Thursday, April 24, 2008

More details on Obama's Social Security plan?

Update: See comments for good reasons why I might be wrong...

This is a couple weeks old, but James Pethkoukis's Capital Commerce blog at U.S. News & World Report has an interview with Obama economic adviser Austan Goolsbee that briefly touches on Social Security reform. Here's the excerpt, followed by a quick comment:

Does Obama want to raise payroll taxes or uncap the income limit to fix Social Security?
What Obama has said is that the problem with Social Security is not a crisis, but that there are fiscal issues associated with the aging of the population. In his view, the place to start is with the regressivity issue of the payroll tax. It is my own view that the majority of Americans do not understand how the payroll tax works. How it is that a guy making $90 million a year is paying the same tax as a guy making $97,000 a year? I think that at best it would strike them as highly weird and at worst grossly unfair.

But he has not called for uncapping it completely, and there are several different ways to do it. But before anybody starts talking about cutting benefits, let's address that issue. It would also get the trust fund to last another 50 or 60 years. An increase in the payroll tax of 6 percentage points phased in over many years or decades doesn't strike me as a dramatic move. They have uncapped the top rates in Ireland, and they have grown extremely well.

The italics are mine. This sentence is interesting since it seems to indicate a willingness to raise the payroll tax rate as well as the cap. (Why? First, it's not clear what raising the payroll tax cap "6 percentage points" means, so I'm inferring it refers to the rate; second, for Obama's plan to lift the cap to keep the trust fund solvent another 50 or 60 years would require it be implemented very quickly, not over decades.) This would seem to indicate a dedication to little or no reductions in scheduled benefits at all.

From my previous work on Obama's tax plan, the payroll tax rate would need to rise by around 4.7 percentage points -- from 12.4 percent to around 17.1 percent -- by 2080 to maintain solvency, even after his plan to lift the tax cap has been implemented.

For what it's worth, Goolsbee's general comments in the first paragraph focus on Social Security taxes distinctly from the benefits that are linked to them. Yes, the Social Security tax is flat when measured up to the cap, and regressive when measured relative to total income. But benefits are paid progressively based on taxes, so low earners can expect to receive a higher return on their contributions than can high earners. Given the program's history, in which taxes and benefits are linked in a single system, to focus on the progressivity of one element rather than the program as a whole seems misguided. The U.S. already has a much higher taxable maximum than most developed countries; eliminating it entirely would put us out of step even with countries with far more extensive welfare states than our own.


Anonymous said...

First, I think you improperly read into Goolsbee's comments about a 6% rise in tax rates phased in. He's clearly talking about phasing in the employees share of the tax for earnings above the max. I don't see how in the context of all his other comments it gets read any other way. In fact talks about rebating some payroll taxes and eliminating some taxation on benefits, so your interpretation just doesn't make sense to me

But then again you see what you want to see because any increase in taxes in anathema to you.

Second, yes, we have a higher ceiling than most countries but: (A) the maximum has fallen from about 90% of earnings to the low 80's all because of changes in the wage distribution, not policy. Who is to say that trend will not continue? (B) those countries have much higher tax rates (and much higher benefits and replacement rates), so the argument about the max is a bit of a canard. You would never accept the higher rates they pay, anyway.

Third, the system as a whole is progressive mainly because of the disability. Faster improvements in mortality accrue to those with higher incomes, as several recent reports have shown.

On another topic, I think Obama's plan should be more open to increasing retirement ages. We need more open and honest discussion about this issue.

As for the other candidates, Mrs. Clinton suggested there were more progressive ways to collect revenue than raising the cap, presuming her commission lets her. What was she talking about? Perhaps non payroll tax mechanisms?

Suppose McCain *has* to raise more money to fund the system, where would he get it? So far, I am assuming that he would inject more progressivity only though adjustments to benefits. Is this correct?

Marc Goldwein said...


I think the 6% referred to the rough increase in the marginal tax rate for someone currently earning above the cap. Of course the real number is 6.2%, and this does not account for the additional 6.2% paid by the employer (which will ultimately be born by the worker).

So while Goolbee is intentionally making this number look lower than it truly is, in economic terms, I don't think he is suggesting any kind of rate increase in addition to an increase in the taxable maximum.

Andrew G. Biggs said...

I'm working with a very marginal wi-fi signal and lost previous comments, so will have to be quick.

I think two things threw me off:
First, Goolsbee referred to just the employee share, which is misleading since almost every economist thinks that almost all the employer tax is born by the employee.
Second, Obama's plan fixes less than half the 75-year problem even if phased in over 10 years. If done over several decades it will do a lot less. This is pretty underwhelming, given how Obama has stressed the problem and how he says this reform is all he needs to do.


For what it's worth, I'd rather lift the rate than the cap; it's more in keeping with how the system has traditionally worked.

While the cap has fallen from 90% to 84% of total wages, 90% is the highest it's ever been while 84% is basically the historical average. In any case, raising to 90% wouldn't be the worse thing on earth, but it's not at all clear why Obama wants to raise it to 100%. He simply doesn't present much of an argument for it.

Bruce Webb said...

Given the vagueness of Obama's precise plan I find it easier to assume that he is implicitly endorsing that of top three advisor Jeffrey Liebman, the eponymous Liebman-MacGiuineas-Samwick NP SS Reform Plan. No element of what Obama has said to date is inconsistent with LMS, not even his pledge to avoid 'privatization', you can define LMSs PRAs (personal retirement accounts) fairly as not really being 'private' at all. Under LMS both control of the investment and annuitized payout remain under governmental control.

If you Osama SS=LMS you get a 1.5% increase across the board in the rate, a partial lift in the cap for another 1.0% increase in payroll equivalent meaning an increase in both rate and incidence which would make much of the above dispute moot.

Then of course by examining the additional 2.02% of payroll equivalent by accepting benefit cuts and a .68% equivalent in an increase in retirement age under LMS and you end up with a result that delivers results of between 89% (I think) and 129% of the current schedule depending on variables including income level, one or two income, and state of survivorship.

Given that Obama has not explicitly rejected Liebman-MacGuineas-Samwick and took Liebman on after drafting of the LMS plan it is not unfair to draw a line between the two.

Bruce Webb said...

Well a couple of words dropped out and I misspelled Maya's name (not that we are on a first name basis) but on my reading there is nothing explicit in Obama's platform that would require us to accept that he was proposing a 100% level vs the LMS 90% level or would extend the incidence beyond wage income.

Bruce Webb said...

Under scoring by the Office of the Chief Actuary of SSA a lift in the cap to 90% of wage income translates to a 1% improvement in a payroll gap currently scored at 1.70% of payroll down from 1.92% at the time LMS was drafted. To suggest that it backfills less than 50% of the gap is justified if you use the 2.02% payroll gap of the 2007 Report but is outdated if only fractionally by the 1.95% gap in the 2008.

Oddly enough not every student is napping in the back row.

Andrew G. Biggs said...

Bruce, your numbers are right but characterization is wrong. I said "Obama's plan fixes less than half the 75-year problem even if phased in over 10 years." Obama would eliminate the cap, but have an exemption for earnings from $100k to $200-$250k. I modeled the effects of exempting to $200k, phased in over 10 years. That fixed less than half the 75-year shortfall. The reason is that there's a big concentration of earnings between $100-$200k.

If the exemption is increased to $250k, as Obama hinted in the Philly debate, and phased in over several decades, as Goolsbee hinted in the interview, my guess is that his plan will fill about 1/3 of the hole.