• Why I’m skeptical that HSAs will work to reduce spending greatly – Part 2

    Earlier, I posted some slides on why I’m skeptical that health savings accounts and cost-sharing won’t work like some hope because those who spend by far the most won’t be influenced by them. I’d like to take some time exploring those people who are high spenders. Remember – the top 5% of spenders account for about half of all personal health spending. Who are they?

    Who likes HSA’s the most? Young, healthy people (check out Figure 4). They’re in the left column. Look at the middle column, though. Americans below the age of 45 make up only 23.5% of the big spenders. Moreover, 38.2% of the big spenders are on Medicare – and none of this reallyapplies to them anyway.

    As you can see in the left column, the bottom 50% of spenders (who account for less than 3% of all personal health spending) are primarily in good, very good, or excellent health. Only about 5% of them are in poor or fair health. Compare that to the top 5% of spenders. More than 43% of them are in poor or fair health. They’re sick. They not only won’t spend less, I’m not even sure they should spend less. They’re not in good health!

    One more chart:

    If we look at the bottom 95% of spenders (who account for about half of all personal health spending), about half of them have no functional limitations or chronic conditions. They’re totally healthy. Only 7.4% of the top 5% of spenders can claim the same. Almost all of the big spenders have a chronic condition, have a chronic condition and a functional limitation, or have a chronic condition and need help with activities of daily life. They are sick, and need help. And they will (and likely should) spend money to get it.

    Health savings accounts, and increased cost sharing, will likely work for those who don’t need the care and are healthy. But they spend so little, that even if they are influenced, we won’t save much. Those who are spending the real money are older, sicker, and need help. If they spend less, it’s more likely they will suffer. And it’s likely many won’t spend less, since there’s no chance of their keeping money in their accounts. If they don’t spend less, then we can’t save real money.

    I get why people think cost sharing and HSAs will work. But when the theoretical possibilities meet the real demographics and patterns of spending, I can’t see how this will work.

    • Another very good post. That satisfies my curiosity about the physical characteristics and habits of the big-spenders.

      But as you said in your reply to John Goodman in the comment section of Part 1, a quarter of the people move from the Top 10% to the bottom 75% from year to year. I agree with your point that 3/4’s of the top spenders remaining from year to year is a fair amount of persistence, and so I don’t think HSAs are going to make a big impact on those permanent big-spenders, particularly as it pertains to the health care units they consume for their chronic condition(s).

      But I suppose I just don’t know enough to say that the $18 Billion less in spending from everyone else is so small that we can be assured that it would put no downward pressure on prices. Also when you consider that there is still some significant amount of give and take between the categories of big-spenders and everyone else, and when you consider that we haven’t itemized which items/services the big-spenders are spending the most for, I’m still left with my common sense heuristic that insuring routine and frequent medical care of the the vast majority people raises prices. I mean, the prices people are paying are paying right now for routine care are prices that, according to the proponents of HSA’s, were set without competition/shopping around.

    • You have to consider the actuarial costs of high deductible health insurance, too. A health plan with a 10K deductible isn’t that much cheaper than one with a 5K deductible because once you go over 5K, you are probably pretty sick and will blow through 10K, so the insurance company doesn’t get a big savings. The individual faces a substantial cost differential if they come down with something expensive — at a time when they probably have other cares, like home help, loss of a job, etc. So high deductible insurance tends to be a bad deal for consumers.

      For my insurance, $1,500 was the deductible where I thought the savings from a lower policy price versus the probable out of pocket cost got me the best deal. But my family often goes over $1,500.

    • Thanks for this clear and insightful analysis.
      HSA and cost-sharing are based on a false assumption that health spending is discretionary.
      In fact, there is very little discretionary health spending. You get sick, you need to spend. You are healthy, you don’t spend. It’s not like deciding to go out for a fancy dinner when you could just open a can of spaghetti at home.
      Once you are sick, most of your spending is out of your control. It is dictated by doctors, insurance, etc. You generally don’t have a choice to choose the “cheap” treatment or the “deluxe” treatment. Doctors tell you what to spend… (and it is usually the “deluxe” treatment).

      • Mark Spohr,

        I think part of the point of HSAs is to take insurance out of many health care decisions. I think your argument applies well to people with serious/chronic conditions. I mean, you get sick, you’re going to get care – whatever care they tell you is needed. But even with routine care, if insurance is picking up the tab, and better yet subsidizing most people’s care, then of course we’ll pay for “deluxe” treatment.

        The author of the post admits that HSAs reduce spending of relatively healthy populations. Presumably these are people that do to the doctor more than never. People in the big-spending category have seemingly endless spending habits, but presumably don’t need to spend endlessly on every kind of health care.

        Not saying they’re a silver bullet, but HSAs can be a part of an effort to treat these different things differently.

        • I think that you also need to remember that our growth in health care costs is largely being driven by procedures. These are expensive and go past any deductibles people are likely to have. HSAs large enough to cover the cost of a CABG or most chemotherapy seem likely to be out of the reach of most people. Also, this would require a huge cultural shift. In over 40 years in medicine in one form or another, I have never heard a patient newly diagnosed with a major disease ask where they can go to get the cheapest care. Basic logistics and familiarity with caregivers/facility predominate.


        • Due to the high cost of just about any medical care today, HSAs would only cover a few office visits and tests. Even a “routine” ER visit can easily deplete an HSA. So HSAs only come into play for people who have minor, self limited problems. They may keep a few people with sniffles away from the doctor (but at the cost of missing strep or a more serious problem). People might also skip important preventive measures (immunizations, HBP screening, etc.) so I really don’t think it’s a good idea to discourage going to the doctor.
          Going to the doctor also has significant costs (time, travel, leave from work, etc.) beyond the doctor fee so I don’t think people look at this as fun discretionary spending like buying a new pair of shoes. People go to the doctor because they are concerned about their health and they need a doctors care. It is not discretionary spending and it should not be discouraged. If you keep people away from the doctor, they will suffer health consequences. The only way you could argue that there are no consequences for skipping medical care is to argue that doctors don’t do any thing of value for patients. There may be some procedures for which this is true but patients don’t have enough information to make this decision with a doctors advice.

          The other issue is patients shopping for “cheaper” care. I just think this is laughable. Patients don’t want the cheapest care; they want the best quality care. Due to the extreme opacity of medical billing practices, it is also almost impossible to find out what most medical care will cost so patients can’t get this information even if they wanted to find the cheapest care.

          I also find this statement odd:
          “People in the big-spending category have seemingly endless spending habits, but presumably don’t need to spend endlessly on every kind of health care.”
          Are you comparing sick people with complex medical problems to compulsive shoppers? What “habit” of their would you like to curtail?

          • Mark,

            I’ve acknowledged that people with chronic/serious medical conditions shouldn’t be expected or asked to spend less. By using the word “habit” I’m only speaking to habits – nothing more, nothing less.

          • I really think there’s nothing to this point you’re making about “cheap” versus high quality. I mean, is it really your view that people are not sensitive to price at all when it comes to medical care, to any kind of medical care? I find that extreme. I’ve tried to acknowledge the places where I believe people wouldn’t price shop, conceding a lot of ground, I thought. But to hear that health care consumers don’t care about price at all?

            And, we’re too far apart on the fiscal/economics benefits of preventative care. I understand that your view is pretty intuitive, but it’s not the only intuitive view if you think about it; a lot of evidence so far shows that preventative care costs more.

            I think I should probably stop posting on this thread and just point out that the contention of those in favor of high costs is that they are caused (at least in part) because there is not enough competition/shopping around. As for opaque pricing, we’re simply talking about HSAs. I’ve acknowledged – I suppose not painstakingly enough – that HSAs are not a silver bullet.

            The context of these recent posts on HSAa was open enough, it seemed, to discuss what role HSAs could play in a variety of health care regimes. I realize that in the American context, the right sometimes seems to view HSAs as a panacea, so I (again perhaps not demonstrably enough) acknowledged that HSAs are no cure all, while trying to carve out a space where I could motivate the view that the right has some wisdom to offer. It didn’t work. Maybe you could read this post:


            Now, I’m not sure Singapore has it exactly right, and I’m not sure whether their system should be more vindicating for the right or left, but it has HSAs, and it eliminates the opaque pricing that concerns us. My immediate instinct is that the single payer catastrophic should not be optional. But anyway, we should be arguing on this level, not on some PolySci 101 level.

            • And by “we” I mean you and me (and probably SAO) as I think this blog in general is adding to the discussion.

        • Another problem with HSAs is that it assumes the patient can make informed decisions about what is the right treatment. I’m sure that most of the waste that can be found on review by professionals looks very different when you are a worried parent or patient in an ER or doctor’s office.

          I’ve never been offered ‘deluxe’ treatment. I’ve been told by doctors what they think the best thing to do is. If the problem is minor, like an earache, then the cost of the best treatment isn’t that great. If the problem is major, like cancer, I want the best chance of beating it. So, I can’t see many occasions when I would make a health care choice that would be a cheaper, second best option, even if I was spending my own money, unless I ran out of money, which is what I want my insurance to protect me against.

          • I imagine there are a lot of places in between cancer and an ear ache. In other words, I don’t imagine that the only two categories of health costs for consumer-patients are for things that are so low cost that shopping around is unmotivated on the one hand, and physical conditions (like cancer) so serious that price falls out the window, on the other. If we disagree on that, there’s probably too great a divide to make much headway with one another.

    • Aaron,

      Why are you jumping through all these theoretical hoops when this question has been studied over and over again by many reputable investigators? Most recently, the RAND Corporation concluded that HSA plans reduce premiums by about 30 percent and produce no adverse health affairs.

      (Please remove above comment)

      • There is no doubt that when we study relatively healthy populations (as the RAND HIE did) that we see reduced spending and relatively no bad health effects. We’ve blogged on that many, many times. The problem is that the vast majority of health care spending is not done by these populations, and so it is difficult to generalize its findings. Spending is not in a normal distribution, where those in the employed population represent how money is spent.

        The study which you are talking about (which I have read) is not being ignored here. Everything I’ve said does not deny that for many, many people in the US, HSAs and increased cost-sharing might have an effect. I think that because of how spending is distributed, those who constitute the vast majority of spending will not be effected as much, and that might make all the difference.

        • A very interesting and informative series of posts, thank you Aaron. This doesn’t mean that HSA’s can’t play a role in controlling health care costs for healthy people, but I agree that they are probably not the silver bullet that their advocates believe them to be.

      • This is what has always bothered me about this study, and why I dont think it is very useful.

        “They go on to acknowledge other limitations of this research. For instance, this study looked only at one year before enrollment and one year after.”

        In my corporation of physicians and advanced practice nurses, I find that it generally takes them about 6 months to a year to figure out how to effectively use a new plan when we change insurers.


    • 1. Won’t save much money – compared to what, exactly?

      If the demand for health care in particular cohort is inelastic, and they need X units of care – then only way to spend less is to reduce the price of X.

      It’s not clear how plans that incorporate cost sharing and embed incentives to seek out care only when necessary will be score any worse on this metric than comprehensive policies that have no mechanisms to encourage prudent use of care.

      Even in patients that have a chronic condition, it seems entirely possible to couple VBD with deductibles so that a diabetic pays nothing for a diabetes related illness but pays a significant copay if they go to the ER in a level 1 trauma center to get a splinter removed instead of going to an urgent care center.

      2. If your premise is correct, then there would be an opening in the market for insurers who can underwrite zero deductible, zero cost sharing plans and sell them for just about the same price as high-deductible plans, no?

      Ditto for enterprises that self-insrue. If the incentives embedded in the plans have zero effect on total spending in the covered population because all of the spending is driven by the sickest cohort – then large companies with many thousands of covered employees should have concluded that high-deductible plans don’t actually save them any money and either reverted back to zero-dollar plans or have done the math and decided not to go down that path. Any idea why that’s not the case?

    • Each year, we do not not know which participants will be in the high-use category, but we do know a certain percentage of them will be.
      What would be cost-effective is to divide the entire group into 2 cohorts – those who have expenses up to $25,000 in a year, and those with $25,000 or more of expenses.
      Actually, 2 different insurers would underwrite each cohort.
      I and 3 of my partners are designing a plan just like that after our 2 successful meetings with Milliman, a well-respected actuarial firm.
      Those that remain relatively healthy, can have a paid-up policy within 2-4 years, worth $25,000 -$50,000,,depending on contributions and claims.
      In a separate group are the high users, those with claims of $25,000-$50,000 and above.
      Everyone pays for this group of users, healthy and sick alike.
      And, they pay at community-rated [premiums, which are discounted about 60-80% from a traditional plan.
      Don Levit

    • No one here has mentioned prices, but that appears to be a big element in why US health care spending per capita is so high.

      As Ezra Klein reports, “the International Federation of Health Plans — a global insurance trade association that includes more than 100 insurers in 25 countries — … surveyed its members on the prices paid for 23 medical services and products in different countries, asking after everything from a routine doctor’s visit to a dose of Lipitor to coronary bypass surgery. And in 22 of 23 cases, Americans are paying higher prices than residents of other developed countries. Usually, we’re paying quite a bit more.” So you’ve got a sizable cohort of elderly, very sick people who require a lot of care that is priced higher than anywhere else. HSA’s can’t touch that.

    • Americans in general are timid about confronting doctors or hospital administrators. Americans also want to save every life that can be saved.

      The countries which have some cost control over health care are either:

      a. like Canada, which had left-wing govenments that were not timid and in fact took on a doctor’s strike when they banned balance billing as part of their Medicare:


      b. like Germany and France, which had powerful civil services and, after World War II, so much poverty that there really was no alternative to socialized medicine.


      c. countries that have been through wars and famines, and have had to let extremely sick people die.

      I bring this up because the high-deductible debate seems to follow a clasic American pattern — i.e. use the patient as a kind of kamikaze of cost control,

      If insurers stopped paying $40,000 to $80,000 for heart surgeries, or stopped paing $150,000 to save premature babies, and stopped paying $5000 a day for intensive care in general, that would save more money in six months than making millions of healthy peiple skip a doctor’s visit.

      Yet that kind of cost control would mean confrontation with big medicine, and it would also mean accepting some deaths or at least death panels.

      My view is that if you want to cut costs, go where the costs are. But be ready for fierce resistance.

      Bob Hertz, The Health Care Crusade

    • My firm switched us to a high-deductible plan with an HSA attached. The firm funds the HSA, which is great and obviously means my coverage is still paid for by someone else.

      But I do have to follow the HSA procedures of being billed by the provider, paying it, and submitting the receipt, and I’ll admit that I am now way more knowledgeable about the costs of healthcare. And, jeez, the cost is ridiculous. But here’s the thing: what the hell can I do about it? It’s near-impossible to comparison-shop. I had to take my asthmatic son to the ER recently. The bill for a couple of nebulizer treatments, some oral steroids, an obviously the related labor for a two-hour visit? $1,412. My other son injured his leg and had to have an x-ray recently. That was $582. I inadvertently went to an out-of-network sports doctor the other day (they used to be in network, now are not) and saw the guy for 15 minutes. That was $488.

      The main message I’ve taken away is that THANK GOD I do not have to pay for this stuff myself. You’d be faced with awful decisions all the time. I wouldn’t have wanted to skip that ER visit for the asthmatic son, but $1,400+ for that visit? I’d think twice if it was self-pay.

      And it doesn’t seem quite right, these prices. Was it “worth” $488 for me to talk to an orthopedist for 15 minutes? Are we really to believe his hourly rate (which includes overhead for his practice, of course) is nearly $2000? This clinic schedules patients every 10 minutes, too.

      The practitioners have pricing power in the relationship among them, the insurers, and the patients — especially when they combine into groups like Sutter Health. There’s no way I can get anyone to bring the costs down, just because I now happen to know that these services are way more expensive than I had thought.

    • Thanks for your post, Sarah.

      As director of the Health Care Crusade, I have had a solution for your problems for several years.

      I call it Health Courts.

      These would be small-claims courts, where anyone can bring in an unconscionable medical bill. The courts would have the power to reduce or even nullify the bill. No debt would be incurred while the bill was under review.


      Bob Hertz

    • I have a HSA plan, too, and think the only place where there is room for comparison shopping that saves money for both the insured and the insurer is in buying generic drugs instead of brand name drugs. That’s the one area where I have been conscious of requesting the generic, both before and after we hit the deductible.

      I’d also note that in contrast the the co-pay type of plans, which always give a small incentive not to see a doctor (i.e., the co-pay), with a HSA plan, once you’ve hit that magic number, you have zero incentive to forego additional treatment/doctor visits. (I assume my plan has an annual or lifetime limit, but I am talking about on the margins, not at the extreme.) So if your kid has a marginal fever that may or may not require the doctor visit, a $20 co-pay might cause you to wait it out, but with nothing on the line, you might be more inclined to just go to the doctor.