Since the release of the Oregon Health Study last week, conservative columnist Ross Douthat and liberal blogger Matt Yglesias have both written articles broaching the idea that the government should offer poor people more cash benefits in place of health benefits. Douthat argues more broadly for re-conceiving health insurance as something akin to homeowners insurance: a backstop against catastrophic losses, not a comprehensive package covering everyday expenses.

They might be right. But they are eliding a key reason that the U.S. -- along with every other advanced country -- has health-care policies aimed at providing people with coverage for both routine and extraordinary expenses. The Douthat and Yglesias plans would each make insurance against catastrophic losses universally available and provide redistribution from the rich to the poor, but they would do much less than Obamacare (or even the status quo) to redistribute from the healthy to the sick.

Some people have the misfortune to have chronic illnesses -- diabetes, HIV infection, kidney disease -- that can cost thousands of dollars a year to treat. A system of catastrophic-only coverage, say one that only covered health expenditures exceeding 10 percent of income, would leave these people poorer while making healthy people better off, even if it came with cash grants funded by the savings from reduced health-care costs.

Comprehensive insurance that covers routine care is not “insurance” in the sense of covering expenses that are unexpected at the individual level. But it is social insurance that covers the unexpected event of being a person with high ongoing needs for routine care. Switching to a catastrophic coverage model would mean giving up this social-insurance component of Medicaid, Medicare, the Obamacare exchange subsidies and every other government policy that aims to make comprehensive insurance available.

Maybe that is an acceptable outcome. The government does not enact policies to compensate people for every instance of bad luck. Doing so for poor health might be especially wrongheaded, since it often arises from a mix of luck and choices; offsetting the financial penalty associated with chronic illness may be reducing people’s incentives to stay healthy. And if health-care costs continue to rise faster than overall inflation, a universal guarantee of comprehensive coverage would require ever-higher marginal tax rates and eventually become untenable.

Still, that a policy will reduce the real incomes of the sick while increasing the real incomes of the healthy -- thereby increasing the inequality of happiness, if not necessarily inequality of income -- is a factor that should weigh against a switch to catastrophic-only coverage. I'd be curious to see whether Douthat and Yglesias disagree.

(Josh Barro is lead writer for the Ticker. Follow him on Twitter.)