I am only going to make the problem worse, but I think the Oregon Medicaid study has rapidly become of the most overdiscussed pieces of policy analysis in Internet history. To see why, you need to see what the three different Medicaid debates really are.
The first debate asks how valuable is Medicaid to low-income families compared with some equally expensive benefit? This is a really fascinating question. One reason is that when we calculate how many families are living below the poverty line, we don’t include the value of in-kind benefits like Medicaid at all. But Medicaid is, among other things, supposed to be an anti-poverty program. So we might want to know how effective Medicaid is at fighting poverty, and to do that we need some sense of how useful Medicaid is to people compared with something like an equivalently costly cash benefit. This would be a great question to research, but unfortunately the Oregon study doesn’t address the issue at all. I would love to see policymakers design a study along these lines, but they haven’t.
The second debate asks do the welfare gains to poor families of getting Medicaid exceed the welfare losses to rich families of the Obamacare tax hikes that pay for them? Frankly if you doubt that the answer to this is “yes” you probably need to have your head examined. The Oregon study doesn’t specifically address this question either, but the finding that giving people free health care benefits them financially (duh) strongly suggests that the obvious “yes” answer is in fact correct.
The debate asks do the Obamacare tax hikes do long-term damage to the economy that outweighs the good of Medicaid expansion? The Oregon study, again, does not address this question. More to the point, this is really a question about taxes and not Medicaid. Far and away the most important thing dividing the economics brain trust of the Republican Party from the economics brain trust of the Democratic Party is the seemingly obscure issue of whether marginal tax rates on investment income have large consequences for long-term economic growth. If you think they do—and Republicans really think they do—then this totally swamps the Medicaid issue. Over time, growing at 2.2 percent rather than 2.3 percent (or whatever) compounds into a vast social and economic catastrophe that would totally swamp any plausible gains in terms of poor people’s blood pressure management.
Of these debates, it’s the third that’s far and away most practically consequential to American politics. If you could change the minds of one camp or the other, you’d see enormous shifts in the terms of public policy debates in the United States. Unfortunately, this is precisely the sort of issue that is not amenable to resolution via RCT or other empirical modes of inquiry. Empirical evidence is relevant to people’s views, of course, but intelligent and well-informed people can be found on both sides of the argument (this is an important subtext of Adam Davidson’s joint profile of Glenn Hubbard and Larry Summers), and it’s going to keep on keeping on. The first debate is the one that would be easiest to explore empirically, and is something I really wish we’d look into since it would at least change how liberals think about a lot of things. But ultimately the political argument in Congress is much more about redistribution than about program design, and even though it’d be convenient for journalists if there were good experiements on the growth impact of redistribution, there really aren’t.
