• There are differences between us

    I try so hard not to focus on politics that sometimes I convince myself that deep down inside we are more alike than different. While that may be true for most things, once in a while I see something that forces me to realize that there are some philosophical divisions not easily overcome. For instance when I read the following from Paul Krugman today, I agreed wholeheartedly:

    The main thing, though, is the strengthened role of and target for the Independent Payment Advisory Board. This can sound like hocus-pocus — but it’s not.

    As I understand it, it would force the board to come up with ways to put Medicare on what amounts to a budget — growing no faster than GDP + 0.5 — and would force Congress to specifically overrule those proposed savings. That’s what cost-control looks like! You have people who actually know about health care and health costs setting priorities for spending, within a budget; in effect, you have an institutional setup which forces Medicare to find ways to say no.

    In fact, I think this is my favorite part of the proposal. I spend a lot of time thinking about health policy, cost-effectiveness, and decision analysis. And I’ve said over and over that I favor a strong IPAB. Moreover:

    [I] think we can have a panel of experts (which should include physicians) try and determine which care isn’t worth the money and stop spending as much public money on that. This will mean that if individuals want to get that care anyway, they have to pay for it themselves.

    Knowing what works and what doesn’t is not easy; it’s complicated. Since we’re spending public money, I think it’s reasonable to say we’re not going to pay as much for things that don’t work as well. This is also part of value based insurance. I’m nodding even as I write this.  Then I read this from Andrew Sullivan:

    Even Krugman is happy. Because the Obama plan includes much greater power for top-down rationing dictated by “people who actually know about health care and health costs”. Gulp.

    The “Gulp” made me pause. Readers of this blog know of my respect for Andrew Sullivan, so the fact that this made him nervous made me uneasy. It forced me to face the fact that there are many, many people out there who disagree with me, who think this is a bad idea.

    I’m not changing my mind. I still think this is the best way to go. But it wil be important to make sure people understand that the IPAB and such will have public oversight and can be over-ruled. This also doesn’t stop people from getting care they want. It stops you and me from having to pay for that care when experts think it doesn’t work. If they still want it, they’re free to pay for it. Moreover, this happens already today. If you have private insurance, believe me, someone is deciding what your insurance will and will not pay for. It’s just not public, and I think that’s a bad thing.

    It comes down to this, again and again. I think that experts are better positioned to make many of these determinations; others think that individuals are. This is not a little difference, it’s actually pretty big. But I’ll make the following concession: if you show me high quality research that proves me wrong, I’ll change my mind. Will those who disagree with me say the same?

    • I think that this is the key:

      “This also doesn’t stop people from getting care they want. It stops you and me from having to pay for that care when experts think it doesn’t work. ”

      The media likes to leave that out of the debate because leaving it out makes the debate more interesting to listeners and so impacts their ability to sell ads.

    • Question: does the history of the Soviet Union from 1917 to 1991 count as “high quality research?” I’m not being facetious–the failures of their economy should have proved to everyone that central planning cannot work.

      I appreciate your open mind on this issue, but I cannot answer your last question in the affirmative. As you point out, these are *philosophical* differences that divide us on these issues. Philosophically, I do not believe that human beings have the capacity to understand something as complex as the market for health care services. And if an all-knowing panel of experts were magically created, I do not believe that they have the capacity to act rationally and altruistically in all circumstances. My political views are shaped by my belief that we are small creatures with limited capacity for knowledge in a very complex universe. I honestly cannot conceive of any evidence that you could provide that would change that belief.

      • No, The Soviet Union does not count. First of all, that’s a country, not a health care system. And there are plenty of health care systems with expert panels that do quite well.

        I believe there are studies showing that co-pays can cause harm (and raise overall costs) in Medicare. If someone shows me studies showing the same for expert panels, I’m willing to read them.

        • Can you give a couple of examples of “health care systems with expert panels that do quite well?” I’d like to know more.

    • “It comes down to this, again and again. I think that experts are better positioned to make many of these determinations; others think that individuals are. This is not a little difference, it’s actually pretty big. But I’ll make the following concession: if you show me high quality research that proves me wrong, I’ll change my mind. Will those who disagree with me say the same?”

      Start here:

      “The Wisdom of the Crowds,” by James Surowiecki
      DARPA’s FutureMAP experiment
      The USS Scorpion project

      Those last are particularly good evidence, as they demonstrate just how bad groups of experts really are in comparison to large groups of people with a financial motive. If that’s not good enough though, please, let me know.

      • I’ve read The Wisdom of Crowds. I don’t believe it applies here. It does in some places.

        Experts can know things that crowds don’t.

        I’m looking for research specific to health care systems. Specific to Medicare would be even better.

        • This is going to be a problem then.

          I like the above examples because they are objectively verifiable. You either find the sunken ship or you dont, you either predict the PM’s assassination or you don’t.
          In medicine, you don’t get objecitve verifiability.
          So, the best thing we can do is either accept the word of a relatively small group of experts because they’re experts or be skeptical of them because they are a relatively small group.
          The evidence we do have from when the there was objective verifiability suggests skepticism to be the more prudent response.

          To your stated objection, I do not disagree with the claim, but rather the conclusion. The whole point of the wisdom of the crowds, is you don’t need expert information. In the Fatal Conceit, this argument is made much better than I am going to be able to reproduce, but the idea is that a million people, each with only a millionth of the relevant information, do not need to the collected sum of their information to be equivalent to the whole – there can be overlap, even strict overlap if you want to interpret the model to the extreme case – and you will still get better results, simply from there being more people making the decision. If people only have limited cognitive ability, and are prone to prejudice and bias, then a smaller group is going to be less able to overcome this than a larger one would be able to. So, even if that larger group is, as a whole, collectiely less well informed, it might not matter.

          A specific case of this was the USS Scorpion I mentioned above. The random outsiders did not have the information about the ship’s last coordinates, they did not have the information about the tidal flows, they did not have the information from the ship’s mission logs. Experts had all of this information, yet the experts would have never found the ship – it’s location was outside even the expert’s 99% confidence interval for where it could have been – but the “market” was able to find it on the first try. Even though the crowd as a whole could not know things the experts could, they were still able to do radically better. This is a big deal, I do not think you should be writing it off because it is not identical to the problem you are looking at.

          • Different Aaron –

            I’ll admit this is the first time I’ve ever heard of the case of the USS Scorpion, but from what I’ve read about it so far, I’m not sure it’s quite as supportive of your argument as you think it is. (And please correct me if I’m wrong on any of this)

            While the Naval experts do seem to have failed in finding it, the stuff that I’ve read seems to indicate that there was a fair amount of expertise involved in the crowd that ended up finding it. A lot of credit seems to be given to Gordon Hamilton, who had a listening station on the Canary Islands. From an article by Stephen Johnson, who wrote the book “Silent Steel” about the sinking of the Scorpion ):

            “When Gordon Hamilton of Columbia University discovered that his La Palma Island (Canary Islands) listening station recorded the loss of Scorpion, the data provided numerous details including the time of the event and the location of the Scorpion’s wreckage by triangulating the signal arrival times using mathematics. (This method was an everyday task for Hamilton who perfected methods of calculating splashdown locations of ballistic missiles during accuracy tests.)”

            Though he was not a part of the official group of experts who initially failed to locate the submarine, he does seem to have leveraged a fair amount of expertise in finding it. Maybe there are others I don’t know about who contributed to finding it with out any knowledge that the average American would not possess, but I have found nothing to that effect. Please point me to examples if you have any.

            You may argue that even still, the initial failure by a group of experts and success by people not included in that group argues against the groups Dr. Carroll references in this article. I would say that because health is such an important issue to most people (as compared to the loss of a submarine, which was likely of interest only to a select subset of people with relevant knowledge) we need a more objective group if we are to control spending. If the public could control it on their own, I don’t think we’d even be having this discussion right now.

            • Here is what I have always had taught to me, and how I have come to understand it:

              The USS Scorpion had a torpedo or something explode somewhere in the middle of the Atlantic Ocean while on some mission, conspricy people say covert mission to spy of some group or another – this is not important. What is important is that the Scorpion was one of two submarines in the world at the time that had some weird nuclear thingey. The navy was worried because simulations they had run said there was no way the ship could have sank the way it did without exposing the radiation in the core, and given that they thought the ship was relatively near the coast line, they were very worried about radiation leaking for the ship.

              So, the Navy got a bunch of really smart people, put them in a locked, windowless room with maps of the ocean, the ship’s logs, mission reports, weather data from the area the ship went down, they ran a really antiquated version of a Monte Carlo simulation using all of this data and generated a really big map of everywhere the ship could possibly be, along with an expected point and various sized confidence intervals. They were conviced the ship had to be in one of a few places. After spending all sorts of time looking for the ship in those places, they gave up, said they would monitor the area for radiation leaks, and move on.

              Dr Hamilton got ahold of listening data from some look out post near where the ship was supposed to be, took the data from the listening post to a bunch of his collegues, they looked it over and everyone guessed where they thought the ship was, they all guessed areas that were right around where the Navy was looking. he then took the data to a bunch of random people he knew that were vaguely associated with marinecraft and salvage, they, with none of the expert knowledge (and this is the important part) guessed where they thought the ship was. Guesses were all over the place, some North Atlantic, some off the coast of Africa, some in the middle of the Meditterranean, etc. But, when Dr. Hamilton ran the collected data from all of these random guesses into a Bayesian simulator, he got a data point that was so far away from where the Navy was looking, it was out of their 99% confidence interval. However, once the Navy looked at this spot, they found the Scorpion – nuclear core intact – about 220 yards away..

    • Dr. Carroll – I agree wholeheartedly.

      To those who would argue that these are decisions better left to individuals rather than some experts . . . why do you go to doctors at all? If these experts have no purpose, than we should just be able to crowdsource medicine. Get your diagnosis and recommendations on treatment from an online forum with as many participants as possible, giving no additional weight to those who have spent a lot of time studying and trying to understand the factors at work.

      I’ll stick with doctors.

      • Experts aren’t better, they are less random.
        You could ask 1,000,00 people how to treat your heartburn and you will get a 1,000,000 bad ideas, but the collected sum of these bad ideas will be better than what you’ll get from 1 doctor.

    • The wisdom of the crowd here is altered by at least a couple of things. First, insurance. Since people do not feel as though they are paying directly for their care, this bit of removal makes their decisions less directly money or market motivated. marketing is also another factor distorting the wisdom of the crowd. I also suspect that there is a bit more emotion attached to this kind of decision making than buying appliances. (One could also look at the recent financial collapse as an example of the wisdom of the crowd.)


    • If the question is whether or not a specific treatment works at all, then perhaps an expert review would beat the market.

      When we get into cost/benefit though, I’m less convinced. Couldn’t conditions like barriers to entry for suppliers of certain treatments be driving up the prices, thus skewing the cost/benfit? Are a panel of medical experts going to be able to tackle these questions as well?

      Bring on the evidence, I’m willing to listen, but I think it would be nothing less than a revolution if the IPAB (or any other panel of experts) really could solve the ‘calculation problem’