Means and Extremes: How Not to Balance the Budget
Tuesday, June 28, 2011
Should we impose a means test on Social Security and Medicare benefits? No.
Last month, the Peter G. Peterson Foundation released budget plans authored by analysts at six think tanks from across the ideological spectrum: the American Enterprise Institute, the Bipartisan Policy Center, the Center for American Progress, the Economic Policy Institute, The Heritage Foundation, and the Roosevelt Institute Campus Network. Along with Joe Antos, Alan Viard, and Alex Brill, I was one of the authors of the proposal from AEI (and I’d note that it represents the authors’ opinions and not those of AEI or other AEI scholars).
All of the plans managed to put the budget on a more-or-less sustainable track through tax increases, spending reductions, or a combination of the two. One plan, from Heritage, stood out in that it went further in cutting spending and it stabilized the debt sooner than the others. The principle difference between Heritage’s plan and AEI’s (which are similar in many other respects) is that Heritage imposed a means test on Social Security and Medicare benefits.
The reason means tests have been unpopular in the past remains true today: they penalize people who work and save.
I’m working on a general article for National Affairs on means testing that will come out in the fall, but since we’ve gotten some questions about how the plans differ and why we didn’t go for a means test in our approach, I thought I’d quickly run through the issues. I’ve talked to the folks at Heritage and welcome a back-and-forth discussion. I’ve worked with Heritage and even written papers published by them, so this should be taken as a disagreement among friends.
A means test makes the payment of a government benefit contingent upon the income or assets—that is, the “means”—of the recipient. Many programs for low earners are means tested, but means tests for Social Security and Medicare are more limited. Retirees may pay income taxes on part of their Social Security benefits and high-income retirees pay higher Medicare premiums, but the effects of these de facto means tests are usually pretty small.
Older individuals are particularly sensitive to marginal rates, because unlike others they have the option of retiring.
Heritage imposed a means test more aggressively, which may reflect a new willingness among conservatives to use means tests to limit spending. Columnist Charles Krauthammer has called for a means test for Social Security “so that Warren Buffett’s check gets redirected to a senior in need.” Likewise, former Minnesota Governor Tim Pawlenty favors means-tested Cost of Living Adjustments (COLAs).
But the reason means tests have been unpopular in the past remains true today: they penalize people who work and save. Some may perceive this as unfair; fair or unfair, means tests have negative effects on incentives to work and save.
Heritage’s means test would reduce Social Security and Medicare benefits for retirees with non-Social Security income over $55,000 and eliminate them for those with incomes over $110,000. The means test is limited: around 9 percent of individuals would have some benefit reduction and around 3.5 percent would lose their benefits completely. (It’s not clear if these percentages would remain stable over time.)
Since a typical person at that income level has Social Security benefits of around $14,000 and Medicare benefits, under Heritage’s plan of around $11,000, Heritage’s means test implies a loss of around 45 cents in benefits for each dollar of income over $55,000. In effect, it’s an implicit marginal tax of 45 percent on top of the 25 percent flat tax rate under Heritage’s proposal, for a total of around 70 percent.
We don’t want to set a precedent that 70 percent marginal rates for anyone are a good policy prescription for addressing the U.S. deficit.
One of the lessons of tax policy is that most people aren’t going to pay that kind of tax rate. My National Affairs article will point to some interesting research finding that older individuals are particularly sensitive to marginal rates, because unlike others they have the option of retiring. If people stop working then they won’t get the income and the government won’t get the spending reductions.
In practice (or using dynamic scoring), I doubt Heritage’s means test would produce nearly the savings that a static projection implies. Some people would limit their earnings to avoid it, as individuals already do with Social Security’s Retirement Earnings Test, which reduces social benefits for early retirees who earn more than around $14,000. If individuals work less to avoid the Heritage means test, that’s bad for them, but it also means lower budget savings. Other retirees might avoid the means test by delaying claiming Social Security benefits until after they stop working completely. The main savings would come from individuals with incomes far above $110,000, for whom the negative marginal incentives wouldn’t apply. But there are only a tiny number of people in this category.
Heritage’s plan projects a reduction in Social Security and Medicare outlays of 1.4 percent of GDP from 2011 to 2015, much of which presumably comes from the means test (although savings may also come from abandoning ObamaCare and other areas). Personally, I do not think this reduction would really happen. Moreover, I don’t think it should happen, in the sense that we don’t want to set a precedent that 70 percent marginal rates for anyone are a good policy prescription for addressing the U.S. deficit.
Heritage imposed a means test more aggressively, which may reflect a new willingness among conservatives to use means tests to limit spending.
To be fair, Heritage’s plan would exempt the first $10,000 in retirees’ incomes from taxes, which would encourage low-income retirees to work. However, for individuals in the means test’s income range of $55,000-110,000, this exemption would actually exacerbate disincentives to work (in econo-speak, they’d have a positive income effect on top of the means test’s negative substitution effect, both of which discourage work). I believe Heritage’s plan would do away with some of the smaller entitlement means tests under current law, although the implicit taxes from these current policies aren’t huge. So, on net, this is a pretty big change.
Because Heritage’s plan (like AEI’s) would shift the general tax code to a consumption base, it might seem that the true marginal rate would not be the sum of the implicit tax through the means test and the 25 percent consumption tax rate. But that’s not right. Consumption taxes alter the incentives whether income, once earned, would be spent or saved, but they don’t change work incentives overall.
If you needed to reduce Social Security and Medicare outlays in a hurry, there is a way around this: cut benefits based on individuals’ lifetime earnings rather than their current incomes. Social Security benefits are based on lifetime earnings, so we already know what people earned. If you cut benefits for retirees with high lifetime earnings, you would give them the incentive to earn more to make up the difference, not the incentive to earn less to avoid the penalty. I’m not saying it would be popular and it wouldn’t be as closely targeted as a means test based on current income, but it would be more effective in generating budget savings and more conducive to encouraging individual work and saving.
Clearly we need to get on top of the budget deficit, and to their credit, Heritage’s folks were willing to make tougher choices than most of the rest of us. But, at least when it comes to the means test, I think the benefits aren’t worth the cost.
Andrew G. Biggs is a resident scholar at the American Enterprise Institute.
FURTHER READING: Biggs has also written “Public Pensions Roll the Dice,” “Is Social Security Middle-Class Welfare?” and “The Way to Cut Taxes and Deficits.” He recently published “Comparing Federal and Private Sector Compensation” with Jason Richwine of the Heritage Foundation.
Image by Darren Wamboldt/Bergman Group.