Ezra Klein writes:

The [U.S. health care] system is currently biased toward the worst form of cost control: rationing by income. Every year, we contain costs by quietly letting 2 million or so more people fall into the ranks of the uninsured.

He makes the first statement above without presenting any evidence. There might be some evidence for it, but he doesn’t present it and I think the evidence is weak. For one thing, two big offsets to his claim are Medicare and Medicaid. Medicare recipients, though relatively wealthy, have relatively low income. And the percent of the population with Medicare has been rising as the percent of the population that is elderly has been rising. Medicaid recipients have both low income and low wealth. Perhaps he means, given his second statement above, that people of a given real income are more and more likely to lose their insurance. For that to be true, the percent of the uninsured with income above $x in real terms would have to be rising. Is it? My impression is that it’s pretty stable.

Also, his second statement is false. Actually tens of millions of Americans a year fall into the ranks of the uninsured, just as tens of millions of Americans a year rise into the ranks of the insured. The point is that most people who go without health insurance at a point in time get it later.

Presumably Klein means that at any given point in a year, the number of people without health insurance is typically 2 million more than the number without health insurance the previous year. If that’s what he means, he’s wrong.

Check out the most recent Census Report on the issue, Income, Poverty, and Health Insurance Coverage in the United States: 2007, issued in August 2008. Table 6 on page 20 shows that the number of American residents without health insurance has risen from about 39 million to 46 million over the last 8 years, an increase of less than 1 million per year, or half the number that Klein asserted. Moreover, the percent of people without health insurance has stayed about constant since 1993.