Monday, March 2, 2009

Why Healthcare Isn't Easily Scalable

One of my mentors (mentioned previously in this blog) mentioned to me a few weeks ago that he thinks there are certain industries that don't scale well, but have scaled anyway because of incorrect lessons learned by CEOs at business schools. His biggest examples were banks and cars. I also see that as potentially being true in software and pharmaceuticals.

I've been thinking, and I'm becoming increasingly convinced that healthcare, as an industry as it exists in the United States, is not a scalable one.

Let's walk through it.

Firstly, because of quality of care regulations, it takes well above-average individual human capital (intelligence, work ethic, responsibility, street smarts, etc) to be a doctor. The government (rightly) wants doctors to be good at their jobs, and typically it takes very capable people to be good doctors. Therefore, medicine requires talented people.

These talented people typically have many job options. Smart people are desired or required in many industries (finance, engineering, medicine, law, etc). In order to attract capable people, medicine needs to offer a pay and quality of life package that is attractive to smart people. Therefore, doctors need to be well-paid.

Now let's walk through history.

Medicare was instituted in 1965. At this time, private sources were responsible for 75% of all health insurance plans ( The government had been taking on more and more health insurance plans over the course of the 50s and 60s, so this was a lower number than historically.

According to the same source, by 1995, the government was responsible for half of health plans. I don't have current numbers, but from looking through a bunch of HMO 10-ks, it seems as if that number has gone up to a little over 60%. In other words, 60% of health insurance plans are through Medicare, Medicaid or other government programs.

This increase has accompanied a fast increase in medical costs (direction of causality unclear to me). As the US economy grows faster than its population, a greater and greater percentage of people want healthcare, increasing demand for a good with a supply whose growth is constrained by the number of people who can enter, as well as facilities and hospital budgets. With increased demand comes increased prices.

Secondly, improved medical technology increases the medical profession's ability to treat a number of diseases. This increases demand for medicine in two ways. Firstly, things that wouldn't have required many medical expenses before because they were untreatable now require heavy medical expenses for sophisticated treatment. Secondly, because we have the ability to treat so many more things, healthcare is becoming seen as a human right rather than a purchasable good. This effect has been growing over time.

Finally, all of these improvements in medical technology cost a lot. Firstly, discovering the advances is costly, and needs to be compensated if future research is to occur. Secondly, the government imposes massively costly testing standards upon medical developments to ascertain safety, increasing the cost of healthcare. Finally, improvements in medical technology spur demand for more improvements in medical technology, which perpetuates spending. Arguably, each incremental medical improvement costs more to develop than prior ones as we tackle low hanging fruit (this is certainly not always the case, as technology discoveries in other sciences make new discoveries easier, but, as with most things, is probably true some of the time).

So you have increased demand as society gets wealthier, increased demand because of improving technology, increased costs of finding further new technology, and a supply constrained by the talent distribution of the country and reduced healthcare budgets relative to demand because of federal deficits. That should lead to some serious growth in medical expenditures, and it has.

If you universalize healthcare, there are a number of possibilities for what happens. Possibility one is that care becomes so expensive that premiums go through the roof, reducing worker wages, increasing copayments and increasing federal budget deficits (which, when they implode, REALLY send premiums through the roof). You also don't see that much improvement in well-being of people who have new health-insurance, because wages need to fall by almost as much as their coverage rises, so you're basically forcing poor people to substitute other consumption for healthcare. I have no normative judgment on whether the last point is good or bad, but generally, this outcome is bad.

Possibility two is that we just don't have enough doctors to service everyone. People with health insurance can't get treated, or can't get treated in time. Lack of healthcare is thereby spread more randomly over the population instead of on the poor. Everyone's quality of care drops as doctors become even more harried and stressed and productivity of our most potentially productive people drops as they get less care than they had. Economy and healthcare quality bad.

Possibility three is that we get enough doctors to service everybody because people substitute to medicine, and because poor people seek out healthcare so much earlier than they otherwise would, healthcare costs actually go DOWN as people stop putting off necessary medical treatments for lack of money to pay for them. This is the dream of universal healthcare.

Possibility four is that universal healthcare merely serves as a redistribution mechanism. This is like possibility one but less extreme. Rich people pay more taxes to the government, which then spends it on poor peoples' healthcare. Healthcare costs go up somewhat for everybody, but much of the cost comes out of the consumption and investment in non-healthcare areas by the better-off. The economy as a whole is worse off, but you fulfill a moral desire to have everybody receive healthcare. Most liberals think this is a great outcome, most conservatives think this is a bad one. As a left leaning moderate, I think this is a mediocre-to-good outcome.

There are a number of other considerations:

1) It's unclear how much moral hazard plays into healthcare decisions. If having health insurance makes people who currently don't have health insurance more likely to engage in activities by which they can get injured or sick, then universal health insurance becomes a big productivity drain and a bigger transfer of money from rich to poor.

2) It's unclear how well insurance companies are able or willing to price insurance to actual individual risk instead of perceived individual risk. If the difference in healthcare costs for obese people vs skinny people or smokers vs nonsmokers is larger than the perceived healthcare cost difference by those buying the insurance, then universal healthcare can represent a societal transfer of wealth from people making good lifestyle decisions to people making bad lifestyle decisions. For example, if an insurance company covers an even number of smokers and nonsmokers, and has to pay $1000 in costs for smokers and $500 in costs for nonsmokers, but smokers and nonsmokers have a general baseline of $750 for insurance costs and smokers aren't willing to go much above that, then you may see a cost of $650 to nonsmokers (who feel they're getting a bargain) and $850 to smokers (who wouldnt enroll if it were higher). Nonsmokers subsidize smokers' insurance.

This may encourage moral hazard (point 1), but whether or not it does, on a moral level, it seems unfair to reward irresponsible people because they chose to be irresponsible in our country instead of another one. As a rule, the poor tend to have higher obesity rates, smoking rates, teenage pregnancy rates, etc. than other classes, so extending universal healthcare could significantly increase moral hazard as well as unrealized costs of bad behavior.

3) People who make bad lifestyle decisions may be able to hide their decisions from insurance companies. If there's a nontrivial percent chance that someone enrolled as a nonsmoker is actually secretly a smoker, then every nonsmoker's health insurance goes up to account for the fact that the health insurance company can't entirely tell if they smoke or not. Both point 2 and point 3 should require comprehensive Pigouvian taxes, but nobody talks about these as part of a health plan, they require constant readjustment as to the exact level, and are often politically sensitive. How this relates to universal healthcare is the same as the second point: the poor tend to have higher obesity rates, smoking rates, teenage pregnancy rates, etc. than other classes, so extending universal healthcare could significantly increase moral hazard as well as unrealized costs of bad behavior.

So you have a LOT of problems with scaling healthcare to a universal level.

What are potential solutions, and how have other countries managed to achieve universal healthcare?

In a sense, we're damned by our own good healthcare.

Population health statistics belie an important fact: US safety standards and physician standards are the toughest in the world. I spent 3 summers working for Harvard Medical School, which verified what you can see by reading through a collection of big pharma 10-Ks: it isn't even close. Drugs are held to a higher standard for use in the US than in other countries. The disparity with doctors is even bigger: physicians are held to a higher training standard than elsewhere to a massive degree.

This should allow other countries piggyback off of the United States to a large degree. The US has a lot of people and a lot of money, so drug companies, when they develop a drug, almost always do so to US standards. Drug companies want to pass US standards, so they typically only greenlight drugs that come in the vicinity of US standards. This means they pass standards in other countries fairly quickly, with very little cost. If government health insurance plans in other countries refuse to shoulder any of the US testing burden (because they have negotiating power, because they can simply refuse to cover the drug for the entire country), they can force healthcare companies to accept lower prices on their drugs - in a sense, they "buy in bulk", and refuse to pay for any US compliance. We shoulder the rest.

Universal healthcare here would allow us to "buy in bulk" in the same way, but fundamentally, our testing standards and last-mover status mean we're going to be paying quite a bit more no matter what. Our last mover status also means that if we "buy in bulk", drug companies aren't going to have anyone left who can pay for costly drug development and will therefore develop fewer drugs. That's a problem roughly akin to the US becoming the world military force (we can't cutback because everyone else already has), but with healthcare the effects are probably more drastic.

Another thing other countries do is they have significantly reduced qualifications necessary for becoming a doctor. Medical school in other countries is shorter, less difficult and accepts more people, so there shouldn't be the supply shortage there is here. As would be expected with more elastic supply, doctors are paid less in other countries. There should be correspondingly fewer talented doctors (the really talented people should go to higher paying jobs) with less training. This is something the United States probably wouldn't be willing to accept, and rightly so.

Other countries have a different sense of care that is "necessary" and is covered by insurance. I heard recently (I have no verification on this) that half of all US insurance company costs are for end-of-life prolonging - when the end is days away but could be prolonged by a few weeks. I was told by this same person that European insurance companies don't cover this care, and therefore most patients are allowed to die at this point.

Finally, other countries culturally, legally and educationally do a better job of preventing many diseases. For example, we have very high obesity and diabetes rates, leading to heart attacks and strokes. We have lots of bad diets, leading to cancer. Our cars are bigger, so we have more deaths in car crashes. We allow guns more freely, so we have higher homicide rates. Etc.

The areas we do have a lead over the rest of the world are alcohol and cigarettes. Just over 20% of the US smokes, compared to mid-30s percentages for much of Europe and way higher in Asia. (cool paper on this: Alcohol, we're at the mean for consumption according to Glaeser and Cutler, but teen consumption is much lower here than in Europe, and we correspondingly have much lower levels of chronic alcohol-related diseases.

Preventatively, we're behind other countries largely due to higher heart attack and stroke rates. This makes universal healthcare much more expensive for us than other countries.

So how do you institute universal healthcare in the United States?

Option 1. Roll the dice. Pray for outcome 3 above. The scant evidence I've seen suggests that 4 (redistribution, economic slowdown, more healthcare) is the likely long term outcome, with 2 (the not enough doctors outcome) in the short term.

Option 2: Reduce testing standards for drugs. Doesn't seem optimal.

Option 3: Reduce qualifications necessary for becoming a doctor. Doesn't seem optimal.

Option 3a: Reduce qualifications necessary for performing basic medical procedures. In other words, let nurses do more and don't spend as much of the doctor's time on simple things. This seems like one good step. Reduces quality of care slightly (for simple things that get botched or seemingly simple things that aren't actually simple) but you have to take a tradeoff somewhere, and this seems like a good place to do it.

Option 4: Reducing the sense of what is necessary care. The end-of-life case is a decent one. It's a moral issue rather than an economic one. If doctors are required to extend life to the best of their ability for each individual patient, as we have in the US, then you majorly increase cost of care for everyone. If doctors are never required to extend life, you have no medical system. Somewhere between those is Europe. I'm not a bioethicist and can't tell you where the threshold should be. This is one option that makes me uncomfortable, but it's an option.

Option 5: Prevention efforts. Pigouvian tax the living heck out of everything bad for your health (cigarettes, guns, trans-fats, etc). Educate better. Regulate better. Creative economic nudges to be healthier. These are by far the best option, but they take time to work.

Option 6: Increase efficiencies. Computerize health records (great private expenditure, shouldn't have been done by the federal government, especially with lousy implementation guidelines. REGULATE, not SUBSIDIZE here). Get rid of excess overhead and bloat in hospitals. This can be good but won't get rid of nearly all of the costs we need to cut.

No matter which of these options you choose, you'll find universal healthcare MUCH harder to institute in the United States than in other countries.

No comments:

Post a Comment