Can't He Just Eat His Doughnut? (Is Obama Setting the Stage for a Social Security Means-Test?) Ramesh Ponnuru, opposing Obama's plan to apply the 6.2% Social Security payroll tax to earnings above $250,000--it now stops at $102,000--says it would undermine the rationale of the system:
Social Security is structured so that the more you pay in, the more you get back. That's what supposedly makes it a compact among the generations and not a welfare program. Actually, what it does is make it an inefficient, disguised welfare program. [E.A.]
A couple of points:
1) Changing how we finance Social Security won't turn it into a welfare program, or unmask it as a welfare program. A "welfare" program pays out benefits according to need whether or not the recipient works--at least that's the distinguishing characteristic of "welfare" people don't like. Social Security, in contrast, pays out benefits only to those who work for them (for the required number of quarters). Because of this "work-test," Social Security wouldn't be "welfare" even if it was funded entirely out of general revenues generated by the regular progressive income tax.
2) That said, funding Social Security through a payroll tax underscores the "work test" by mimicking the contributions in an ordinary pension plan. In this Contributory Model, you pay in part of your paycheck until you've paid enough to "cover" your benefits, then (if you keep earning) you don't have to contribute any more. Some liberals may hate this "capped" payroll tax as regressive, but it's served the system well, emphasizing that you only get benefits through contributions, that rich and poor both contribute as well as benefit, that benefits are a finite foundation for retirement and not part of some general liberal redistributional impulse.
When Obama fuzzes up the "contribution" part of the tax--by starting up the payroll tax again above $250,000 in income--he risks undermining support for the system (even though he'd keep other regressive "contribution" aspects in place, including a) taxing low-income workers from their first dollar** and b) taxing only earned income and not investment income.) The new Obamified system wouldn't be welfare, but in order to make that point defenders would have to rely more heavily on the "work test."
3) The third basis of support for the system--one emphasized by its traditional defenders--is a political calculation based on crass economic self-interest. Call this the Crass Calculation. Specifically, the system has to benefit enough people far enough up the income ladder to seem worthwhile to them. After all, even if it's not a welfare program it might not be a program voters want to support. The political fear is that the middle class or upper middle class will say "hey we're contributing all this money but not getting much back in benefits, so the hell with this system." By asking the affluent to pay what by historical standards is a big extra chunk of their income--at least 6.2% at the margin, maybe double that if you assume workers wind up paying for their employer's additional 6.2% share, definitely double that for the self-employed--Obama risks provoking the bailout reaction. That's one reason he creates his "doughnut hole" of no payroll tax between $102,000 and $250,000--there are lots of voters in that range he doesn't want to chase away. (If he has to eat the doughnut hole to get more revenue, the whole plan could collapse, politically.)
Krugman's right that the size of this one Obama hike in marginal tax rates for the rich has been vastly underplayed in the press. High-marginal rates in themselves are not a good thing--they encourage tax evasion, for starters, as well as elaborate legal schemes that funnel money not into the most productive uses, but into tax shelters that avoid the high marginal rates (by, say, reclassifying it as capital gains). Lots of laywers and bankers take their cut in these wasteful shelter shenanigans. If Obama really will hike the top rate into the 60% pre-Reagan range, as Krugman and the Tax Policy Center suggest, that's a big deal. And if 6-12 points of that 60% will come from Obama's Social Security hike, we'd better make sure we're getting something pretty important in exchange.
There's an alternative to raising Social Security taxes on the rich, after all. It's cutting Social Security benefits for the rich. The simplest way to do that is through a "means test"--people who were wealthy when they retired would have their benefits drastically cut. Maybe at the very top they'd get no benefits at all. The sums that could be saved through a means-test are staggering, hundreds of billions--money that, as Krugman notes, Obama will need for health care.
Means-testing wouldn't turn Social Security into welfare--again, because of the "work test." Social Security would become government insurance for workers against the possibility that they won't be very well-off in retirement! Yes, means-testing might undermine political support for the system, by violating the "contribution" model (nobody could even think they were getting their contributions back) and by altering the Crude Calculation so that those who confidently expected to be affluent would have little selfish reason to support the stystem. But--and here's the key point--Obama's already diluting or dispensing with those other two sources of support. He's fuzzing the Contributory Model. He's undermining the Crude Calclation. He can still rely on the work test--he can say, "Hey this is to help old folks who've worked all their lives," etc.. But as long as you're relying on the work test, why not rely on the work test to justify a more radical, more progressive and more lucrative reform of the system that would not just tax the rich but stop shelling out unneeded benefits to them.
There might be reasons. Maybe Obama feels it's politically easier to tax the rich than to cut their benefits. But it's not wildly obvious why this should be so. Both tax hikes and benefit cuts involve imposing economic harm. But (a) tax hikes are immediately felt by people of all ages. Benefit cuts impact most immediately only on the old. (b) High tax rates, as noted, have bad side effects that hurt the whole economy. Means-testing wouldn't have those effects (though it might have others, such as discouraging savings). (c) Means-testing could be sold as a shrinking of the program--as Ponnuru points out, why run lots of money through the system by taxing the rich in order to pay benefits to the rich?