The Washington PostDemocracy Dies in Darkness

Weekend interview: Peter Diamond on Social Security, privatization proposals and the grand bargain he’d like to see

Peter Diamond is an Institute professor emeritus of economics at the Massachusetts Institute of Technology, where he taught from 1966 to 2011. In 2010, he was nominated to be a governor of the Federal Reserve, but withdrew his nomination after he drew Republican opposition in the Senate. He shared the 2010 Nobel prize in economics with Dale Mortensen and Christopher Pissarides for their work on "search costs" in labor markets — or the frictions that arise when employers need to find new workers,and vice versa.

He is an expert on public finance, especially Social Security, and has consulted for various government agencies on the program throughout the years. With Peter Orszag, he coauthored Saving Social Security: A Balanced Approach, which proposed shoring up the program through gradual, progressive benefit cuts and increases in the payroll tax rate. We talked yesterday about Mitt Romney and Paul Ryan's current and past proposals regarding the program, and where Diamond thinks we ought to go from here. A lightly edited transcript follows.

What do you make of the Romney campaign’s proposal to change the Social Security benefit formula for high earners?

It’s an adjustment in the benefit formula for when you first claim benefits - it’s not the cost of living adjustment after you’re retired. Currently the formula for initial benefits (when you start being eligible for benefits) goes up with the average wage index. Progressive price indexing (PPI) would substitute a mix of wage and price growth, depending on where you are in the earnings distribution. Romney hasn’t specified how many people are affected, but other plans using PPI lower benefits for 60 percent of workers, not just very high earners.

Do you think there’s a responsible way to make a change like that?

As an isolated question, there’s no good answer. As I said in the Bloomberg View op-ed that Peter Orszag and I did on the Romney plan, the central question is how much you do with benefit cuts, and how much do you do with more revenue. Once you’ve sorted out an appropriate balance, the benefit cut portion could be PPI or an increase in the age for full benefits (Romney’s plan has both) or some other change. — By the way, it drives me crazy when people talk about “raising the retirement age.” Sometimes they mean the age for full benefits, which used to be 65 and sometimes they mean the age when you can first take benefits, which is 62, and sometimes they mean both.

As a particular way of doing a portion of benefit decreases, neither PPI nor increasing the age for full benefits is my favorite but neither is horrific. The central point is, are you doing it all on benefit cuts or are you doing part with increased revenue? What Peter and I wrote on Bloomberg view was to describe the plan that was put forward by Senators Graham and Paul and Lee, which has exactly those two ingredients that Romney includes. But, each of the ingredients has parameters that can be tweaked so that the balance between the two could be anything you want, and Romney has not specified the details. The Senators’ plan, which the Office of the Actuary evaluated, did three months a year of increase in the age for full benefits and then drops the rate of increase to one-half a month a year. You could start at two, you could drop at a different rate, etc. They run the increases rapidly for a while and then they slow it down. They tweak the two forms of benefit cut to come scantly ahead for actuarial balance for 75 years.

What’s really interesting, to my mind, on this proposal is despite the fact that we’re an aging society, compared to today, they actually reduce the fraction of GDP spent on Social Security. I find that stunning – a greater need but decreased spending relative to GDP. Again, you see that if you go to the Office of the Actuary website and look at the figures relating to their plan. It’s a bit less than half a percent of GDP decrease at a time when we’re aging.

Whereas if you kept benefits the same you’d expect the opposite.

Right. You’d have the benefit level keeping pace and then the aging would require you to be increasing spending.

What do you make of the criticism that raising the age for full Social Security benefits would disproportionately hit those at the bottom of the income scale, who haven’t seen life expectancy rise as much?

It’s a mostly a red herring. Let me elaborate. The way I think about increasing the age for full benefits is that it has nothing to do with retirement. Maybe some people will retire later because it’s a signal in some form, but I’m skeptical that that matters a whole lot. People are thinking about different ages to retire and for any age that you retire there’s a smaller benefit if the age for full benefits is higher. So increasing the age for full benefits reduces the benefit level at any age at which you retire. You think of retiring at 62 now, you get an actuarial reduction of 25 percent from the current full benefit age of 66. If the age for full benefits is 70 the benefit is lower – a 45 percent reduction. Instead of 75 percent of full benefits, you get 55 percent. So it’s a particular way of cutting benefits for people retiring at different ages.

What I don’t like about it is that while it’s fairly close to a uniform across-the-board cut the way benefit reductions for early benefit claiming is structured, it isn’t exactly that and if you look at it, it’s a bigger benefit cut to people retiring at 62.

What do we know about the people who retire at 62? On average, shorter life expectancy and lower earnings than people retiring at later ages. If anyone stood up and said, “Instead of doing uniform across the board cuts, let’s make them a little worse for people who have shorter life expectancies and lower earnings,” they’d be laughed at. Anything that reduces benefits is going to hurt everybody. It’s going to hit people with short life expectancies, it’s going to hit people with high life expectancies. But we should not make it worse for those retiring earliest.

It’s interesting, the issue of whether phrasing it as a benefit cut or a retirement age increase plays better. And I don’t know, I’m not a politician, but back in 83 there was a clear sense that increasing the age for full benefits  was thought to play better.

Now, if you change the age for initial benefits, age 62, that has a differential impact because the people who can’t claim at 62 and would have done so have lost a year of benefits, and that’s a year relative to what their life expectancy is. That hits people particularly with short life expectancies.

Do policies like the payroll tax cut concern you, in terms of financing Social Security?

No. That required the Treasury to provide the revenue as if the cut wasn’t there. It’s a particular way of cutting general revenues. It has nothing to do with the actuarial balance.

Right, but as a precedent does it concern you, or do you think any future cuts will happen the same way?

It hasn’t worried me. I don’t know that anyone posed the question to me before. It has a lot of appeal as a temporary tax cut approach. So I don’t know that it really worries me.

Ryan seemed to back off privatization in the debate…

No, he has just said Romney’s the president and sets the policy. I wouldn’t call that backing off privatization.

Fair enough. But looking back at his privatization proposals, one potential problem that arises is that many more people would be buying stocks, which could increase their prices and cause all kinds of disruptions. Does that worry you?

I’ve always focused on the other side of it. Basically what we’re talking about is individual accounts that come out of existing payroll tax revenues. What’s happening is that the federal government is losing revenues, and to cover that there’ll be a lot more federal borrowing. They’re basically lending money that is borrowed from the public to individuals who then turn around and buy stocks and corporate bonds with it. I refer to it as lending because in all of these plans you give up some of your future benefits.

The federal government has a cash flow problem, and you worry about the debt held by the public because it needs to be rolled over unlike debt held by Social Security. It is the need to roll over debt which leads to the sort of thing we’re seeing in Greece or Spain (though I don’t think that’s likely in the US anytime soon), it adds to that problem. I worry about the debt held by the public, which is more salient and makes the situation more risky than the question you raise, which is to what extent it raises stock prices and so lowers returns.

Remember two things. It’s a slowly phased in operation. There are people who will be buying the additional federal debt, and in part this would be coming out of their buying less in stocks. The first pass cut at it is if we thought about this in a representative agent way, where the representative agent integrates the portfolio inside and outside, it’d be a wash. We don’t live in that world, obviously. At least half of the payroll taxpayers have no stocks, they don’t buy stocks. So that’s new demand. I think the portfolio change by people who do hold stocks would be a large part of the response.

I haven’t thought of the rate of return issue being a big one. I think the rate of returns on stocks is driven by the rate of return on capital and that depends on the rate of aggregate savings. Maybe you get a little lower return on capital. I don’t see that as a big effect. And if it comes from higher investment, that is valuable.

But let’s circle back to the Ryan plans. The Ryan-Sununu plan was in 2005. Ryan did a new plan all by himself in 2010. The borrowing from the public was still there but greatly reduced because of changes in the plan. The Ryan-Sununu plan’s impact on borrowing from the public was astonishing. There’s a Center on Budget and Policy Priorities study of all the different plans around in 2005, and it asked the question, “How much does it impact debt held by the public in 2050.”

And the Ryan-Sununu plan increased the debt held by the public by more than 90 percent of GDP. All of the individual account plans that were out there at the time required increases in the debt held by the public. It was only my plan with Peter Orszag, and Bob Ball’s plan that decreased it. And of course we didn’t have individual accounts.

I just don’t see him as moving away from privatization — he said in the debate he still likes individual accounts. What fascinated me when I looked recently at Ryan’s 2010 plan, which I’d never done before, is that he wants to raise payroll taxes. But he does not increase the maximum subject to the payroll tax. And he doesn’t change the payroll tax rate. What he does is change the base that the payroll tax affects by including what’s now excluded from the payroll tax relating to health insurance benefits. It’s both what your employer spends for you and what’s currently not taxable when you spend it yourself through your employer. So he’s increasing the tax base, and he’s increasing revenues by a significant amount. 1.13 percent of payroll’s worth of increased revenues over 75 years.

But the tax increase only applies up to the Social Security cap. So if you’re earning less than $110,000 and get your health insurance through your employer, your payroll tax payment goes up. If you earn more than $110,000, your payroll tax doesn’t go up. So this is a tax increase only for the bottom 94 percent, not the top six percent.

And it isn’t a tax increase if you get your money from capital rather than wages.

Right, but that would be true if you raised the payroll tax rate as well. Within the context of the traditional social security revenue stream, this Ryan plan is an increase in taxes for lower earners but not for higher earners.

Another potential issue with these privatization plans is that most don’t let you pick and choose the stocks you invest in. You can pick from a menu of portfolios that the Social Security Administration puts together. Doesn’t that give administrators a lot of power over the stock market?

Yes, potentially. But the basic idea is to imitate what’s done for federal civil servants. And so it’s broad index funds, and as such there’s very little scope for the organizers to tinker with it.

So it’s not like you’re setting up the world’s biggest sovereign wealth fund.

Exactly. When that first started for civil servants, it came back in along with the 1983 Social Security reform and kicked in in 1984. There were only three funds. There was, I think it was an S&P 500 index fund, a corporate bond fund and a fund that just invested in US Treasuries. That was it. They’ve expanded it now. You can also buy some foreign stocks and the domestic stocks that are not in the S&P 500, and you can have a lifecycle mix of these. The idea here is this is not managed money. Individuals still can adjust the stock-bond risk to adjust for their degree of. But it’s all done very broadly, not picking individual companies.

The idea here is really to hold down, two things, to hold down the costs of running the system, and the second idea is to try to protect people from doing really dumb things.

Last summer you wrote a piece for The New Republic calling for a grand bargain in which you pair long-term deficit reduction, in the form of fixing Social Security, with action to address the jobs crisis. Whoever’s elected, it seems probable that a debt deal of some kind will be enacted next year. Are you still as optimistic about the potential for a deal like that?

[Laughs] I don’t know that I ever was optimistic. Let me frame this the way I think about these things. We have an unemployment crisis, in my view. The impacts, both on long-term unemployed and young people, are going to affect them for years and years. This is something we need to be addressing right away.

We have a debt problem. We don’t have a debt crisis. If you look at CBO projections, and ask “At what stage in the debt to GDP ratio do you think the debt market might start to look for a sovereign risk premium?” I think we have a decade at least, perhaps two before there’s even a hint of that problem. What we should be doing now is something that improves economic growth, and there’s a list of the usual suspects, and addressing the debt part should come into force slowly, although legislated soon. I’m obviously not alone in this. The same perspective is coming from Ben Bernanke and the IMF.

So what can we do in terms of debt concerns that will have credibility that they really will happen, and represent good policy, not bad policies? Across the board spending cuts make bad policy. The answer to my mind is, let’s just fix Social Security because without individual accounts it has a big impact on debt held by the public. For example, my reform proposal with Peter Orszag knocked 20 percent of GDP from debt held by the public 45 years on. And restoring actuarial balance is something we need. As every Social Security analyst will tell you, the sooner you restore actuarial balance, the easier and fairer it is to do.

Generally the political process has treated Social Security by itself and not as an integrated part of the budget process. In the 1990s, under Clinton, there was a budget rule that you couldn’t use Social Security to justify changes elsewhere in the budget. I really don’t like the idea that we might change the COLA for budgetary reasons. I think the analysis should focus on a whole package for Social Security. The countries that end up integrating their Social Security systems with their annual budget processes make a mess of it. We haven’t gotten into that kind of game.

So I dislike integrating Social Security as part of a grand bargain unless you carve out a Social Security reform that restores actuarial balance, and does what has traditionally been our political approach.

I can’t think of a better thing to be doing than Social Security precisely over its credibility. If we pass a fix, a bipartisan fix, it’s not going to be undone. Most of the other things that are on the table will stay on the table down the road and do not have the same credibility. It’s the ideal thing to be doing.

Update: This post has been amended at Diamond's request to clarify some statements. None of the updates change the substance of what was said.