I like this post by Megan McArdle. I agree with many of her philosophical principles, and think it’s a good idea for us to figure out what our principles are, before arguing about health care plans. Do we disagree on our principles, or do we disagree about which plan will most effectively satisfy these principles? Here are a few from her list:
- We have some obligations to future generations, if not necessarily future individuals within those generations. Extreme thought experiment to clarify the principle: we cannot strip mine the earth and leave them to die.
- States have an absolute right to tax their citizens to provide public goods, i.e. goods that are broadly beneficial but non-excludable. They have a right to enact other laws, such as public health rules, to achieve similar ends. Both rights are constrained by the basic rights of their citizens. You may perhaps quarantine Typhoid Mary. You may not shoot her.
- Societies have a right to organize themselves to improve the justice of their income distribution. That organization may include taxation. It may also include property rights, or outlawing behavior like blackmail.
- Just income distribution is not just a matter of relative position, but also of how the income is acquired, and absolute need. I do not have any moral claim whatsoever on a dime of Warren Buffett’s fortune, because I have a perfectly adequate lifestyle. I still wouldn’t have any claim on his fortune if he suddenly got 100 times richer, provided that he acquired that money through means that we regard as licit.
- Societies should strive to organize themselves so that everyone in the society can, if they desire, acquire the means to provide their basic needs.
- There is no per-se right to health care, since “health care” is not a thing, but a shifting collection of goods and services with amorphous boundaries. Health care is a subset of the modern “basic needs” package, and therefore falls under broader distributional justice claims. No matter what your distributional justice intuitions are, it would be perfectly acceptable, if impractical, to give very sick people the cash required to treat their cancer, and let them blow it on a trip around the world.
- Taxation should strive to equalize the personal cost of taxation among all members of society, not the dollar amount or the percentage of income. That is, it is appropriate for Warren Buffet to pay a higher percentage of his income in taxes for shared public goods than I do, because the personal cost of taking 25% of his income is much lower than the personal cost of taking 25% of mine.
Let’s say that we agree with those principles, in particular, that middle and upper class folks should help pay for lower income people to have their basic needs met.
I might add that goods or services for which evidence of benefit is lacking should not count as contributing towards basic needs being met, even if, unknown to us, that is the case. For example, middle and upper class folks shouldn’t be forced to pay for unproven treatments (there could be exceptions to this).
When it comes to health care, we spend way more than other nations. Bob Somerby keeps reminding us about these statistics:
Total spending on health care, per person, 2007
United States: $7290
Canada: $3895
France: $3601
Germany: $3588
United Kingdom: $2992
Italy: $2686
Spain: $2671
Japan: $2581 (2006)
We’re paying more than twice as much, every year, per person. Yet, we do not seem to have better outcomes. What is really important is to understand where we are wasting money (because clearly we are). Unlike other countries, we lose some money to insurance companies making a profit. We also make it very easy for pharmaceutical companies to extend patent life (e.g., by finding a new indication), which keeps the cost of drugs high. Finally, I suspect that we overconsume health care products.
It was recently estimated that 46% of treatments have unknown effectiveness. There are all kinds of ways that treatments might look more effective in research publications than they really. See here, here and here for a few examples. This suggests to me that we are probably spending way too much on useless treatments. Based largely on the RAND experiment, Robin Hanson argued that medical spending could be cut in half. The RAND experiment found that people randomized to the full health coverage group spent 40% more on health care, but did not have better outcomes. While variations in health care spending do not seem to explain differences in outcomes, other types of variations do (lifestyle, environment). Phillip Longman has a very interesting essay on the topic:
A child born today can expect to live a full 30 years longer than one born in 1900. Improvements in medicine, however, played a surprisingly small role in this achievement. Public health experts agree that it contributed no more than five of those 30 years.This may seem counterintuitive given the attention society pays to medical breakthroughs. But the changes in living and working conditions over the last century are the real reason. American cities at the turn of the last century stank of coal dust, manure, and rotting garbage. Most people still used latrines and outhouses. As recently as 1913, industrial accidents killed 23,000 Americans annually. Milk and meat were often spoiled; the water supply untreated. Trichinellosis, a dangerous parasite found in meat, infected 16 percent of the population, while food-borne bacteria such as salmonella, clostridium, and staphylococcus killed millions, especially children, 10 percent of whom died before their first birthday.
During the first half of the 20th century, living and working conditions improved vastly for most Americans. Workplace fatalities dropped 90 percent. This, combined with public health measures such as mosquito control, quarantines, and food inspections, led to dramatic declines in premature death. In 1900, 194 of every 100,000 U.S. residents died from tuberculosis. By 1940, before the advent of any effective medical treatment, reductions in over-crowded tenements combined with quarantine efforts had reduced the death rate by three-fourths.
…
Consider the startling difference in mortality between Utah and Nevada. These two contiguous states are similar in demographics, climate, access to health care, and average income. Yet Nevada’s infant mortality rate is 40 percent higher than Utah’s, and Nevada adults face an increased likelihood of premature death. As health-care economists Victor Fuchs and Nathan Rosenberg have pointed out, it’s hard not to attribute much of that difference to the fact that 70 percent of Utah’s population follows the strictures of the Mormon Church, which requires abstinence from tobacco, alcohol, premarital sex, and divorce. Nevada, with its freewheeling, laissez-faire culture, has the highest incidence of smoking-related death in the country; Utah the lowest. Utah has the nation’s highest birthrate, but the lowest incidence of unwed teenage mothers. Culture and behavior seem to trump access to health care in improving human life span.
So, if we believe that we should take care of people’s basic needs, and that our current health care system is inefficient, and that we consume too much health care, this suggests to me a fairly simple solution: the government provides all of its citizens with a minimal health care package (a type of single payer system). This includes coverage for catastrophic events and emergency care in general. If you have a broken leg, you can get it taken care of for free. If you’re in a car accident, your ambulance ride and life saving surgery won’t cost you. No one will ever go into bankruptcy over health care expenses again. I’d imagine the coverage would also cover annual physicals and some treatments that are known to be effective. But I’m picturing something far less comprehensive than coverage currently provided by most employers. Should taxpayer dollars pay for Prozac when it doesn’t seem to be better than placebo? Should we pay for cholesterol lowering drugs when the number needed to treat is 100 or more (i.e., if you give 100 people the drug 1 of them will benefit, on average)? I don’t think so.
Employers could offer as a benefit to employees additional health insurance that would cover more treatments. Similarly, those that could afford it could choose to purchase additional health insurance.
Of course, we live in a world of lobbyists. How we would prevent lobbyists from getting their favorite useless treatment covered by a government insurance plan is a major challenge.
Another valid concern about a single payer system is that innovation will suffer. For example, Megan McArdle says:
As long as people don’t know that there are cancer treatments they’re not getting, they’re happy. Once they find out, satisfaction plunges. But the reason that people in Britain know about things like herceptin for early stage breast cancer is a robust private market in the US that experiments with this sort of thing.
So in the absence of a robust private US market, my assumption is that the government will focus on the apparent at the expense of the hard-to-measure. Innovation benefits future constituents who aren’t voting now. Producing it is very expensive. On the other hand, cutting costs pleases voters this instant.
We should be concerned about a decrease in incentives for innovation. It could be that a system that covers everyone saves lives now, but at the cost of innovation. This could cost future generations many more life years.
However, two things to consider. If we adopt a plan like I proposed, there would still be plenty of people purchasing private insurance. Thus, there would still be financial incentives for new drugs and devices. Further, the government currently does fund innovative research. NIH pays for the basic science, and then Pharma uses the knowledge to develop new drugs. The gov’t certainly could let companies compete for research money for device and drug development, in the same way that researchers currently compete for NIH dollars. In my opinion, the old model of funding research without a plan for how the findings could lead to the creation of valuable products is pretty much dead anyway.
There are real tradeoffs between different policies, and we shouldn’t pretend that that is not the case.
N=1 