Healthcare Theory Podcast. I'm your host, Nikhil Reddy, and every week we interview the entrepreneurs and thought leaders behind the future of healthcare care to see what's gone wrong with our system and how we can fix it.
00:00:14
Speaker
Today we're joined by Dr. Joseph Newhouse, one of the most influential figures in health economics. Dr. Newhouse is a professor of health policy and health management at Harvard, and he spent decades shaping how we think about healthcare financing, Medicare, and insurance design.
00:00:30
Speaker
He's actually best known as a principal investigator of the RAND Health Insurance Experiment, which is the largest health policy study ever conducted in U.S. history, and it's fundamentally changed our understanding of how cost sharing influences healthcare use and outcomes.
The RAND Health Insurance Experiment: An Overview
00:00:44
Speaker
His work has guided health economists, policymakers, presidential administrations ranging from Nixon's to Obama's and continues to still have strong influence today.
00:00:56
Speaker
Hi, Dr. Newhouse. Thanks so much for coming on today. It's pleasure to be with you. Of course, it's great to have you. And I'd love to start by talking about the RAND health insurance experiment that you led in the 1970s.
00:01:10
Speaker
From what I know, I mean, it totally reshaped how we think about how cost sharing and deductibles, copayments can affect how we use health care. And before we get into the result results, can you share the story of how the experiment came to be?
00:01:24
Speaker
What questions did you set out to answer? And what was the climate like in the early that made a randomized health insurance study like this at this scale so necessary? How it came about is somewhat chance events. I mean, was a young researcher fresh out of graduate school, submitted ah research grant proposal, uh,
00:01:48
Speaker
an in the jargon of and NIH, to the government to study demand for care. I thought demand was, the empirical evidence was virtually non-existent in terms of designing an insurance policy.
00:02:05
Speaker
Most studies at that time put in a dummy variable for whether a person had insurance or not, ah which wasn't very helpful if you wanted to know what difference it made to have a higher or a lower deductible or a higher or lower coinsurance rate.
00:02:21
Speaker
um So I had gotten access to some data from actually surveys at the University of Chicago ah on people's stated use of services in response to an interview and some details on their insurance policy, although most people couldn't answer very much about the details on their insurance policy.
00:02:47
Speaker
Now, as it happened, this was 1969 or 1970 when I sent this proposal in. And as it happened, the Nixon administration had taken office in January of 1969. And Medicaid had been enacted states.
00:03:07
Speaker
as a program in nineteen sixty six at the option of states and Many states had taken it up by 1969, not all of them.
Political Climate and Experiment Genesis
00:03:18
Speaker
The traditional Republican ah position on cost-sharing was that people should pay something for care. And the traditional ah Democratic position had been that care should be free, or as they said, actually care was a right.
00:03:38
Speaker
although exactly what that meant in terms of care that people were entitled to was left undefined. So when the Nixon administration came into power, they confronted the question that Medicaid had been enacted with ah no coinsurance, no cost sharing.
00:03:56
Speaker
And they established a task force to determine if Medicaid should have cost sharing. And there was a an economist of but my cohort actually and from graduate school, who was at that time at the Office of Economic Opportunity, which was ah the office that managed the war on poverty programs that the Johnson administration had established.
00:04:27
Speaker
And he was seconded to this task force, which was over what is now DHHS, but what that time was called Health Education and Welfare. And He concluded that wasn't really anything known about this subject. What would cost sharing do if it were applied in Medicaid? but It was all ideology.
00:04:49
Speaker
And he, this office had been sponsoring ah at that point, a number of experiments in ah what was called income maintenance or welfare reform.
Design and Challenges of the RAND Experiment
00:05:04
Speaker
And the idea was, this was an idea that was a, ah espoused by both Milton Friedman and James Tobin, who was ah on the left, so a bipartisan kind of support.
00:05:17
Speaker
and The idea was you'd give people a certain guaranteed income and then you'd tax it away with earnings. And these experiments were testing the effect on labor supply of different guarantee levels and different tax rates. A colleague who had been secunded to this task force wondered if you could take the same experiment idea of an experiment and apply it in ah health insurance, particularly the Medicaid context.
00:05:42
Speaker
And he came out to see me. He'd he gotten a copy of this, my grant application, and he thought that was a good, what I was proposing was a good idea, but he should be even broader. I had not not even thought thought about the notion of running an experiment.
00:05:57
Speaker
ah For one thing, I would have been way too costly in terms of the agency I was applying to for a grant. But and so i as we talked about it, it seemed like a natural application to me. So he he said he would give me a small grant to design an experiment with i did over about the next six or nine months. And the design actually that was pretty close to what actually transpired.
00:06:26
Speaker
And then the Office of Economic Opportunity decided it would hold a competition to see who should run the experiment. And we won that competition. And then ah the next step in the saga was that nixon and Nixon was reelected in 1972, and And he decided he wanted to abolish the Office of Economic Opportunity because it was a Lyndon Johnson program, which he didn't agree with.
00:06:58
Speaker
ah But the research and demonstration office, which was sponsoring what I was doing, was shifted over to Department of Health, Health, Education, Welfare.
00:07:08
Speaker
So yeah then there was actually a an attack from the right wing of the political party Republican Party on that experiment that Labeled it, uh, the Nixon administration plot to introduce national health insurance, which was a little strange, but I wound up having to go explain to the secretary of health, education and welfare, what this was all about.
00:07:34
Speaker
So he decided the experiment could go ahead in one site, which was Dayton, Ohio. And then there was. would be a review of what had happened before it could ah be allowed to expand to what I had proposed was six sites and 2000 families.
00:07:56
Speaker
And
00:08:01
Speaker
that, so we went ahead in Dayton and, but by the time we had ah had a year's experience or so in Dayton, Nixon had resigned and ah Gerald Ford was president.
Key Findings on Cost-Sharing and Health Outcomes
00:08:15
Speaker
And he, so then that came up for review. And it was so controversial that it actually went to him. i mean, I wasn't present, but yeah he spent about 45 minutes with his staff on whether the RAND experiment should be allowed to go ahead. And he finally, this seems amazing to me that a president of the United States should spend 45 minutes of his time on this project. But he did. And he was convinced by his staff who back to to go ahead. And so that's and then then things evolved as you can read in my book, Free For All and lessons from the RAND experiment.
00:08:58
Speaker
ah So that was how the experiment came to be. It's actually very interesting. I think like when you hear about these studies in retrospect, like it often seems maybe like I would say simpler, but the actual process of it coming about isn't you necessarily think about, right? You think about the results because oftentimes like that's the implications that are still with us today. But um it makes a lot of sense though. of course, it's like a landmark, huge study. So I guess it makes have so many different things and hoops you to go through, but it's amazing. It finally got done because I mean, the findings are still being used and they're still being cited today, the nearly, I guess, decades later. And I find it really interesting. It showed that people, when there's cost sharing, they use fewer health services, but
00:09:40
Speaker
That doesn't necessarily mean health health outcomes change drastically, but I'd love if you could like walk through some of those key lessons from the health insurance experiment and how did it impact discussions of like health policy at the time? And and especially today too, like do you think it's still having an influence in how we think about health policy and healthcare care utilization to today?
00:10:01
Speaker
Well, I think it still has influence among health policy analysts, uh, and certainly still used by the Congressional Budget Office.
00:10:13
Speaker
Uh, obviously there's still a big element of ideology, uh, in, um, people's positions on health insurance as we see with the efforts to repeal the ACA and the debates over Medicaid now. So what, what is,
00:10:32
Speaker
ah I think the takeaway lessons for me was that I was, ah and certainly my, the physicians I was working with like Robert Brooke were taken aback by the fact that the free plan seemed to have no difference in health outcomes with one exception oh relative to the ah cost sharing plans, even though people were making ah but almost twice as many physician visits per year and using more services generally.
00:11:11
Speaker
So the, um,
00:11:15
Speaker
you could explain that, uh, as some serve as the services weren't effective, which didn't seem too plausible because we knew people were getting, for example, anti hypertensives that were lowering blood pressure and,
00:11:33
Speaker
Um, or that our measures of health status were, not, uh, good, which I didn't believe.
00:11:44
Speaker
And, um or the third explanation was that the, we were talking about marginal services, even on the cost sharing plans, people were getting two or three physician visits a year. So we were talking about the effect of one, one or two more physician visits and some more hospitalizations.
00:12:03
Speaker
And I hadn't really considered before that the possibility that some of the services on average were actually harmful.
00:12:14
Speaker
yeah Maybe I should have because there were certainly malpractice suits in those days, but they they weren't all that common. But I later worked on a study of medical error ah and in new york state in colorado and utah that showed disturbingly high incidence of medical error.
00:12:35
Speaker
It may be tens of thousands of people appeared to each year were killed by medical errors. um And that that was course the extreme end, but short of that, many people had prolonged hospital stays or became disabled from medical care.
Impact on Medicaid and Administrative Challenges
00:12:56
Speaker
um And, and we, there was became later it became clear that there was a disturbingly high fraction of misdiagnosis and um so it it became more plausible to me that at least at the time of the 70s when x-ray treatments were less efficacious than they are now that the negative effects of medical of additional care could outweigh the positive effects of additional care um and I mean most
00:13:32
Speaker
course, the average effect of medical care could still be quite high and the marginal effect could be quite low. But this, these data, of course, all that are almost half a century old. and So whether that is still true of medical care is an open question.
00:13:50
Speaker
um And this I, okay, so I said there was one exception to this. And the big exception was low income hypertensives or people with high blood pressure,
00:14:02
Speaker
um, where there were efficacious treatments and they had a meaningful, uh, reduction in mortality. And by meaningful, I mean, think around a likelihood of mortality in any given year was reduced probably on the order of 15%. So as I recall, yeah the immediate policy implication, think we're two, twofold. One is that it,
00:14:31
Speaker
David Wright- somewhat stopped in its tracks the movement to put cost sharing into medicaid, even though by the time it the experiment came out it. David Wright- The Reagan administration was in charge they they did put in cost sharing for higher income medicaid beneficiaries eligibility had been expanded in some states. David Wright- But the dominant.
00:15:01
Speaker
At that time, the dominant push in Medicaid for more cost-sharing, I think, was somewhat blunted. and And there there became some nominal cost-sharing for drugs, but it it was still ah pretty minor, I think, relative to what a lot of Republicans would have liked to have seen or thought they wanted to see.
00:15:24
Speaker
Now, on the other hand, many employers took started to increase cost sharing. I think they, ah particularly deductibles in my own view that got somewhat out of hand later, but I think that was because ah healthcare care costs were going up rapidly in the 1980s and um employers had ah traditionally shifted
00:15:59
Speaker
the increases from year to year to wage increases. And they were getting uncomfortable with, uh, giving, mean um lower wage increases. There was particularly high inflation at the beginning of the nineteen eighty s Um, so-called Volcker recession was induced to lower the inflation rate.
00:16:22
Speaker
Um, but, uh, that, um, The high inflation let employers ah real wages more easily than if inflation was low because inflation was 10%, they could give a 7% increase in the worker. Didn't feel as much as if there'd been a 3% inflation and a 0% increase um or they got less pushback. So.
00:16:52
Speaker
um But then deductibles now I think are um onerous for many people. The other thing that the experiment had done was it it tried to income relate the deductible. Well, it did income relate to it. It didn't try to, it did.
00:17:10
Speaker
um And it capped it. So there was a stop loss feature. And ah what I learned from there is it very hard administratively to um we income relate ah cost sharing.
00:17:28
Speaker
um And you actually see this with the ac Affordable Care Act. ah And because the income, it does try to in up use income to really affect cost sharing.
00:17:44
Speaker
But the ah the IRS, if I I'm going to report my 2025 income to the IRS next April 2026. And and if I'm going to use that to income relate cost sharing, it will be for 2027 because I will already have an insurance policy for 2026. So there's a two year lag income. And then that raises the possibility. Well, if I income related it and my income changes, which it does, um should that
00:18:20
Speaker
be factored in if it's the government made my plan too generous should it claw back money and the ACA actually said yes but put a cap on the amount it could claw back and similarly should it give me money after the fact ah if my income fell well that may have been too late to do any real good health wise so it um I had not really appreciated how difficult that
Complexities of Insurance Models
00:18:55
Speaker
would be. Maybe if I'd had more familiarity with the income tax code, ah I would have because it's exceedingly complicated, which is why people use accountants to prepare their income taxes.
00:19:11
Speaker
But was... ah i was relatively naive at that point. But that was certainly a lesson um I learned. And what do you do, for example, if the family, you say the family's income is some number of thousands of dollars, and then there's a divorce in the middle of the year? How do you, what do you do about their health insurance? And the husband and the wife aren't speaking with each other. And there's, ah so how are they going to communicate about who uses health services and how it applies to their income related
00:19:46
Speaker
deductible. um I mean, that's a problem also if you have a family deductible. But anyway, that's a lot of administrative complications. I was in effect running a small insurance company, ah which I have to say my graduate school training in economics had not really prepared me to do.
00:20:08
Speaker
But i I subcontracted the claims processing to a health and welfare plans administrator and So I learned a lot about claims processing as well, ah as well about surveys. I mean, I to say my graduate school in economics in those days did not really talk about how to do surveys.
00:20:30
Speaker
So anyway, that that was some color on things I learned from the RAND experiment.
Provider Payments and Healthcare Systems
00:20:40
Speaker
Yeah, i think that's really interesting. Of course, like it's such a multifaceted area of how we end up paying for healthcare care and the decisions that we make around that is always going to be changing. But I'd love to get into i mean the provider side of things too. It's not just consumers.
00:20:54
Speaker
The way we pay for providers and capitate those costs matters. I would love to hear, I mean, seeing as you studied in Germany and also you've seen the health systems in many different countries, what how is our...
00:21:06
Speaker
health system design providers different for other countries? And why does that difference matter?
00:21:14
Speaker
Well, the other countries vary, our and our payment systems vary by the recipient. But I suppose most people immediately think of physicians.
00:21:30
Speaker
And many of the if not most of the Other countries like us predominantly use a fee-for-service system.
00:21:41
Speaker
And I say that even though we have been moving toward capitating physician groups, the groups tend to reimburse individual physicians on a fee-for-service basis. they don't al Kaiser notably doesn't do that. They're on salary.
00:22:00
Speaker
But um sal the salaries are determined largely by what physicians in the fee for service system earned. So, um, uh, and the, I mean, not all systems use that. The British, um, mostly use capitation for general practitioners and their specialists are salaried, but the German system that you mentioned, uh, the French system, for example, both, both use fee for service.
00:22:33
Speaker
And in terms of that, like if so, of course, health care is I mean, it's very different in the UK and Canada versus Germany and France and even more different parts the United States. We have a kind of a more fragmented system with something uninsured, someone public insurance, someone private insurance. But in terms of the capitation for providers, like, of course, it's different. But what does it actually mean for like everyday individuals and how they receive health care, how they how their health care is priced? Like what is like the tangible effects of these differences in capitation across different countries?
00:23:03
Speaker
you First of all, ah Canada the and the UK are quite different in themselves. Canada's system is a fee-for-service system for physicians. and so the um So if you get into health economics, there's a very famous paper by but of my former colleagues, Tom McGuire and Mark Pauley, that says physicians are interested in improving the welfare of their patients, which I think is undisputed, but then um they are also interested in ah their earnings, um which is not
00:23:54
Speaker
at all rat a radical idea for economists, but the it is a Physicians according to their ethical standards are not supposed to consider that and ah Indeed they vary ah In how much they waited I mean some physicians give free care ah some physicians work in low-income areas where patients are uninsured ah or Have high Medicaid practices where Medicaid pays them less and
00:24:32
Speaker
um those physicians in this context would be considered more altruistic than others and the degree of altruism undoubtedly differs by specialty as well on average um so that um and to go to your question about international differences that
00:25:00
Speaker
difference in the degree of altruism may well um differ internationally as well although i have not seen any real efforts to assess that um and indeed our while our physicians earn high fees their their incomes are not so different from the German physicians, I believe. And in any event, our physicians' specialty mix differs radically from most other
Conclusion: Lasting Impact on Healthcare Policy
00:25:40
Speaker
countries. We have ah about 90% of our physicians are specialists, whereas the typical countries, closer to 50 or even less, oh specialists. And specialists, of course, earn ah more money virtually everywhere.
00:26:00
Speaker
And I think it's definitely interesting. i mean, of course, you've also had, and and beyond just providers, there's like a huge difference in terms of, I mean, even most other countries like Germany also have like an active health care health insurance system, which I know is the topic you've touched on before. But the US is also notable because both the private and public insurers that tend to kind of dictate a health care is paid for. They also are actively involved in how we receive health care.
00:26:23
Speaker
um Dr. Pauly actually was one of our press guests on this podcast too. So um it's great to see that you've worked with And I really appreciate you coming on the podcast today. It's been great hearing what how you've not only seen the physician side of behavior, but really the RAND Health Insurance Experiment and how that's totally changed how we view healthcare care today.
00:26:41
Speaker
Oh, it's been a pleasure talking to you.
00:30:55
Speaker
Thanks for listening to The Healthcare Theory. Every Tuesday, expect a new episode on the platform of your choice. You can find us on Spotify, Apple Music, YouTube, any streaming platform you can imagine.
00:31:07
Speaker
We'll also be posting more short-form educational content on Instagram and TikTok. And if you really want to learn more about what's gone wrong with healthcare care and how you can help, check out our blog at thehealthcaretheory.org. Repeat, thehealthcaretheory.org. Again, i appreciate you tuning in and I hope to see you again soon.