Means Testing

This post by Peter Frase (h/t Wonkblog) on the pitfalls of a means-testing welfare state is well taken. Frase concerns himself mostly with the political vulnerability and inherent ethical tension of non-universal benefits, but there's also the obvious and thorny policy problem that means testing imposes an effective penalty on additional earnings at low income levels. Obamacare's exchange subsidies, for example, which are calculated based on income, impose a de facto marginal rate of about 15% by Kevin Drum's calculation. Medicaid, of course, creates heavy distortions by cutting off benefits sharply at a given multiple of the poverty line. Nongovernmental program designs, such as need-based financial aid, can create similar problems--under a certain level of income, any additional earnings are effectively vaporized by an equal reduction in benefits. I would actually be really interested in studying the real-world effects of these distorted incentives; they may be small in practice (though I doubt it), but it's that issue, rather than metaphysical concerns about justice, that should really undergird Frase's claim that "ultimately, means tested benefits tend to make the poor less free and less autonomous than the affluent."

I do take issue with one aspect of Frase's characterization of the politics of means testing:

"The fact that the state must adjudicate these issues–and must do so continually over time, since a person’s status is constantly subject to change–means that benefit recipients are constantly subject to arbitrary bureaucratic domination. Universal benefits, on the other hand, require relatively little meddling in people’s lives: in a country with universal health care, the only consideration for the state is whether or not you are a citizen."

It would be nice if this were true, but universal benefits seem to be just as subject to legislative whims as means-tested ones. Medicare, which is not means-tested, has certainly come dangerously close to the Congressional chopping block (on the beneficiary side, that is; Obamacare, of course, contains about $700 billion in provider-side cuts), and may not be out of danger yet. Social Security, too, was one strong Democratic Congress away from being stripped to a shadow of itself in George W. Bush's second term. I suppose by "bureaucratic domination" Frase may be referring more to the regulatory side than to Congress, but it's definitely not the case that universal programs are blessed with intrinsic stability. Frase's overall point, however, is vindicated by the fact that universality seems to correlate well with popularity; as he says, Medicare and Social Security are famously resilient--not for lack of trying on the part of the right--while the means-tested Medicaid, food stamps, and TANF programs are perennial targets of budget-cutting zeal.

It's absolutely the case that future welfare state projects should eschew means testing, for all the reasons Frase cites plus the inherently suboptimal policy tradeoffs that means testing entails. I'm not particularly concerned by the Obamacare subsidies; they're a necessary evil in a bill full of necessary evils, and I think their political impact will be heavily muted by the fact that so few loud voices, particularly on the right, have actually bothered to understand or explain the workings of the bill in the course of their demagoguery. They also affect a rather small proportion of the population, and the de facto marginal penalty they impose is pretty, well, marginal. But going forward, the left should work to create universal and universally popular benefits. Single-payer health care, universal childcare and preschool, a guaranteed basic income, and what you might call a carbon dividend (straightforward universal redistribution of the proceeds of a carbon tax) all represent fruitful areas of focus for the left, as they would combine superb policy with the political and ethical benefits of non-means-tested programs.