• The moral hazard – ctd.

    Austin writes me at the crack of dawn about my post:

    I agree with the conclusion of this post, but the set-up bothers me in one way. When care is free or cheap people get more not because they want it per se (colonoscopies is a good example), but because their doctors prescribe it. If my doc says, “An XYZ might help a bit,” and it is free to me, I’ll do it. If my doc says, “An XYZ might help a bit,” and it cost me $1,000 I’m going to ask WAY more questions and am much more likely not to get XYZ.

    This is when I wish I had an html tag for kidding on the square.  I agree my example was hyperbolic.  But the point remains the same.  When people talk about how the moral hazard results in more care consumption, it’s always in hypotheticals.  You can make an argument (and a compelling one at that) with straw men that seem to make sense.  I see very few concrete examples that happen in the real world.

    In practice, I can remember almost no instances where I told a patient what to do and then had a discussion of what it will cost.  And you will have to trust me that almost everyone I see is very, very poor.  You can blame me for that, but that’s how clinicians are trained.  Fix it if you want, but it has nothing to do with consumers.  I look at a set of clinical symptoms and history, and the patient, and decide based on that what the best therapy is.  The only thing I ever really bring up in terms of cost has to do with pharmaceuticals.  I will tell patients that the generic is often just as good and much cheaper, but you’d be surprised how few docs do even that.

    These arguments for the moral hazard as a major problem in health care also ignore the fact that something like 80% of health care spending is consumed by 20% of people.  Those people are not like the rest of us.  How much of that is spending is likely moral hazard related?  I’d guess little.

    I’m not denying that the moral hazard exists.  It does.  I’m denying that trying to limit it through brute force is a good idea.  That’s what the research shows, and I will always take research over theory and arguments.

    Austin continues:

    I believe there is provider induced demand, as the example illustrates (this goes way back; one can cite Arrow; invoke information asymmetry). There are two ways to combat it being discussed today: (1) on the public side, ACOs and related bundling/capitation and (2) on the private side CDHPs. I think we’re going to see more of both. The writing is on the wall. They can both work to limit care (ration!!!). Which is better and for whom?

    CDHPs make some sense for younger, healthier, non-poor, and not cognitively impaired folks (most of the working-age population). I’d support a CDHP model if it had income-related cost sharing and, moreover, included value-based design (a la Chernew).

    However, for the elderly, chronically ill, very poor, cognitively impaired (e.g., most of the public program population), CDHPs make no sense. For them, ACOs are sensible. Put the risk on the provider and manage quality.

    Moral hazard is real. Providers and patients play a role. It is all about money. But how the risk should be allocated depends on which population you’re dealing with.

    Well, sure, as I said in the original post, brute force reducing the moral hazard for healthy people is great.  But how do you differentiate?  And, as John Nyman argues, some people who are previously healthy get really sick and the fact that a moral hazard exists may make them more likely to get care – which is a good thing.

    I think that one of the biggest differences between liberals (with a little l) and conservatives (with a little c) is where they want the push to come from.  Many conservatives believe that the lever should be the consumer, that by increasing cost-sharing, you use the individual to drive the cost of the system.  Thus medical savings accounts and high deductible health care plans.  Many liberals believe that the lever should be the system.  You should limit money into the whole thing and force the many moving parts providing care to do so more efficiently.  Thus single payer systems.

    I come down somewhere in the middle.  I’d say that for the stuff we agree everyone should get, that comprises the base set of quality health care, you ignore the moral hazard.  We want people to get that, and we should make it as easy as possible.  But for stuff that is unnecessary – and there is a lot of it – we let people get additional insurance to cover that.  Or we cost-share that.  Or we make them pay for it themselves.

    I can already hear the howling.  Yes, that is “rationing”.  But we have to get over the idea that rationing is a four letter word.  We should want to ration stuff that is not really necessary.  And we should start to get used to the idea that a lot of what we do falls into that category.

    UPDATE: I’m not saying this last thing would be easy. It would be very complicated and politically unpopular.  But it has to be done.

    • Aaron, I made the following comment under your original post, only to read this follow-up after the fact. I think the questions in my final paragraph still stand:

      I followed you here after reading your delightful State Fair post, so of course I’m not at all caught up on a lot of the things you’ve discussed prior to…last week, I guess. If there’s a post from the past that would speak to my observation, I’d love to see it.

      It seems to me that you addressed part of the Moral Hazard issue without addressing another. The issue, as I see it, is not that people either consume too much health care or can’t afford enough. The issue is that under our current system the true *cost* of any product or service in the health care industry is obscured by the third-party model. With my high-deductible, HSA-compatible plan, I get blood work done and then receive a bill that shows the unadjusted charges as roughly $400. Then there’s a line that says, in effect, “discount for being a Carefirst member” and a giant subtraction, then the total comes to…something like $17. Now, how in the world am I to know what it really cost to do those tests? And how likely is it that anyone (at least on the patient side of the equation) really knows what anything – from bloodwork to a routine physical to an MRI, knows what anything costs? Taking out the third party in the transaction, or fundamentally changing that party’s role in it, seems like it would go a long way to exposing consumers to the true costs (and prices), and providers to a more robust market for their services – in which, ultimately, the providers would have to face a bit of healthy competition due to a more engaged consumer base.

      Stossel’s fillet mignon post covers a lot of ground, and you can argue his conclusions, but I think most people would agree that if you make two things of vastly different quality available at the exact same price, you will tend to see an increased demand for the item of higher quality. If I’m missing something there, I’d love to know. If I’m right, though, I think it’s also fair to say that over time people will become totally inured to the underlying value/cost of fillet mignon.

      And this, to me leads to the problem we have today. We change the oil in our car because we know that it’s in our long-term interest to take care of the engine. A robust market for this service means that it is very inexpensive (compared to the total cost of owning a car, and compared to the cost of replacing it prematurely) to get an oil change. On what basis can we be confident that removing/transforming the third party to medical transactions will NOT lower cost and prices over time; and on what basis can we legitimately project that increasing the third party role, or turning the government into that third party WILL lower costs and prices?

    • I think I could count on one hand the number of times a patient has asked me what something would cost. That would cover over 25 years. It would take a major cultural change to start having people ask about and negotiate fees.


    • Trey,

      I’m not arguing that the moral hazard does not exist. Yes, people who are not exposed to costs are likely to want more. What I think is mistaken is the idea that just increasing cost-sharing across the board will move things in the right direction. I don’t think it’s that simple. That’s the flaw in Stossel’s argument. The filet problem is from an unbelievably simple system. Health care is nothing like that. He’s not wrong in terms of grocery insurance; he’s wrong in terms of health insurance.

      You ask what evidence I have one way or the other. The evidence I have is from studies. I talked about the RAND HIE above. Tomorrow’s post will show more, and how cost sharing can actually lead to increased costs as well.

      As I said in the end of my post, I have no problem with increased cost-sharing for unnecessary care. But increasing cost sharing for everything, including what we think patients need, isn’t the best alternative.

      Thanks for the comments!

    • Trey,

      I appreciate your post, because I was just going to argue esoteric points in the other direction without context. So, anyone reading this, please make sure to first read what Trey wrote.

      Now, Trey argues that we don’t know the true cost of a service under our current system and that by eliminating the third-party payor we could all see the true cost. If the cash market is somewhat efficient, we should already see the true cash price being reflected in what cash-paying patients pay for hospital services. This tends to be not far from the list price, in my personal experience. This would correspond to the $400 list price Trey mentions for blood work. The reason Carefirst is able to get a lower price is that as a group, it has the ability to negotiate lower prices through its control of its members and its size. In health care, providers routinely provide discounts and/or rebates to payors in order to get their business. There’s a pretty well established literature showing that payors with a higher degree of control get the best negotiated prices. So, the $17 in Trey’s example only applies if you have plans that can influence to what provider a patient goes and what services he will pay for (i.e. managed care) . Without these third party payors, we would expect to see less negotiation with providers and greater asymmetries between providers and patients, leading to higher prices. In the example, we’d all be stuck paying $400 a blood test.

      As an example, please think of the last time you or a relative went to the ER and had to pay cash. They don’t give you any discount relative to that list price. An ER visit can easily cost $1400.

      On that note, one of the big problems with consumer driven health plans (“CDHPs”), where there’s a high deductible and often some health savings account cash, is that patients don’t get the negotiated discounts that managed care plans can get. At least this was my experience. Some years ago I signed up for an Aetna plan and within 3 months hurt my foot. It turned out it cost substantially more to see my general practitioner at his cash price ($150 without any x-rays) than I’d gotten up to then in my health savings account. Yet the same GP was complaining that he gets something like $35 a visit from insurance. So, basically, I was getting charged an arm and a leg more with a CDHP than with the regular HMO through the same company. Without patients being able to get negotiated rates (which are often closely guarded secrets), CDHPs seem like a rip off.

    • I had no idea people didn’t pressure doctors about cost…
      I haven’t seen a doctor for more than check-ups/shots since I was 3, but whenever I go to a dentist I always say “right, I’m paying cash. What can you do that’s cheap?”

      They must think I’m the worst patient ever.

    • GrandArch: I don’t think your beliefs are substantially supported by reality. Without third party payers, we wouldn’t all be stuck paying $400 for a blood test, because the test doesn’t remotely come close to costing $400 to perform and so a remotely functioning market would put downward pressure on the price. If the lab only received $17 to perform $400 tests, they would be out of business in short order; since labs exist, the tests obviously don’t cost anything close to $400 to perform, they cost something close to $17. The lab might claim that $400 is the “list price” of the test, but that’s a red herring because no one ever pays anything near that price. In reality, the negotiated $17 represents the price both parties have agreed to pay for the service. If all health care spending converted to cash payments overnight, labs could try to charge $400, but they would quickly be undercut through competition. If consumers are the lever, they could use a comparison shopping website to choose a lab just like they shop for books or gasoline or milk. If the system is the lever, the single payer can use the same competition between labs to set a price. Either way, we don’t need third party payers for prices to reach equilibrium, that’s what markets do.

      You also state that hospitals are charging list price to uninsured patients. Virtually every hospital will give a cash discount and will further discount their rates for prompt payment. Most hospitals will write off fees for uninsured patients on a sliding scale basis. Certainly medical debt and medical bankruptcy are problems, but hospitals tend to adjust rates for uninsured patients according to their ability to pay. If Bill Gates walked into the ER without insurance, he might pay $1,500, but that doesn’t really reflect the cost of the services. The uninsured patient’s initial bill might be for $1,500, but a short discussion with the billing office would drop that figure substantially. Some states actually limit the rates that can be charged to uninsured patients as well.

      Furthermore, US hospitals have to provide that ER visit right away and deal with the billing later (at least to the point of stabilizing an emergency medical condition). You’ll still get the services regardless of ability to pay. Hospitals aren’t going to sue and attempt to garnish your wages over a $1,500 bill when you’re offering to settle for something similar to what they would get from insurance anyway.

      In short, medical billing is a lot more like buying a car than ordering a hamburger.

    • @ Zach,
      Thanks for the points – it’s always helpful figuring out what others agree on and don’t. A few thoughts:
      – I don’t actually claim that blood tests cost $400 and the negotiated rate is $17. I could check some databases of average costs insurers paid, but suffice it to say that I claim that the negotiated rate is lower than the cash price, which may be lower than the list price.

      – As for price competition pushing down prices more than what insurers can do, the economics is somewhat mixed. Contracting theory shows that agents with better knowledge than the principal (i.e. the patient) can divert resources to themselves. As patients, we often don’t know the value of a treatment to our particular situations and we don’t have the resources to investigate them. Insurer P&T committees evaluate treatments and technologies. However, as you suggest, insurers often accept any treatment that’s effective and with flat copays, patients lose an ability to make cost conscious decisions.

      – We might be able to test the varying effects by comparing the negotiated rates of insurers with net prices charged to cash customers and even utilization trends for health plans with coinsurance (a percentage of the total cost as opposed to a flat co-pay). Unfortunately, there’s a high degree of endogeneity here,

      – As for evidence that the negotiated rate is lower, my specialization is in pharmaceuticals. Von Olsen (2001), “Pharmaceutical Discounting Under Federal Law: State Program Opportunities,” cites to Schondelmeyer, with tables of private sector pricing, various other prices and AWP. Needless to say that the cash price is the highest price and insurers with the greatest degree of control (e.g. Veterans Affairs) get the best prices (e.g. 34.6% of list price). Other sources (I’ll have to find them) suggest that prescription assistance programs for the poor are not large enough to substantially reduce the cash price. Obviously, pharmacies have much more difficulty discounting than do hospitals. I’ll keep an eye out for such evidence related to hospitals that takes into account cash discounts and the like.