Universal Healthcare in the US

There is a major factor that people aren't factoring in when they talk about universal healthcare. In short, healthcare development almost certainly involves a "learning by doing" model of technological progress.

Many drugs and treatments start off as very, very expensive when they go out into the field. Because they're expensive, they typically see use only in areas where they potentially offer the greatest improvement to patient outcomes (note that this is not "cost-adjusted" improvement - it's actual improvement).

A few things happen.

1) Average cost of care goes up.

2) Average quality of care goes up if the developed treatment is effective.

3) Any problems with the current technology are documented as a result of thousands of uses across the country. Some percentage of these doctors work on improvements, and many more of them report feedback to the manufacturer and insurance companies.

4) Manufacturers work on improving the technology, both by making it cheaper and by fixing problems with current technology.

Through further development, the technology becomes more effective and cheaper.


Yes, I understand that some technologies and drugs hang on even when they show they can never be better than a cheaper status quo, even without adjusting for cost, and eliminating these from the lineup can reduce the cost of care without killing quality of care.

However, the problem with a "cost-containment" public health insurance plan (which isn't a factor in private insurance because of competition and the lack of any insurance plan with a dominant market share) is that if you don't cover expensive but medically effective treatments because they aren't cost-effective, you a) significantly reduce the number of potential improvements to technology identified through use, b) reduce the number of doctors working on improvements, and c) reduce manufacturer incentives to improve the technology because getting a public insurance plan to change what it covers is very time consuming, costly, difficult, and highly uncertain because of lobbying by the current standard technology.

Now, the fact that Europe has socialized health care doesn't really matter for this process - the US sets the standard for healthcare treatments and healthcare development, it has far fewer basic improvements to make to hospitals that European systems still need to make, and, most importantly, the US, as the largest healthcare market in the world, provides a big enough market to incentivize development anyway. European countries can piggyback off of what we do, because they can observe our development and then jump on technology when it hits cost-efficacy. We can't do that, because we're the market that still incentivizes development.

In other words, Obama's public healthcare plan takes an exceedingly static view of helathcare costs, instead of the dynamic one that actually occurs. Healthcare costs have gone way up because we've developed more new treatments in the last 20 years than at any point in history. Rising healthcare costs are happening in Europe, as well - it's not some artifact of an inefficient or badly incentivized system.

In short, rising healthcare costs are not a bad thing - they're an indicator of accelerating medical progress. A point I won't cover, but still an important one, is that they may also the result of a societal willingness to spend a greater proportion of income on extending and improving life vs buying more material goods (not covered in this blog post because it is explained well by Mankiw and other writers)

I'm not saying that the US is as efficient as it could be - there are plenty of structural changes that could be made (and I like computerizing health records as one piece of that puzzle). But I'm not sure that universal healthcare is the right one.

To illustrate the potential effects, I've linked to a paper published in European Radiology called Cost containment and diffusion of MRI: oil and water? The situation in Europe. Turns out MRI is significantly underpenetrated in Europe relative to what it should be for optimal healthcare.
(link to whole article here.)
We can look at the development of the CT scan, the MRI, AIDS medicines, chemotherapy, and biotherapeutic (injectable) medications for similar examples.


To those who say that US medical care doesn't benefit from all the extra technology (and I know a lot of you who argue this), you are making a terrible argument. Gary Becker explains why. (Link, you'll notice, from Mankiw)



And as a final aside, to those of you who challenge the notion that a learning-by-doing model could make sense, I'd direct you to the competition between Boeing, Airbus and Mcdonnell Douglas that eventually drove MD out of business cuz they couldn't make planes fast enough. I think this is the link, science direct is being maintained so if it's wrong, I'll fix it.