Recently in Social Security Category

Andrew Samwick informs us of the Treasury Department's series of issue briefs on Social Security reform. The first one is already up on the Treasury website.

He is pleased to see that the Treasury issue brief points out that:

Delay reduces the options for distributing the financial burden of reform across generations because delay exempts additional generations from sharing in the financial consequences of reform.

Samwick said something similar in the early days of his blog (October 2004).

...each year that elapses without reform causes the burden of financing the unfunded obligations to be shifted away from one more birth cohort that crosses the threshold of being "at or near retirement." The more we wait, the larger the burden on future generations...

I couldn't agree more. My comment in Feburary 2005, when talk of reform hit its peak was similar.

The more gradual the transition, the more the transition cost can be spread out. The sooner we start, the more gradual we can afford to be.

Now to be sure, I'm not one of those who thinks the system is in imminent danger. However, the fact that the system is on solid ground for a number of years to come reinforces the idea that any reform can be gradual. If we start now with a cautious and reasoned plan, the result will be better than if we wait another ten years and need to consider something less gradual.

As I put it in my first post on the subject:

Could the right sort of adjustment/modification/reform of Social Security in the next few years make future retired persons better off? Almost without question, yes.

Is Social Security on the verge of becoming the nation's biggest fiscal problem if it's not fixed in this presidential term of office? Definitely no.

...

But I also concede that you can't just take any old privatization scheme off the shelf and say that it will be a Pareto improvement.

I think that still sums up my thoughts on the matter.

The Social Security Board of Trustees Annual Report is on the web. Have at it.

Remember when you were a kid and you couldn't wait to see the odometer in the family car turn over to some multiple of 10,000. Of course in today's cars the odometers are digital, so it's about as exciting as seeing your digital watch tick past midnight. Those old odometers looked like they really had to work to turn that last digit.

And so it is with the U.S. population clock which, as I write this, stands at 299,998,288. You can check the current number for yourself by going to the Census Bureau web site. In the morning, the big three-oh-oh (million) will have been reached. The Census Bureau's clock, like my odometer, is digital. The addition, be it by birth or by immigration, will arrive seemingly effortlessly. I think that is a good analogy. We're not running out of room in this country by any stretch of the imagination. Our society and economy will absorb the 300,000,000th person as readily as the 299,999,999th. And by 2050, we'll be adding number 400,000,000--a fact which causes Joel Kotkin to opine in the Wall Street Journal:

Unless there is some sort of cultural revolution, most people, particularly families, are likely to continue migrating to places where they can acquire a spot of land and a little privacy. And despite the much ballyhooed "return to the city" by aging boomers, most experts suggest that most are either staying in the suburbs or moving to towns farther out in the hinterland. At least 30% of Americans, according to surveys by the National Association of Realtors and the Fannie Mae Foundation, express the desire to move to the country or a small environment, far more than live there now. The scale of this dispersion depends largely on urban governance. If cities cannot, due to economic or regulatory constraints, provide sufficient job opportunities, people and businesses naturally will flee elsewhere. Other factors, such as preserving family-friendly neighborhoods and stamping out a nascent resurgence in crime, will also be critical.

Yes, we will find room for number 400,000,000 too. There is room for a few of you out here. Sorry, no "for sale" signs on my block. The few that were available this summer have long since been sold. Unlike some places, the real estate market here seems to be approximately in equilibrium. Growth is proceeding sensibly. I still have to get used to the sight of a new apartment complex a mile or two south of us. More room for the next hundred million Americans.

So rejoice at this milestone. Malthus was wrong. We're not doomed by population growth. While a growing country has always presented certain challenges, it has been our innovative responses to those challenges that have made this country what it is. And it's a reason that people keep coming.

To the 300,000,000th American: Welcome! We're glad you're here.

First the nonsense. Mark Thoma quotes a Financial Times piece by Rep. Jim Kolbe. For the sake of brevity, the last line from Kolbe says it all.

Far more people would confront the need to rein in entitlements if they understood how they are putting our foreign aid budget in a straitjacket.

By entitlements, Kolbe is speaking broadly to include Social Security, Medicare, etc. To which Mark responds,

I don't have time to deal with this properly, so hopefully comments or other bloggers can put this into it's proper place, but the amount of foreign aid we give relative to the size of the budget, spending on the war, and so on, is miniscule, low among developed countries on a per capita basis (this says 27 billion in 2004).

Indeed. When I read Mark's post, I couldn't help but wonder just how I would properly deal with such a statement. Foreign aid is often smaller than the forecast error in the budget. True, foreign aid would be one among many programs fighting it out for the remaining funds if there were broad cuts in discretionary programs. But my assessment of the situation is that there is a substantial enough lobby to prevent any serious reduction in aid. There are other, larger, more promising targets for the budget axe. Now, might there be some political reason for setting up a false tradeoff between foreign aid and Social Security? Of course. One doesn't have to be all that cynical to see that. There are always political reasons for framing the rhetoric of budget priorities. So take that and see where you end up.

By the way, I'm no huge fan of the current system of foreign aid. What we do spend is often used less than effectively. That doesn't change my opinion of Kolbe's odd comparison.

So that's the nonsense. PGL at Angry Bear chimes in as well, and in a comment to that post he offers up some sense about Social Security. For context, the comments at AB went in the direction of the "trust fund" as often happens when Social Security is the topic. PGL writes,

On this no trust fund argument, consider this analogy. A 53-year old worker has accumulated funds in a private retirement account to which he'll continue to add savings draw from his salry for the next 12 years. In 2018 he retires and wishes to draw funds from his account from 2019 onwards. Imagine his reaction if he were told that his account weren't really there. That's the rightwing argument. Something tells me that this 53-year old would not accept it in the least.

That is a nice illustration to make the point that the trust fund represents a commitment to play by the rules of the game (but see below). And that is precisely how I see the trust fund. It is an accounting device that represents the promise to follow the law as currently written. The accounting device (whether a ledger sheet or a file cabinet full of bonds) is not the important thing. The promise is. As I said a year-and-a-half ago:

We need to be careful about what we mean when we discuss (or imply the possibility of) default lest we fall into rhetorical traps. The "worthless IOU" argument is itself worthless. There are undoubtedly other cases where it has been used to get the attention of the reader and then been cast aside when it has served its purpose.

My opinion on that has not changed.

To elaborate briefly on PGL's point, the government wouldn't tell the holder of a private account that his account no longer exists. Rather, they could allow inflation to eat its purchasing power away, or they could raise the taxes on investment income. The rules of the game can be changed in many ways--some ways are easier and/or more visible than others, and that should not be forgotten. But the basic thrust of his comment is on the mark.

And that brings us back full circle to the Kolbe article about "entitlements" in general and the trade-offs that they imply for present and future Congresses. Entitlements are, in effect, promises. People assumed that those promises would be kept when they made decisions such as how much to save for retirement. Just remember that when people plan how much to contribute to their private retirement funds, they are also making assumptions about future tax rates and inflation rates. However, I'm not sure that we treat those assumptions like promises--we know that tax rates and inflation rates are too fluid to project out 30 years. That is a distinction worth remembering.

If you've been following this blog for the last few months, you know that Mark Thoma and I have engaged in a running discussion on Social Security. (PGL of Angry Bear commented on many of those posts, and others commented occasionally.)

And so, I bring you this recent dispatch from Thoma which I quote in part: (It's quite long, and I'm skipping over the part on insurance against poverty, a topic we have both covered-- Thoma here and me here).

A common solution to the moral hazard problem is deductibles and co-payments that cause individual to share in the cost of a negative outcome. Two proposed solutions to market failure in retirement savings markets have this deductible feature. The first is well-known, opt-out, add-on accounts. These tend to work better when the default outcome is very unattractive, and the difference between the outcome with active participation and the outcome with passive participation can be viewed as the deductible the individual pays for behaving in a manner leading to sub-optimal outcomes. The other is a money match program where the government matches, say dollar for dollar but the exact ratio could vary, money put into retirement savings accounts. In designing a system, I would implement a combination such as an opt-out add-on accounts with the amount matched in some ratio by the government (and capped) unless one of my really smart economic advisers convinced me some other program would work better.

The opt-out solution has the problem that short-run considerations may interfere with following the optimal long-run plan, and there is an argument for forced individual saving because of this. A family responding to the pressures of the moment may make decisions that, in retrospect, were not the best in the long-run, and forced saving overcomes the short-run temptations (IRAs have this feature). Forced retirement saving is not my preferred solution, but it needs to be mentioned, and, if the opt-out system appeared to allow sub-optimal behavior to emerge even with the incentives described above in place, I would consider dropping add-on accounts and adding forced retirement saving to the forced poverty insurance payment. That would result in a system resembling the current system in many ways. (emphasis mine)

All in all, it's a good post. In the part I'm not quoting, he has a little to say about how to design the poverty insurance aspect of Social Security. Of course, there are issues that are too much for one blog post, so there are no final answers here. But if you have forgotten the earlier discussion, this will refresh your memory.

In the part quoted above, Mark admits that add-on accounts may not be the optimal solution either. It doesn't totally overcome the incentive problem. He doesn't give forced savings a ringing endoresement, but says "Forced retirement saving is not my preferred solution, but it needs to be mentioned, and, if the opt-out system appeared to allow sub-optimal behavior to emerge even with the incentives described above in place, I would consider dropping add-on accounts and adding forced retirement saving to the forced poverty insurance payment."

First of all, forced retirement saving would (or at least could) look a lot like a system of private accounts much like the privatizers have been talking about. In fact, I should probably say "partial privatizers" since most of us who have advocated private accounts at one time or another have really been arguing for partial privatization. Partial privatization would effectively separate the poverty insurance from the retirement savings. The former could quite likely remain solvent in pay-as-you-go form while the latter becomes a system of "forced retirement savings."

Of course, I am assuming that the forced retirement savings go into an personal account. That would not look much like what we have now, so I have to differ with Mark a little bit, at least on the specifics of how I would implement the "forced savings." (Since he doesn't mention private accounts, I have to admit the possibility that he wants forced saving without private accounts--Mark, if you're reading this, please clarify.) Here's why. If we accept Mark's implication that the forced savings looks much like what we have now (without private accounts), then what is the point? One more way to divide up the FICA tax? Would a certain percentage go to a poverty insurance fund and a certain percentage go to a retirement fund? What sort of rate of return would it yield? If it resembled the rate of return the average person receives on their OASI contribution, there would be open revolt! You'd have people begging for private accounts so fast it would make your head spin.

Perhaps Mark would advocate funding the program by letting the government invest in other assets (i.e. stocks) rather than just government bonds--in effect, fully funding the retirement savings but not giving individuals private accounts (and control over the risk/expected return trade-off). He mentions this elsewhere in his post. But this has been suggested before and it went nowhere.

Honestly, I don't see how you sell a forced savings program without private accounts. My return from a forced savings program without private accounts is going to be determined by politics rather than the market.

On the other hand, partial privatization is really just forced savings with private accounts (and choice over your risk/expected return trade-off). At the risk of sounding like a broken record, I think that could be done quite effectively if it were phased in with younger workers first.

That said, Mark's post is a step in the right direction. He and I definitely agree that it would be better to separate the insurance from the retirement savings. The details are still left to be debated. I prefer the "forced saving" with private accounts over the add-on opt-out accounts. Whether I could reluctantly support the latter is something I'm currently pondering. Government matching money for retirement savings is worth exploring (Mark mentions this in his post as well).

The debate continues. The real question is whether any progress will be made in Washington this year. If the second half of the year is like the first, I doubt it.

UPDATE: Has Daniel Altman been reading our blogs?

With apologies to Brad DeLong, I think this deserves a "Why, oh why..." sort of intro.

Here's the Washington Post article about DeMint's proposal for private accounts. They catch on to the flaw.

Key Republican lawmakers, scrambling to keep President Bush's Social Security proposals afloat, plan next week to embrace an idea that many have avoided thus far: funding personal retirement accounts with surplus revenue that now pays for other government programs.

There's a reason they avoided it. It's a bad idea.

The strategy is controversial because it would create new budget problems. Either the diverted money would have to be replaced with new taxes, or Congress would have to slash programs now funded by Social Security's excess payroll taxes.

and...

Aides close to Graham, Santorum and DeMint said their proposal will not preclude broader initiatives, including several that have been floating for months. "It is not a comprehensive reform," one aide said. "It is sort of a first step toward gaining momentum." A Graham aide said the senator still thinks Congress must trim benefits and raise taxes to safeguard Social Security's future.

Another GOP staffer familiar with Tuesday's scheduled event said the participants will acknowledge that redirecting surplus Social Security funds will create budget problems elsewhere.

"No one is going to find a magic bullet," the staffer said. "There's going to be real pain at some point."

Among other things, the proposal would be horribly confusing to implement. Are they proposing to divide up whatever the Social Security surplus turns out to be in a given year among millions of private accounts for people? And these accounts just hold government bonds? The surplus is set to shrink, you know. So over time, the additions to the private accounts shrinks? At least Al Gore's proposal was more practical. This one just seems odd to me. The only thing I can possibly come up with is that they figure that they will add-on or carve-out more contributions for the private accounts later. But if that isn't spelled out at the beginning, it could lead to trouble down the road. Do you raise payroll taxes down the road to pay for low income taxes today?

If they want a "lockbox" as a commitment mechanism, that's one thing. But the addition of private accounts would be needlessly costly added to a plan like this. It doesn't make any sense. There are better ways. It seems like they are trying to blend Gore's lockbox with a minimal sort of private account. The result would be more costly and confusing than either Gore's lockbox or a straight-up diversion of some percentage of the payroll tax to private accounts.

I like the principle of private accounts--I really do. But when something like this comes out of left field ("right" field?), it seems to indicate that there is not a well thought out strategy here. DeMint has this plan, Bennett has another, there are already House and Senate bills (imperfect as they are) languishing in limbo.

Socialsecuritychoice.org is already spinning this as free lunch. They don't use those words, but they also don't acknowledge the costs. Not acknowledging the costs constitutes "spin" from my point of view.

Mark Thoma is disgusted with it too, as is PGL (Angry Bear). Thoma goes to great lengths to smack it down, which is really unnecessary. "No free lunch" says it all. The rest is commentary.

UPDATE: It's worse than I thought. (h/t Mark Thoma)

Probably. Today's NY Times editorializes thusly:

Rather than accept defeat and consider alternatives, Mr. Bush is becoming even more feckless as public and political opposition mounts. On Tuesday, in a lame ploy to draw the Democrats to the table, he gave tepid approval to a proposal by Robert Bennett, the stalwart conservative senator from Utah, to restore the system's solvency in a way that would not include private accounts - all the while saying that he was not prepared to give up private accounts.

Let me interject to say that I too thought that the "tepid approval" of Bennett's plan together with the continued push for private accounts was weird to say the least. I wonder what Mr. Bennett makes of it.

Mr. Bennett's plan includes drastic and unnecessarily large cuts in Social Security benefits, but at least he is being straightforward in offering a plan that addresses the real problem Americans want solved. A group of four Republican representatives have meanwhile offered a proposal that would, in effect, abandon efforts to restore solvency in order to resuscitate those doomed, unwanted, unwise private accounts.

Enough is enough. Mr. Bush must either put forth a complete plan - including details of the risks, benefit cuts and borrowing costs that privatization would entail - or abandon his quest. Anything other than that is wasting his own and, by extension, the American people's time.

It's getting harder and harder for me to imagine a world in which private accounts become part of Social Security before the end of this presidential term of office. If you've been following the blog for a while, you know that I've been supportive of the general idea of private accounts (see here, for example) but not at all bashful about criticizing various proposals that have been floated (see here and here, for example). As the discussion evolved, I came to see Social Security as a very good income insurance program and a lousy retirement investment. The real questions are how much income insurance is optimal and how progressive it should be. I became frustrated with the notion of "rate of return" as applied to Social Security (see the comments to this post). To the extent that Social Security is insurance, the term "rate of return" is awkward to say the least. Because Social Security is partly (mostly?) insurance, any calculated rate of return to the average participant is sure to be dominated by the return to TIPS--probably even dominated by the return on passbook savings accounts. So many people would be better off with private accounts, but if that comes at the expense of the social insurance, it's not an unambigously good thing.

All the while, President Bush kept talking about IOUs and file cabinets. In fact, he did it again on Wednesday. Ugh. It's like the movie Groundhog Day. Same thing over and over again. Even for a supporter of private accounts it's getting old.

And it's not working. Republicans are distancing themselves from private accounts (e.g. Bennett). The polls don't look good for private accounts. Midterm elections will be coming up quickly before anything can get passed. Even Andrew Samwick is losing hope. After testifying before Congress, he notes,

It is not clear that we accomplished much through this hearing. I got no indication from the Representatives that they would be looking to get a bipartisan bill through their Committee. I hope that I'm wrong.

I'm afraid he's probably right.

A few years ago (2001 to be exact), I put the number 2008 on the whiteboard in my office. When people asked about it, I said that's the next time that Social Security reform will be seriously debated. At the start of this year, I feared that I would turn out to be wrong. Well, I was wrong if you're talking about the blogosphere, but if we're talking about any serious chance of anything really happening in the halls of congress I still have a chance. It will only happen if a Democratic candidate engages by endorsing something like Brad DeLong's idea, (not Al Gore's "lockbox").

Not that I would mind having either or both candidates in 2008 touting some version of a plan to increase national (and personal) savings. I think private accounts, if done right, would work. But it is nearly impossible to imagine a world in which politicians could come together to do it right. The last 6 months have been very discouraging in that regard. Doing it right means phasing in private accounts (and the carve-outs) very slowly with younger workers. Since younger workers don't vote in large numbers, what appears to me to be the sensible way to do it economically ends up being political suicide.

It is with much disappointment that I say this. I have lost hope that a Social Security reform plan that is anything close to optimal will be politically acceptable, and I won't throw my support behind plans that are flawed. If that means no private accounts because the demographic who would benefit from them doesn't vote, then so be it. It's a miserable conclusion if there ever was one, but I fear it is so. Time to look for a second best solution for 2008. Enterprising Republicans and Democrats seeking the White House should get to work on it now.

Most of you probably have heard most of what is in this article before if you've been following the debate. But in case you haven't...

Beware the Easy Fix for Social Security (NY Times)

First, I don't know if "Jane Galt" read my post on Social Security as insurance, but she comes to a similar conclusion.

What's the best solution, then? I'd say we're converging on it: a system of minimal government insurance for those who have been unlucky, in life or investments, combined with a regulated forced savings plan to make sure that those who aren't unlucky aren't tempted to free-ride on society, and incentives to employers to encourage additional savings among employees. This won't make anyone ideologically happy. But it seems like the least intrusive, most fair, most economically sound possibility.

Your ideological happiness (or lack thereof), I'd say, rides on how minimal the insurance. Hence, Mark Thoma writes in a comment to my post:

I agree entirely and have noted before that the question of whether a person views the current system as just insurance or insurance plus savings depends upon their (normative) view of the level of support that ought to be insured, the minimum acceptable level of societal support. Much of the argument you hear is needless because it is really an argument over this difference and there is no right answer.

Yup.

So now, Winterspeak asks (also in my comments section):

So what risk are we left with -- the risk that the retiree has not saved enough before he retired? I don't see how this is a risk. The retiree 1) chose when to retire and 2) saved throughout his life. This is not a risk, this is a bad choice.

Not to mention the bad choice of leaving all your retirement in the stock market too long.

I am supportive of (and understand) forced savings so ppl don't make the bad choice I've detailed above. But I cannot see the insurance aspect so social security. I don't understand why insurance against unexpected poverty should be limited to old people.

One crucial reason for a different sort of program to insure against old age poverty is that most of us are of a sensibility that makes us uncomfortable making senior citizens work to pull themselves out of poverty. For younger people, we have unemployment insurance and TANF, both of which have limits which induce young people to get back into the workforce. I think there is good reason for something different for those over a pre-defined, socially acceptable age.

As for the distinction between risk and a bad choice, I'd offer the following. Too much risk is sometimes the result of a "bad choice." In a free society, some people will make what I would consider to be bad choices and take on too much risk. Some will lose. Yet, I am uncomfortable watching them be reduced to poverty (when they have less of an opportunity to pull themselves out). So I like the idea of a safety net. But I also understand moral hazard, so I know that a safety net that is too generous will just encourage folks to make bad choices (take on too much risk, let it ride on the stock market too long, etc.).

Hence, the problem simply boils down to how generous we want that safety net to be. That is an appropriate question for society to consider through the political system. (And a question on which economists can and should offer sound theory and evidence.) Or, as Mark says (in a comment to my post),

You can argue anywhere from an absolute bare subsistence level to a fairly generous level of support, but there is no correct answer as to what is fair that we will all agree upon other than that a floor exists. And I suspect there are a few who would even argue against a minimal level of support.

And while we're at it, let's consider another source of risk that could bite a lot of people in the next few decades--the demise of the corporate pension system. I would be remiss if I didn't mention this:

Unions Threaten to Strike After Pension Default at United

This is the tip of the iceberg. There is also the more subtle pension shaving due to reductions in health benefits for retirees in a number of companies. Did the workers in these situations make a bad decision? The fact is, corporations may no longer be the appropriate vehicle for bearing this sort of risk.

Back to "Jane":

The defined benefit corporate pension plan has been, for a long time, the holy grail of liberals. It was lavish and safe. It is also dying. Not that it was ever that prevalent in the first place, mind you; liberals who lionize the Golden Days of the fifties and sixties seem to believe that everyone worked for either IBM or GM, when in fact most jobs, just like today, were with small businesses.

But the corporate pension was certainly *more* prevalent. Unfortunately, time has revealed its cracks; companies aren't very good vehicles for managing this sort of risk. Time is the biggest one; pensions require companies to plan over time horizons that span 30 or 40 years. That was fine in the cozy, protected, and highly regulated environment of the 50s and 60s, but when the market changed, the pension promises couldn't. This is what (among other things) is dragging down the major airlines; I expect that within the next decade we will also see Ford and GM default on their pension promises.

She goes on to say that there are problems with having the government provide this insurance as well (which I take it is why she argues for "minimal" government insurance). All of this leading up to her conclusion I quoted at the top of this post.

If we could get broad agreement on a coherent system that provides incentives for saving, together with an open and honest debate about the appropriate level of social insurance, that would be ideal.

A couple days ago I said,

And what, you ask, do I think the real debate is (or should be)?

I'll have some thoughts on that later. It's probably not what you think I'm going to say.

Social Security benefits are progressive--benefitting lower income workers with higher replacement rates. The system is widely touted as Insurance against poverty in Old Age or in the case where a spouse or dependents Survive the main breadwinner. OASI=Old Age and Survivor's Insurance.

However, benefits are at least marginally tied to your lifetime wage profile--like a defined contribution plan or a forced savings program. So what is it? Insurance or forced savings? Brad DeLong and Mark Thoma emphasize the insurance aspect. Actually, they take particular objection to those who call it welfare. Most assuredly, Social Security is not welfare, so we've got no argument there. Angry Bear comes closer to a direct comparison of the insurance vs. savings question. He says,

...all of the privatization plans I've seen discussed would replace Social Security with a defined contribution style system in which the more a person makes, and the luckier that person is, the more that person will have at retirement, and vice-versa. The insurance function of pooling and attenuating risk is totally removed; or, more accurately, reversed.

So I think it is useful to think of a Social Security program like ours as having an insurance component to prevent poverty and indexed benefits as a political appeasement to those who are less likely to need the insurance.

Where do you draw that line? What is the appropriate level of insurance? Can you have private accounts (add-on or carve-out) and maintain the insurance aspect? How do we get from where we are to where we want to be?

This is the discussion that we aren't having, but should have. It doesn't matter where you stand on trust funds or private accounts to discuss the appropriate level of social insurance. If we could get a handle on that question, then we could make progress on the rest of it. The consequences of this question run deep. If you believe that the insurance portion of the system should be enhanced above what it is now, then private accounts are a non-starter. If you want to allow people to self-insure more than they do now, then you would likely think that private accounts are a vehicle for them to choose their level of risk. This is a normative question. This is a matter of your view on how much risk is acceptable in a person's lifetime consumption profile, how replacement rates should vary with lifetime income, etc.

Trust funds are a distraction when you look at the problem this way. All these posts about the trust fund are arguing over minutia--how real are the assets, how shortsighted the credit market is for not recognizing that the trust fund has been raided, and so on. The bonds are real. The promise to pay benefits will be kept (subject to modifications by Congress, but the modifications are not likely to be severe). The trust fund is being used by both sides as a rhetorical device to advance their point of view by quibbling over the value of a bond. These matters, while interesting, are nothing compared to a frank discussion of the role of Social Security as insurance and the appropriateness of having the government run a pension program whether it's with private accounts or not. This discussion would not take place in a vacuum, but would necessarily take place in the context of the overall tax and spending environment at the macro (government finance) and micro (overall tax progressivity and the effect on households) levels.

Debating the progressive price indexing of the Pozen plan is a start. Brad DeLong's piece in Slate is a good opening salvo. In it, he implicitly acknowledges the political reasons for less draconian cuts.

Pozen's proposal caps the maximum Social Security retirement benefit at roughly $22,500 dollars a year (adjusted for inflation). Bush's private-accounts plan—which would allow people to contribute 4 percent of their wages—makes retirees repay the taxes they diverted into private accounts out of their standard Social Security benefit. Medicare premiums are already deducted from your Social Security check. Deduct the claw-back for the private-accounts diversion as well, and by late in this century the odds are that—at least for the upper middle class—the standard Social Security check would be zero. Social Security would no longer be a universal program: It would be a program in which the half of America that is richer and more powerful and more likely to vote sees large chunks of its money going in and nothing coming out. (emphasis mine)

Appeasement for those less likely to need insurance might be a political necessity to keep the program viable. This has always been a standard argument against means testing.

I say again, this is the right discussion to have.

I, for one, would approach such a discussion with an open mind.

Let's look at what the President said in his news conference.

We're saying, you ought to have the right to set up a personal saving account so you can earn a better rate of return on your own money than the government can.

And it's that difference between the rate of return, between what the government gets on your money and what a conservative mix of bonds and stocks can get on your money that will make an enormous difference, and a person being able to build his or her own nest egg that the government cannot spend.

Now, it's very important for our fellow citizens to understand there is not a bank account here in Washington, D.C., where we take your payroll taxes and hold it for you and then give it back to you when you retire. Our system here is called pay-as-you-go. You pay into the system through your payroll taxes, and the government spends it. It spends the money on the current retirees, and with the money left over, it funds other government programs. And all that's left behind is file cabinets full of IOUs. (emphasis mine)

What's wrong with the italicized part? If you said that what the government "gets" on your money is irrelevant, you're a winner. Why did he say it this way? No clue. Rhetorical flourish maybe. Maybe he thinks it sounds good to put the idea in people's minds that individuals can do better than the government. Sounds nice, but it's both irrelevant and technically wrong.

Hence PGL at Angry Bear writes:

So if I own a Treasury bond, that is a real asset but it is not a real asset when a Trust Fund owns a Treasury bond. And suppose I recently purchased a bond paying a 4.8% nominal return. Bush is saying this is somehow a better return than the Trust Fund gets – even though its bond also pays a nominal return equal to 4.8%.

Good catch, my friend. That is what Bush appears to be saying. And you're also correct to point out that it can't possibly be right.

But it's also irrelevant. Why? You could set the nominal interest on those non-traded bonds to be anything you want. It has no impact on the benefits that retired people will receive (a point I made here). The interest on the bonds does not figure into the benefit formula for an individual. Higher interest on the bonds would just mean that the general fund has to pony up more cash when the bonds come due--restoring "solvency" at the stroke of a pen by making it explicitly a general fund problem. It's irrelevant to the discussion of private accounts!

To the individual, it matters not "what the government gets" but "what the government gives" (the benefit formula).

So, PGL calls out Don Luskin for saying

Younger workers should have the option of putting a portion of their payroll taxes into a voluntary personal account which will allow them to build a nest egg that belongs to them. This money will give workers an opportunity to receive a higher rate of return than the current Social Security System can provide. (emphasis mine)

But to be fair, this is really just a paraphrase of the Moynihan-Parson's Commission (2001, p.38).

The river of lousy rhetoric runs deep indeed. Let's get it right. The pro-privatization side is doing a terrible job of explaining itself and it plays right into the hands of the opposition. It's a shame, because this just bogs down the real debate.

And what, you ask, do I think the real debate is (or should be)?

I'll have some thoughts on that later. It's probably not what you think I'm going to say.

Continued from the last post.

Question #5: Would privatization be a free lunch? (See numerous posts at Angry Bear) Economists are trained to be very skeptical of free lunch claims. It's hard to answer yes to that question. However, if it were reworded: are the opportunities for some small efficiency gains? That asks about the same thing but sounds better. (Ok, ok, I'll quit being evasive.) Not much of a free lunch--not to the extent that some claim. Some efficiency gains might be possible, mostly for the very young. This has a lot to do with how people's portfolios will change if there was privatization. It also has to do with Ricardian Equivalence. Here's what I said in January. I was responding to PGL of Angry Bear:

PGL writes that if people already have balanced portfolios, including the bonds implicit in the Trust Fund and their 401(k)s and they suddenly receive the Trust Fund, they will not buy stocks. The implication is that they would buy bonds, effectively replacing the bonds in the Trust Fund and keeping their risk/return ratio the same.

Our logic is pretty much the same. If Social Security was fully funded, I would absolutely 100% agree. But it's not. So I don't. Not 100% anyway. I have to ask myself this question. Suppose the government gave me a $1000 refund on my FICA tax for this year and offered me two options. 1) Give it back to Social Security and forget the whole thing ever happened or 2) buy a $1000 government bond (TIPS, so I don't have to worry about inflation) and keep rolling it over until I retire.

I choose door number 2. In a heartbeat. In other words, I don't think that the bonds implicit in the Trust Fund are the same as a bond in my hand (or my private account). If I did think that they were the same, I wouldn't care--option 1 and option 2 would be equivalent. (UPDATE/CLARIFICATION: It's not by a wide margin that I choose door number 2, but I would choose it. The implication that has for PGL's comment is that while I don't agree 100%, I think he's close. Close enough that there should be more discussion on this point.)

Another reason I'd choose door number 2 that I didn't mention is that I would just like to own the bond myself. Once I own those bonds, would I sell some to buy stocks? I would if my portfolio was previously at a boundary--if I had no stocks and was borrowing constrained. I make the same point in these comments.

No large scale free lunches, but some small scale efficiency gains for mostly young workers. The wealthy and the over-40 set would probably be unaffected. In other words, Ricardian Equivalence would be a good first order approximation. Barro and Becker would be right as a first order approximation.

Question #6: Should private accounts be carved out or added on? I think a carve out would be better if it was done right (sensible assumptions about equity growth, less Draconian cuts). It would also be worse if it was done badly (use your imagination). To have any value, an add-on would have to accomplish something that the present array of 401(k)s and IRAs does not. If it encourages young people to save, it would be better than nothing.

Question #7: What is the significance of 2041? I'll be 69 years old. That's about it.

In other words, the date the trust fund is exhausted is a moving target. (Wasn't it just 2042?) If productivity growth stays healthy and other good things happen, that date will be pushed back. It has been pushed back before. But by the same token, it won't be pushed back to infinity. Since initial benefits are indexed to wage growth, it is harder for productivity gains to push that date back than if initial benefits were indexed to price growth. (See my response to the question on whether or not there is a crisis--it's not a crisis, and 2041 isn't set in stone.)

Final point: The more gradual the change to the Social Security system, the better. The sooner we start, the more gradual we can afford to be.

Closing thoughts: A reasoned debate is still possible. My rationale for these posts is to summarize where I've been and collect my thoughts in one place for my own reference (and yours if you're still listening) as we go into the serious debates on this issue over the summer. If you're still with me after these two posts, you can decide for yourself where I am on the ideological spectrum--I've been called all kinds of things. But I endeavor to be honest. I try to be as consistent as I can be. I love a good discussion of economics, both theory and policy. This has been, and promises to continue to be, both enjoyable and enlightening. As I look back, I haven't changed my position much, but discussions with you, my readers, have refined them.

Thanks to everyone with whom I have discussed these issues through e-mail or blog comments. I look forward to more. Have at it.

It was a pretty interesting week for Social Security blogging. See Angry Bear (pgl), macroblog (David Altig), and Economist's View (Mark Thoma) just to name a few. I figured I'd put down some thoughts--sort of in the form of an interview with myself. Some of it is retrospective. Let's start with the basic question.

Question #1: Is Social Security in crisis? No crisis, but we could do better.Here's what I said in January.

Could the right sort of adjustment/modification/reform of Social Security in the next few years make future retired persons better off? Almost without question, yes.

Is Social Security on the verge of becoming the nation's biggest fiscal problem if it's not fixed in this presidential term of office? Definitely no.

No change in my opinion there.

Question #2: Are the bonds in the trust fund worthless IOUs? No. However, I do not have a claim to those bonds other than through my participation in the Social Security system. The payout I will receive when I retire will be determined by the SSA's formula rather than by the return on those bonds. The bonds represent the government's commitment to fund our retirement based on that formula.

I don't like photo-ops with file cabinets, and I'm not crazy about comments like this from President Bush's press conference:

Now, it's very important for our fellow citizens to understand there is not a bank account here in Washington, D.C., where we take your payroll taxes and hold it for you and then give it back to you when you retire. Our system here is called pay-as-you-go. You pay into the system through your payroll taxes, and the government spends it. It spends the money on the current retirees, and with the money left over, it funds other government programs. And all that's left behind is file cabinets full of IOUs.

The reason I don't like that rhetoric is that it downplays the government's commitment to funding the retirement of the next generation. Some would say that it's tantamount to threatening default. You be the judge of that. I'm pretty sure Bush is not actually threatening default (he'll be long out of office by the time it would happen anyway). It's just not the way I would say it. The political and moral commitment cannot be understated. But a political and moral commitment is not a claim to a specific quantity of bonds.

And here's what I said about it once before. (Feb. 21)

We need to be careful about what we mean when we discuss (or imply the possibility of) default lest we fall into rhetorical traps. The "worthless IOU" argument is itself worthless. There are undoubtedly other cases where it has been used to get the attention of the reader and then been cast aside when it has served its purpose.

That characterizes my current sentiment as well. So far so good.

Question #3: Would privatization solve the solvency problem? Nope. Might privatization make a marginal contribution towards achieving solvency? This depends crucially on the assumptions one makes concerning growth of the stock market and how much the traditional defined-benefits can be reduced as a result. Assume a high enough growth in the stock market and it does contribute towards solvency. I don't buy the upper range of those claims. With reasonable assumptions, the contribution is likely to be small.

However, I don't think that this is a reason to reject partial privatization out of hand. If you want to reject privatization, do so on the merits. Judge privatization by what it would do, not by what it would not do.

See also this post.

Question #4: What about wage vs. price indexing? I can only find one post where I addressed this directly. I am cautiously in favor of the switch, as I was when I wrote that post. My biggest concern is that expectations matter, and I wouldn't want to change the indexing for people who are close to retirement. Maybe for younger workers who have time to plan (and save more--with or without privatization).

Pozen's plan, which the President has embraced, is worth discussing.

Under the current system, the benefits set at retirement are supposed to grow, on average, at the same pace as wages, so that the comparative living standards of retirees, while generally lower than working Americans, do not erode below today's levels.

Mr. Pozen's plan would maintain that schedule only for the bottom 30 percent of the work force - those with average annual earnings up to $25,000.

At the top, those earning more than the taxable limit - expected to be about $113,000 in 2012 when the plan would start, would have future benefits uncoupled from wages and linked instead to inflation, which tends to grow at a pace about 1.1 percentage points slower than wages. In the middle, benefits would be indexed by a mix of prices and wages.

2012 is a little soon for me unless it is phased in slowly, but it's a starting point for negotiation.

I'll continue this in the next post.

Headline (NY Times): Survey Finds Many Have Poor Grasp of Basic Economics

You guessed it. This is not simply a story about the need to improve economics education and basic financial literacy.

Their implied conclusion: Many don't have the skills to cope with an ownership society.

...there was confusion about the purpose of mutual funds, with some students stating that they provided higher returns than individual stocks, and others stating that they guaranteed a steadier income. Only 15 percent of students understood that the purpose of mutual funds was to provide diversification.

Mr. [Alan] Krueger, who contributes a column for the business section of The New York Times, said these findings were disturbing, given the big increase in the number of households that hold stocks and mutual funds.

"Many Americans are potentially open to scams because they don't understand the purpose of the financial markets," he said yesterday.

I agree that's disturbing. But they did not report how adults did on that question (or adult owners of mutual funds). The millions of adults who actually own mutual funds probably understand their purpose a little better than that.

Other analysts said they thought that the findings added to a growing body of evidence that the typical American is poorly equipped to take advantage of what proponents call the ownership society: a future in which individuals are free to invest their own retirement money, rather than having to accept the returns offered by the Social Security program or a group retirement program at work, like a pension plan. Many surveys have shown the public has doubts about the Social Security program, with young people, in particular, confident that they could do better by investing on their own.

Yet even their concern is poorly informed, according to the Employee Benefits Research Institute, a nonpartisan research organization that is financed by companies and labor unions. The institute's own research showed that fewer than 20 percent of workers thought that Social Security would be their primary source of income in retirement, even though Social Security is currently the primary income source for more than two-thirds of retirees.

I don't doubt the validity of these statistics (and I agree that some people are probably deluding themselves), but I do want to give this some additional thought. Is it realistic in the current environment with 401(k) plans, IRAs and so on, to expect that the number of future retirees (current workers in the survey) whose primary income source will be Social Security will equal the portion of current retirees who rely so heavily on Social Security? I don't think so. Now, do I think there will be as massive a shift as the survey implies? Of course not. Would a shift be a good thing? I think so. This story just doesn't give me enough information to say any more than this--it may not be as bad as it seems.

I'll bet I know what people said on this one.

Even though the Social Security Administration sends all participating workers individual annual statements, the institute found that only 18 percent of Americans know at what age they will be eligible to retire with full benefits.

I'm sure a lot of them said 65, but if they are under 44 years old it's 67. As retirement nears, they'll get the message. This question worries me less than some of the others.

The article concludes:

The purpose of the National Council on Economic Education is to raise the public's understanding of the economy. It created a basic standard for high school level financial literacy in 1997 and conducted its first survey in 1999. The economic literacy of both students and adults has improved since then, but only slightly.

Talk about burying the lede. No details on just how much it has improved. I'll try to research that for you and report what I find. We can clearly do better, and I think we also have to remember that if the stakes were higher and a partial privatization ever came to pass, I think we would. After all, if people didn't have to make choices about the allocation of scarce resources because they were all made for us, I (and many of you, my economist readers) would be out of a job.

Our job is to present these concepts and issues in our courses, discuss them on our blogs, and work with the MSM to get the message out.

The fact that so many people do participate in 401(k) plans, contribute to Roth IRAs, and so on, is very encouraging indeed. Rather than wringing our hands, let's redouble our efforts.

N.B. Even if privatization does not happen, I do not shrink from anything I say in this post. Economic and financial literacy is more important than this single issue.

UPDATE: Don at Cafe Hayek has much more.

Jane Galt questions the value of the Social Security trust fund.

Upon even a moment's reflection, it's obvious that the trust fund does not exist in the way that its proponents are claiming -- as a guarantee of benefits -- because the bonds are not obligations to Social Security beneficiaries. They are obligations to the Social Security Administration. (emphasis in original)

She continues:

Let's say it's 2018, and Congress is running out of money, as the Social Security system stops paying money into the government coffers, and starts taking it out. Congress cuts benefits to the point where Social Security taxes are once again a net contributor to federal revenue. Have we violated the trust fund? Nope. Congress is still paying the interest on those bonds; it's just that the interest they pay is immediately lent back to the federal government. Congress could knock benefits to zero, and keep recording interest payments into the trust fund for all eternity, without violating anything except its constituents expectations.

Now, is this likely? Probably not, because the political cost would be high. But the point is that the continuation of benefits depends on the political cost of offending a highly motivated interest group, not the existance of this trust fund. And the effect on the government of continuing those benefits--forcing it to raise taxes, cut other spending, or borrow money to pay for them--is exactly the same whether or not the payments are recorded on the books as "interest on bonds" or "contribution from general revenue". (emphasis mine)

That last part is what tells me that she understands the concept of the trust fund. I gather, however, that she is just a little less optimistic than I am about the government's willingness and/or ability to make good on their past promises.

But that's ok. That's where the discussion should be focused. In order for any progress to be made on this issue, we all must agree that the generational contract is what is important and that contract has nothing to do with how many bonds are in the SSA's file cabinet. If you happen to interpret the part of my sentence in italics as meaning that the trust fund consists of "worthless IOUs," that is your interpretation. I would, however, say that they are no more or less valuable than the IOUs (T-bills) that the government issues to the public. They are not worthless in that sense. They will be paid to the SSA and not repudiated. But their connection to actual benefits paid is a political, not an accounting, relationship.

To put it another way, those who will retire between 2018 and 2042 should consider this question. How do you think of the government's promise to pay you Social Security benefits?

a) a promise of a stream of income related by a formula taking into account the earnings over my lifetime, indexed by the growth of wages, and once I start earning it, adjusted for increases in the cost of living.

b) a claim to a share of the bonds in the trust fund.

Just because the answer is (a) that doesn't mean that the trust fund is worthless. If we can't agree on that, I don't think we're going to get very far. This sort of thing doesn't help.

Unfortunately, it is easier (for both sides) to politicize the trust fund than it is to discuss the generational contract. It's inevitable (for reasons that economists on both sides understand quite well, I think), but it's very unfortunate just the same.

The title of this NY Times article is, "When It Comes to Managing Retirement, Many People Simply Can't."

I'll admit it, the headline made me read the article. Score one for the Times. But there's a lot more in that article than just that one-liner. A more accurate title would be, "Some People Enter Retirement With Too Little In Their Private Accounts; Reasons Vary." You now see why I'm an economics professor and not a headline writer.

One very important reason mentioned in the article was that some people start saving too late in their working years. 401(k)s and IRAs are a relatively new invention. Two people mentioned in the article saved for only 17 years. A 22 year old college grad could save for more than 40 years if they start right after they get their first job. This is a pretty good argument for making private accounts a part of Social Security (whether they are a carve out or an add on is less relevant to my point here than just getting 22 year olds to do it automatically and steadily--let's save the carve out/add on argument for another day). Saving for retirement must begin early. Rome wasn't built in a day. I would expect that as time goes on and later generations of workers (who have had 401(k)s for a longer period of time) retire, the results will improve.

Consider this:

Ms. [Annika] Sunden [of Stockholm University and Boston College's retirement studies center] and Alicia Munnell, director of the Boston College retirement studies center, estimated that a worker making the minimum ideal contribution of 6 percent of wages to a 401(k) plan, with a company match of a further 3 percent, could do better than with a traditional pension.

If the account achieved a 4.6 percent annual return on top of inflation, the assumption adopted by President Bush based on historical returns, a worker retiring at 62 after 30 years of savings and a final salary of $52,650 would amass more than $350,000. This could provide annuity income for the rest of the person's life of roughly $31,000 a year, higher than the $26,500 such a worker would get from a typical defined-contribution pension.

Disciplined saving is a good thing. But...

In 2001, the most recent year for which comprehensive figures are available, the Federal Reserve's survey of consumer finances found that the median savings in a 55-to-64-year-old American's 401(k) or individual retirement account added up to $42,000, less than one-eighth the amount needed at 62 to achieve the retirement income estimated by Ms. Munnell and Ms. Sunden.

$42,000 is not much of a retirement nest egg. Obviously people are starting late and not saving enough. I think we've pretty much established that now. Towards the end, the article says,

Fewer than 10 percent of eligible workers contribute the maximum amount to their 401(k), and about a quarter contribute nothing at all. Many younger workers empty their 401(k) accounts when they change jobs - postponing saving for retirement.

Low savings are not the only problem. After they retire, workers must manage their savings, a task often complicated by unexpected expenses.

It's hard to imagine something worse than cashing out your 401(k) when switching jobs. If you have to use it as a cushion to tide you over to the next job (if you really must), then at least put back what you don't spend when you take a new job. The article ends with an account of how difficult retirement planning can be:

When Robert Stacy took an early retirement package from U S West more than six years ago, at 53, he had a traditional pension. But he took only half as an annuity, worth $900 a month, and the other half as a lump sum. His total savings at the time, including his 401(k) and the proceeds from the sale of his house, added up to almost $600,000.

The Stacys expected this money would be enough to roam through the country carefree in their new R.V. But six years later, the stash is down to $400,000. And the couple is facing higher health expenses since Qwest, which took over U S West in 2000, started charging hefty premiums on its retirees' health plans. So the Stacys, too, decided to take up jobs in retirement. "If I had to choose again," Mr. Stacy said, "I would have taken it all as an annuity."

Maybe so. But nothing here suggests that Mr. Stacy didn't understand his options. This seemed like a strange way to end the article.

Anyway, if the intent of the piece is to scare people away from private accounts for Social Security, it doesn't deliver the goods. The whole idea of private accounts should be to get younger workers used to saving and taking ownership of their retirement funds. Judging individual retirement accounts by the fact that most people haven't been contributing to them for very long strengthens my contention that starting young is the most important thing. An ownership society will take generations to build, but every day we wait just puts it off further.

The private accounts being discussed for Social Security would not (to my knowledge) allow workers to cash them out when changing jobs like a 401(k). That's one more problem eliminated.

Furthermore, the Hagel bill (S. 540) specifies that the private accounts be converted to an annuity on retirement. You won't have the same regret about your Social Security private accounts that Mr. Stacy did about his pension. Still another problem solved.

(Another nice thing about S. 540 compared to H.R. 530 is that the eligibility age is lower--45. I'd argue for even lower, but at least this is moving in the right direction.)

The Times article could certainly be interpreted as an attempt to show the perils of relying on private accounts for retirement (especially if you only read the headline). On the contrary, it makes an effective argument that disciplined saving over one's working life is very important. It also makes the point that once retirement hits, those savings should not be left to chance. Neither of these points is controversial.

Advocates of private accounts need to start making these very uncontroversial points.

Thanks to a comment from Calculated Risk (by way of Movie Guy), I looked at some publications on the other trust funds. This is an interesting one that came up in the search.

I didn't find any long term projections but did find some history and some clarifications on how they are handled. There really are a lot of them, fortunately most are extremely small in the grand scheme of the budget. Even the larger pension plans are small compared to Social Security (and let's not even talk about Medicare). Plus, I suspect the demographics are different. Military pensions, for example, would have income/outgo profiles that respond to the demographics of the military rather than the demographics of the work force in general (which is the problem for Social Security).

One other thing I noticed while reading the publication is that the 1999 balance of the Civil Service Retirement and Disability Fund was around 6 times its 1999 revenue. Compared to Social Security where the 1999 trust fund balance was not even twice its 1999 revenue. I have my theory on why this is the case.

All in all interesting, but still nothing that worries me greatly.

Calculated Risk makes a good point. There are other trust funds, like the Civil Service Retirement System, will face demographic problems similar to that faced by Social Security. Acutally, I haven't seen long term demographic projections for these funds, though I assume such projections exist. These other funds are "on-budget," which means that you should be able to find the short term projectons in the budget (something to do this weekend).

Granted, the magnitude of the problem will be much smaller, so I am confident that this can be handled. Also, I would be interested in hearing the details of how these trust funds work. Civil service pensions would seem to more closely resemble employer managed pensions than a general Social Security system. I would think that the administrative details would be a little different.

So, while this doesn't particularly trouble me, it is a good point. Most of all, it highlights the problem of lumping every thing into one "unified" budget. I like the idea of generational accounting for Social Security and separating it (and perhaps the other trust funds) from the general fund.

Dead Parrot vs. Angry Bear. Quite a mental image, isn't it? Anyway, lots of stuff to address, and I can't figure out where to break it into two posts. So here goes...

If you haven't been following it, today's participants in the Social Security free-for-all have been PGL of Angry Bear and Victor of Dead Parrot Society. They cover a lot of ground, and I'm not going to trace through it all. Today, I will cover just one aspect of their posts. This is a long post, but I have a couple of tidbits for you that I think are worth reading.

One observation I have is that the debate on the econoblogs is moving towards more general issues of government financing. Not that I have a problem with that. Might be a good thing.

PGL and Victor are arguing over whether increases or decreases in specific taxes affect the overall trend in spending. This is sometimes called the "starve the beast" theory. Cut taxes and spending will fall. Obviously it's not that simple in reality. It certainly hasn't been the case this time around. But the discussion then turns to the experiences of the 1980s and 1990s.

So, was the Social Security surplus that big of a deal during the Reagan administration? Nope. Not at all. You can say that Reagan raided the lockbox if you want, but at the time, the lockbox was the government equivalent of petty cash. Year to year fluctuations in the on-budget deficit were larger than the Social Security surplus. No evidence of "raiding the lockbox" yet, not in any meaningful sense. (Ah, but read on!)

The Bush (41) years were different. The size of the off-budget surplus (mostly Social Security) stayed relatively constant. Meanwhile the deficit took off. Looking at the numbers, you can't make a convincing case that Bush (41) was looking to the trust fund as a contemporaneous source of revenue. I suppose he could have been forward looking, but I'll leave that for you to decide.

The tax increases of the early '90s were perfectly timed (for everyone except Bush 41). The recovery was well underway by the time they really started to bite. Revenue growth outpaced spending growth. The on budget deficit fell for 8 years (becoming a surplus for 2 of those years). Meanwhile, the off budget surplus took off for demographic reasons that have very little to do with what was taking place "on budget." So, did Clinton raid the lockbox? No more (or less) in substance than Reagan or Bush (41). There was, however, a more subtle raid taking place--one that continued into Bush (43) as the following question makes clear.

What was the peak surplus at the end of the last cycle? 236 billion or 86 billion? If you say 236 billion, you're guilty of raiding the trust fund. (The off budget surplus was 150 billion in 2000.) The trust fund was getting huge just in time to make our already pretty good deficit fighting effort look even better. But the two have very little to do with each other. So who can blame the Democrats for touting hundreds of billions of dollars of surpluses? I certainly don't! (And don't say they didn't make those claims and rhetorically raid the trust fund--the proof is here. Furthermore, Gore wouldn't have made the lockbox such a big deal in his campaign if they hadn't already noticed that their hands were in it.) The current administration has simply continued on this path. In fact, the off budget surplus hasn't grown much since Bush (43) took office. The on budget deficit has. Like father, like son.

The punch line is that there really isn't much connection between the growth of the trust fund and the changes in trends in taxes or spending, at least not contemporaneously. That explains my reasoning for being skeptical of the "starve the beast" theory as well as my skepticism over any cause and effect connection between the tax increases in the early '90s and the fact that the economy boomed and government's share of GDP fell during the Clinton administration. There you have it--I'm an equal opportunity skeptic.

So, I don't mind making this debate at least partly about government financing in general. The general fund deficit is a big problem worth talking about independently of Social Security. I happen to think that if Social Security reform is done correctly, it won't make the general fund problem much better or worse. So let's do Social Security reform correctly and tackle the general fund problem separately.

PGL often claims that the payroll tax increases were to prefund the baby boomers' retirements (and thus he contends that claiming that we are still essentially pay-as-you-go is equivalent to saying Reagan lied in 1983). The first part is true, at least rhetorically. The trust fund is a meaningful promise to pay--fully funded politically and morally (but not in a strict accounting sense). Because of this, your mileage may vary on the conclusion he draws. However, it's interesting that the first time this issue came up was in 1985, just as Social Security was being taken off budget. Reagan wanted a one year freeze on cost of living adjustments to Social Security. This would save $6 billion in FY1986. Imagine! He wanted to raid it for $6 billion and make the $200 billion dollar deficit look $6 billion smaller! I don't know what Reagan was thinking, but I think it's plausible to think that he either thought that in 20 years we would solve this problem or that we wouldn't be talking about such huge amounts being at stake. Again, you are free to make up your own mind. (By the way, this information is from the May 14, 1985 New York Times. If you have Lexis-Nexis, do a search for the headline "Two kinds of deficits" from that day's paper. It's an article worth printing and saving if you have access. I'd excerpt from it if this wasn't so long already. Maybe another day. To boil it down to whether the Times writer thought that Reagan was lying--his answer would have been a firm yes... and no.)

All historical data in this post are from the 2005 Economic Report of the President, Table B78.

Ok, that's enough for today. As always, your comments are welcome.

UPDATE: PGL responds. Also, take note of a comment by CalculatedRisk.

Let's start with a quote from macroblog.

I won't even bother to link to previous posts, but if there was ever a policy-driven transaction without consequences for national saving, it would have to be issuing explicit debt for already promised benefit payments. (OK, I lied -- here's the link.)

In the same spirit, here are links to a couple of my previous posts. I think Altig's comment above makes excellent sense as a first approximation. In other words, in a perfect world or a simplified model it should hold. The extent to which it doesn't hold exactly would, in my estimation, have to do with market imperfections and other "real world" problems. So when Brad DeLong says,

My position is that we really don't know what the impact of having the Treasury sell $4.5 trillion more of government bonds and then having individuals invest that $4.5 trillion in their private Social Security accounts will be. I would bet that there's at least a 50-50 chance that it will be a wash as far as national savings is concerned. I would also bet that there's at least a 20% chance that it will shrink national savings significantly--that people will regard their private accounts as relatively close substitutes for their 401(k)s and other assets, and so reduce the amounts they commit to funding their other retirement savings.

I have to give him credit for honesty. DeLong also says,

What I object to are assertions that people know that the effect on national saving will be a wash. They don't know this. What I object to are assertions that worriers--like me and Alan Greenspan--should "stop railing about the budget impact [of the Bush Social Security plan]. The... increase in the budget deficit won't place a new burden on future generations." There's reason to hope that this is the case, and I think it is better than a 50-50 bet. But as Uncle Alan said, it's important to go slowly: if it is a big mistake, we need to find that out in time to stop it.

I also said this more than a month ago.

Back to the issue of what happens to national saving, here's why I think there might be some ambiguity. Let's go back to macroblog again.

I have already conceded that the trust fund should be treated as a meaningful promise to pay future benefits.

Me too. Of course, there are various ways to shave the amount that we have to pay (like raising the retirement age or changing the indexing formula, to give two examples). One possible problem for Ricardian Equivalence is financial market uncertainty about whether and how much we'll shave that amount. If privatization changes those expectations, RE will not hold exactly.

Another possible problem would be if privatization had a very different generational effect compared to waiting until 2042 and the generations react in systematically different ways. This is a serious research question, not something you can answer off the top of your head.

I'm not sure how much either of these would impact national saving, but it's worth talking about.

Once again, the more gradual the transition, the better the chance for success. The sooner we start, the more gradual we can afford to be.

H.R. 530 is not gradual. If this ends up being the plan for consideration, I'll be as worried as Greenspan and DeLong.

Today's best Social Security analysis is by Arnold Kling. By the way, if you don't read his blog, you should.

The memorandum to which he refers is can be found at the SSA website.

He gets it right. At issue is the rate at which a worker trades off benefits when he or she contributes to a private account. If I'm reading it as he does, it appears that this proposal shaves the benefits by 3%, which means the private accounts have to return at least that much for the worker to break even. His problem with that is that 3% is too high. Alas, it looks like this is, as he says, "dictated by the Social Security actuaries."

I have to agree with Kling that 3% makes the system look more solvent, but might bias people against private accounts (or lead to inflated claims about the returns on those accounts--something which also appears to be happening).

Kling says,

"Actuarial scoring" of Social Security and the reform proposals is a menace. The assumptions used in scoring are rigid and misleading. If you want to be in a position to evaluate Social Security reform objectively and accurately, the first thing you have to do is throw out the actuarial analysis. Pay no attention to the "trust fund," the "solvency" of the system, or other noise generated by the actuaries.

They certainly are rigid. And it's true that the solvency or lack thereof depends on those rigid assumptions. I'd be more comfortable with a lower number, but that means that private accounts would have to be introduced much more gradually (to satisfy anyone's definition of actuarially sound). Of course, that wouldn't bother me one bit.

Oh, and one more thing. The proposal in the memo, which apparently is now going by the name of H.R. 530, lets anyone under age 55 start a private account. Read it!

I give you David Altig and Jagadeesh Gokhale's idea from back in 1996, nine years ago. (Yes, the David Altig of macroblog.)

Calculations using the current distributions of Social Security benefits by age and sex (assuming a 1.8 percent internal rate of return on the contributions of those included in the present system, and an 8 percent return on investments in private capital markets) suggest that 42 is the appropriate cutoff age. With this as the dividing line, 18 percent of the contributions of those age 42 and younger would be sufficient to provide those age 43 and older with benefits at least equal to those received under the current system. For younger workers, future benefits may be greater than those offered by the current system, because their contributions will reap the higher private rate of return for an even longer period.

Having dithered for nine years, I'd say the appropriate cutoff is at a younger age now. And this paragraph from the same 1996 article is delicious in its irony.

For an individual starting from scratch, the rate of return from a funded system will clearly exceed that of the unfunded scheme. However, if the cutoff age below which individuals are shifted to the privatized system is too high (say, 55), some workers would not have enough remaining years to exploit the increased private returns, leaving them worse off than before.

I don't know whether to laugh or cry.

Via the Club for Growth:

U.S. Rep. Martin Sabo (D-MN) announced this week that he will introduce a bill in Congress that he says will "guarantee Social Security solvency through 2080" by increasing the interest rate on Treasury bonds in the Social Security trust fund.

PGL at Angry Bear responds first. I'm going to take a different approach though. Rather than make this another post about what I think the trust fund is or is not (I've been pretty clear about that in the last few weeks), I'm just going to explain why I think this idea is not a good one.

My complaint is simple. Even if we keep the current Social Security system, I would just ask that the system's income and outgo be balanced in the long run. Right now, that's where the infamous 2018 and 2042 dates come into play. The plus side of the ledger wins until 2018, then the negative side of the ledger wins for many years into the future--hitting a zero balance in 2042. PGL often comments that the Republicans want to raid the trust fund or that Reagan lied in 1983. I respond as follows. If we raise taxes (or borrow less!) from 2018 to 2042 to pay back the bonds in the trust fund, then David Ricardo, for one, would not call Reagan a liar.

(The fact that many people reading this will find the previous sentence unrealistic is a political issue, not an economic one.)

The problem is that the world does not end in 2042. Beyond that year, the system goes negative and would probably stay negative for a long time, forcing the general fund to subsidize Social Security, perhaps indefinitely (unless there is another baby boom). I think we can all agree that the pay-as-you-go system is not in long run balance when the long run is taken to be past 2042. That would bother me. At that point, I would want us to raise the retirement age, raise payroll taxes, or cut benefits. Of course, I really would prefer that we never get to that point in the first place. That's why I support doing something now, while there is a lot of time to do something gradually.

Anyway, back to Rep. Sabo's plan. His plan would simply shift the financing from the pay-as-you-go Social Security system to the general fund... until (at least) 2080. He's talking about doing exactly what I would try to avoid... for 40 years. This is more than just a counterpoint to the "raiding the trust fund" argument. This would institutionalize a long run imbalance in the pay-as-you-go Social Security system--a long period of deficit without a corresponding surplus. In my rank ordering of approaches to the Social Security issue, this ranks right below doing nothing.

The discussion that started at Angry Bear and macroblog (not to mention the others that have piled on) got me thinking.

We use this word "default" a lot in these debates over the appropriate way to pay for war, pay to restructure Social Security, and otherwise finance our debts. When most people hear the word "default," they probably think of firms which have gone out of business without paying off its debts. Bonds from companies that face this kind of risk carry the adjective "junk" due to the fact that you may not see any return if the firm goes under. Perhaps some people relate default to third-world debt crises. But are the third-world debt crises or the kind of default that results from a firm going out of business appropriate models for the U.S. economy?

I don't think so.

(*I can hear the shuffling of feet and clearing of throats out there in blogland*)

Let me explain. The part of Krauthammer's piece that had everyone in an uproar was the part about the IOUs in the trust fund being worthless. OK, he doesn't say "worthless," he says,

These pieces of paper might be useful for rolling cigars. They will not fund your retirement.

Whatever. This is the rhetoric of default. This is the (overly) dramatic statement of how bad things are. This is the heart of the "worthless IOU" argument.

But it doesn't fly. Krauthammer himself follows with,

To cover retiree benefits, the government will have to exhaust all of its FICA tax revenue and come up with the rest -- by borrowing on the world market, raising taxes or cutting other government programs. (emphasis mine)

And in so doing, he reveals that this is a General Fund problem. The worthless IOU argument was just a distraction. Whether he meant it to be a distraction or not, I can't say. It's an easy trap to fall into, but I'm sure that some writers have set the trap deliberately.

The real problem is not some theoretical point in the future when the U.S. repudiates its debt. Not likely to happen. What we need to be on guard against is the prospect of the insidious default that is caused by inflation (if the Fed becomes complicit in the borrowing and becomes the buyer of last resort for the IOUs). Another version of default was pointed out by Dave and I will repeat the two paragraphs that make the point extremely well.

Suppose I hold a Treasury security. That, of course, is a payment the government owes to me, and I have every expectation that it will be made. But if, for some reason, there has been a miscalculation, a change in economic circumstances, a change in policy, the government may find that it has to raise my taxes to obtain the revenues to honor those payments. In doing so, it has effectively reduced the return on that security. Distortionary price effects aside -- granted, a major qualification -- why should it matter to me how it happens? Lower my social security benefits, raise my income taxes, whatever. It all amounts to a haircut on that Treasury payment to me.

Because the distributional aspects of these things can matter, blanket haircuts are probably a pretty bad idea -- foreigners, for example, finance a good chunk of our collective borrowing, and they aren't likely to appreciate the opportunity to finance our fiscal imbalances on an ongoing basis. Changes in tax and transfer policies are the way we go because they can be targeted (which gets us to positive versus normative questions, which I'll address below.) But the basic economic distinction is one without a difference.

Precisely.

Bottom line: We need to be careful about what we mean when we discuss (or imply the possibility of) default lest we fall into rhetorical traps. The "worthless IOU" argument is itself worthless. There are undoubtedly other cases where it has been used to get the attention of the reader and then been cast aside when it has served its purpose. Better to be upfront about the more likely (and just as troubling) consequences of too much spending, if that's the point you would like to make.

Krauthammer's argument does not require the "worthless IOU" concept and would be better off without it. And I think we should keep the discussion of Social Security reform honest by refraining from using that argument unless you are willing to apply it to the General Fund problem as well. I am not. I'm more inclined towards Dave's comments noted above.

PGL at Angry Bear critiqued a recent Charles Krauthammer article for essentially calling the IOUs in the Social Security trustfund worthless. Specifically, Krauthammer says,

Let's start with basics. The Social Security system has no trust fund. No lock box. When you pay your payroll tax every year, the money is not converted into gold bars and shipped to some desert island, ready for retrieval when you turn 65. The system is pay-as-you-go. The money goes to support that year's Social Security recipients. What's left over is "loaned" to the federal Treasury. And gets entirely spent. It vanishes. In return, a piece of paper gets deposited in a vault in West Virginia saying that the left hand of the government owes money to the right hand of the government.

These pieces of paper might be useful for rolling cigars. They will not fund your retirement. Your Leisure World greens fees will be coming from the payroll taxes of young people during the years you grow old.

And PGL replies,

If you really believe piece of paper called financial assets are worth nothing, I’ll gladly take all of your cash, funds in your bank accounts, and other financial assets. In turn, I’ll even buy a week’s worth of groceries.

Dave at macroblog posts a detailed response that defends Krauthammer's essential point while pointing out the comingling of positive and normative analysis in Krauthammer's article.

I'll go one step further. Krauthammer is setting up a straw man with the "worthless IOU" argument. But really folks, we've been over this before. This is a variation on the topic that Dave, PGL, and I addressed a couple weeks ago. Click here for my post which has links to the others.

Dave appears to remember that discussion since he correctly points out the positive/normative distinction. Here's what I said back then,

The trust fund is a social contract. In that sense, it's fully funded politically and morally. But I see nothing in the accounting structure of Social Security to suggest that it is fully funded in the strict economic sense.

The meaning of a default here is a breakdown of that social contract. That has real political/economic implications just like a default through the financial markets.

I also called the trust fund a "useful fiction." It's clearly not worthless, but it's just not fully funded in an economic (or accounting) sense. And I can say that with a perfectly straight face while refusing to accept Krauthammer's straw man the way he clearly intends it. Actually, I just read the whole article again leaving out those two paragraphs. It's much better that way. He actually makes a good point about the need to enact reform now, before the outgo exceeds the income in 2018. He doesn't say so in so many words, but I think his argument fits quite well with the view that the Social Security problem is really a general fund problem. (That's a view that I think PGL could accept.)

Dave's explanation of the distinction between Treasury securities and Federal Reserve notes is an interesting way of lifting the discussion to a more abstract level.

But if, for some reason, there has been a miscalculation, a change in economic circumstances, a change in policy, the government may find that it has to raise my taxes to obtain the revenues to honor those payments. In doing so, it has effectively reduced the return on that security. Distortionary price effects aside -- granted, a major qualification -- why should it matter to me how it happens? Lower my social security benefits, raise my income taxes, whatever. It all amounts to a haircut on that Treasury payment to me.

PGL still doesn't totally buy it (see the update at his post), but Dave responds directly to PGL in this comment to his own post.

The thing is, it is seemingly much easier to find your road to collapse by trying to finance imbalances with inflation than it is from explicit taxation or spending reductions. So we appear willing to countenance more "little defaults" with Treasury securities than we are with money. And I'll repeat myself -- if the government raises taxes and reduces the return to the debt I hold in the process, they default. We put up with it (a) because we can verify the state of the world and identify cause and effect; and (b) we can throw the rascals out of office for getting us into the situation if we want to (an idea I know you can embrace).

I think that captures the essential idea. Allow me to just add this point (saying the same thing in different language). PGL asks,

So if the Federal government defaults on its IOUs, what does the Federal Reserve rely on to honor its commitments?

If the Fed is really independent, it can honor its commitment without regard to the Federal government defaulting on its IOUs. The Fed would "simply" have to refuse to be complicit and maintain the value of the dollar (perhaps by allowing interest rates to rise--not pretty, but then refusing to be complicit in a default entails sacrifice). The reverse is the real problem. If the Fed is not independent, that's when its commitment to price stability goes out the window. If the government defaults on its commitments, the Fed is either complicit or not. If it is complicit, we've got all the problems PGL is worried about and then some. If not, then we're ok. I would further state that if the executive and legislative branches believed in the independence of the Fed, they will be less likely to default (in Dave's sense of many "little defaults") because they know they will bear the political cost (part b of Dave's comment).

Thoughts?

Brad Setser points us to an interesting tidbit from the LA Times:

During the transition, less money would be paid into the system even as it paid out at current benefit levels. Bush said earlier this week that he would not rule out paying those transition costs by raising the current wage cap of $90,000 that can be taxed for retirement.

On Thursday, a number of conservatives said that directly contradicts Bush's earlier promise that he would refuse to raise taxes.

At least he waited until after the election--something his Dad didn't do.

But seriously, it's not necessarily a bad idea. You probably wouldn't have to raise it much to have a pretty good bump in revenue (Club for Growth arguments notwithstanding). There are many good reasons to keep the growth of the "cap" small and gradual, but it shouldn't be treated as a sacred cow that arbitrarily cannot be touched.

After all, the benefit structure of Social Security is meant to be progressive. The payroll tax is not. I've never understood this. So, a small, gradual increase of the cap (in addition to the usual inflation adjustment) is not a deal breaker for me, as long as it is just that--small and gradual.

UPDATE: The NY Times has an editorial that makes one think of this as a dynamic bargaining problem. They even propose a solution.

Paul Krugman's latest NY Times piece gets it partly right. First, the not-so-good:

But privatization "as a general model," he said, "has in it the seeds of developing full funding by its very nature." Nice metaphor, but what does it mean? Clearly, he was trying to create the impression of links where none exist.

Swing and miss. Private accounts must be fully funded. (There are no credible proposals to the contrary.) Ergo, if fully funding Social Security is a goal, private accounts represent a means to achieving that goal. Granted, it's not the only way, but it is a way. Seems to me that's the link he [Greenspan] was going for, and it's correct as far as it goes, whether you agree with private accounts or not.

He does get a couple of hits though:

Privatizers claim that financial markets won't be disturbed by all that borrowing because the Bush plan prescribes offsetting cuts in guaranteed benefits for the workers who open private accounts. Mr. Greenspan, who does know a thing or two about markets, put his finger on the reason why those prospective future benefit cuts wouldn't offset current borrowing in the eyes of investors: "Well, the problem is that you cannot commit future Congresses to stay with that."

Valid point in general, so it's a hit. However, I think decision day on this is still many years in the future. The financial markets will not pull the trigger until and unless private accounts are unsuccessful to the point that Congress will be tempted to break the commitment to lower benefits. If you are confident about private accounts, you won't worry about this as much. If you don't think private accounts will deliever the goods, it's an entirely rational point to raise.

Yet the chairman managed to avoid admitting the obvious - that borrowing on the scale the Bush plan requires would substantially increase the risk of a financial crisis. And the headlines didn't emphasize his concession that crucial critiques of the Bush plan are right. As he surely intended, the headlines emphasized his support for privatization.

Two ideas in one paragraph. The first sentence belongs with the paragraph which preceeded it (above). The last part is correct. The headlines did emphasize his [Greenspan's] support for privatization. No doubt about that. Is it what he intended? Well, you could argue that he's been a central banker long enough to know what the headlines would be if he said certain things. The fact that he said them anyway is circumstantial evidence in Krugman's favor.

I can't really comment on the last part of his column directly, though I wish I could. I'll explain why and do the best I can.

One last point: a disturbing thing about Wednesday's hearing was the deference with which Democratic senators treated Mr. Greenspan. They acted as if he were still playing his proper role, acting as a nonpartisan source of economic advice. After the hearing, rather than challenging Mr. Greenspan's testimony, they tried to spin it in their favor.

But Mr. Greenspan is no longer entitled to such deference. By repeatedly shilling for whatever the Bush administration wants, he has betrayed the trust placed in Fed chairmen, and deserves to be treated as just another partisan hack.

I did not see the hearings. I was working on more pressing matters on Wednesday, and on Thursday I was teaching. C-Span sadly has not rebroadcast the hearings, at least not to my knowledge (and I have checked their website every day). Often they replay these in the evenings or later in the week, and I'm very disappointed that they haven't done so this time. I can say that from previous hearings that I have seen on C-Span, the treatment was anything but deferential (at least since the recession) from the Democrats. One of the reasons I watch them is for the entertaining way in which the Democrats verbally lash him and he responds in such a soft-spoken, even-handed way.

However, I did find this quote on Rueters:

"I do have to express skepticism that telling workers losing their jobs ... 'Do not despair. Private accounts are coming' will be less a morale booster than I think you implied," said Rep. Barney Frank, a Massachusetts Democrat.

I suppose you could deliver that line in a deferential manner. It could also be delivered with the implication that he's a partisan hack. Unfortunately, I did not see the delivery, just the words.

This one, for example, doesn't cut it.

While Social Security reform cannot and must not include tax increases, there is an opportunity to increase economic growth and government revenues by cutting effective marginal tax rates. This is why there must be no “phase-ins” and no “caps” on the 4% of wages that workers should be able to put in their Personal Social Security Accounts. With Personal Accounts with no “phase-ins” and no “caps”, Social Security Reform will act as a 4 percentage-point cut in marginal tax rates on everyone making less than $90,000 per year. This would give a significant boost to economic growth and, over time, Federal revenues.

Anyone selling privatization as a means to growth is selling snake oil. If you want to sell it on other grounds, like participating in a growing economy or being a member of the "ownership society," that's one thing. And I'm sympathetic to privatization for a number of those reasons. But holding up Social Security privatization along with a Laffer curve and saying that it will "give a significant boost to economic growth and, over time, Federal revenues" is raising expectations a little higher than I'm comfortable with.

[Note: I was born and raised a Minnesotan with a gift for understatement. You are free to interpret the last sentence in light of that fact.]

Thanks to PGL at Angry Bear for the link.

Read here

Not enough time for deep analysis tonight, but suffice to say that The Economist is supportive of privatization. They also provide a counterpoint to the many critics who point to the British experience as a reason to avoid privatization. More here.

If you don't have a subscription to The Economist, get thee to a library once a week. Good stuff.

UPDATE: Brad DeLong is not amused. Now, I should clarify that while I still think The Economist is good reading, I clearly don't follow them blindly.

While DeLong makes some good points, he doesn't address this article or these comments specifically.

However, in both Sweden and Australia, as in Chile, the new accounts are mandatory, while in America they will be voluntary. To find out how voluntary accounts have worked, Americans need to look at Britain, where Margaret Thatcher introduced them in 1988.

That should be enough to put them off the idea for good, according to Mr Bush's critics. In their version of history, Britain's experience was a disaster, in which people who opted for individual accounts were made worse-off by pension mis-selling. Fortunately, the critics are wrong.

To be sure, pensions were mis-sold in the late 1980s and early 1990s: the bill for putting things right was £12 billion ($22 billion). But the mis-selling was out of employers' defined-benefit plans, not out of the state system. Many people were lured away from generous employers' plans into funded individual pensions when they would have been better off staying with their employers' schemes. “Mis-selling was not about people being sold private pensions when state pensions would have been better for them,” says Philip Booth, the editorial director of the Institute of Economic Affairs, a think-tank.

Britain's mis-selling scandal occurred within a distinctive pension system that had long allowed employers to provide part of the overall state benefit in return for rebates on part of their payroll taxes. In the late 1980s, this right to “contract out” was extended to individuals, who were also given the right to leave their employers' plans. In America, as Olivia Mitchell, a member of Mr Bush's pensions commission in 2001, points out, there is no “contracting out” for private workers in Social Security and the new individual accounts will form part of Social Security.

I haven't heard that argument in the American media. I do hear all kinds of talk about how we shouldn't let our system be like Britain's. I'd welcome more on this point from those in the know.

DeLong concludes, quite fairly,

To head off this moral hazard meltdown, a plan should--and the Bush plan appears to--very tightly constrain where the investments can go.

The moral hazard is the temptation to bail out the private accounts whose investments go bad.

I agree, and I did read some of the Bush plan tonight too. He's right. It does appear to offer some safeguards. I'm not completely sold on the Bush plan though. It seems different from the three proposals that had been discussed earlier (maybe just more detailed--I didn't sit with all of them side by side). I think we need to scrutinize it more carefully. And that's what I intend to do. I hope you will join me over the next few months.

I still learn more from one issue of The Economist than from (at least) two issues of American newsmagazines.

Thanks to Angry Bear for the link.

It reads in part:

Under the current White House proposal, individuals could put about a third of their Social Security contributions into six highly diversified index funds.

These index funds would include ordinary Treasury bonds, inflation-indexed Treasury bonds, corporate bonds, small-cap stocks, large-cap stocks and an international index fund. At retirement, individuals could choose to purchase an annuity with their accumulated wealth.

In theory, those with little wealth should invest in the safest investments, probably inflation-indexed Treasury bonds, and choose to purchase the annuity when it is offered. But these actions would not have much of an impact on their retirement income because the safe return on inflation-indexed Treasury bonds is not all that different from the safe returns offered by Social Security.

We agree in our advocacy of inflation indexed bonds as the safe investment and in our estimation that the returns will be similar to that offered by the current system. Personally, if there was an over/under bet on that question, I'd take "over", but it would be close.

But it gets worse. Social Security is structured as a "pay as you go" plan. A transition to private accounts would mean huge borrowing by the federal government to cover the Social Security payments owed to today's recipients.

There are those who argue that there would be no impact on financial markets from such borrowing, as it is merely replacing one liability (future Social Security payments) with another (future payments on Treasury obligations).

I have stated previously that I almost agree, but not to the point of saying that there would be no impact. Less impact than many people would think? Yes.

But this seems to me to be quite unrealistic. The Social Security obligations are "soft obligations" - all it takes to change them is an act of Congress. And indeed, we are discussing such changes right now.

By contrast, Treasury bonds are "hard obligations." They must be paid no matter what.

It is likely that financial markets would react quite negatively to the huge amount of debt required to finance private accounts. And such accounts offer little benefit to individuals anyway: those with little income should invest in safe assets, and the additional flexibility offered to high-income individuals is not very relevant.

The fact that we currently have "soft" obligations creates a time consistency problem and is a compelling reason for reforming the system. (Recall this discussion from the archives of Macroblog.)

And so on that basis, I am skeptical that the financial markets would react too severely. It would be true if you really believed that benefits would be cut substantially in 2042 when the trust fund runs dry (if we did nothing to reform Social Security). If you've been following my posts, you know that I don't believe that.

I also don't quite accept the claim that "additional flexibility offered to high-income individuals is not very relevant." I think age matters. I think high income young people would appreciate being able to increase the equity portion of their portfolio. Remember, too, that I would start the reform with the very young. They have the most to benefit, and the transition cost is lower when you start with the very young.

So while I don't agree with some of the article, it does correctly address two important aspects of the debate separately. It is a meaningful contribution to the debate.

From the section of the budget on the fiscal outlook, which discusses Social Security in broad sweeping terms:

The effect of creating personal accounts would be to protect some portion of worker payroll taxes from the reach of the Federal Treasury. The Federal Government might therefore have to increase its borrowing in the private capital markets by an amount equal to the annual flow of payroll taxes into the personal accounts. While the creation of personal accounts would increase Federal borrowing, they would also reduce the Federal Government’s future spending obligations by an equal amount. Both effects must be shown and understood together—both the size of near-term investments in personal accounts, under reasonable participation assumptions, and the corresponding amount by which reform would reduce future unfunded obligations. It is misleading to measure only the near-term impacts.

Under most circumstances in which the Federal Government borrows funds, it is because the Government is spending more on goods, services, and transfer payments than it is taking in. In the case of personal accounts, however, the reduction in national savings from the Federal Government’s increased borrowing exactly matches the increase in private saving that occurs through the personal accounts. There is no net reduction in national saving arising from this arrangement, nor is there any reduction in the flow of saving available to the private sector. For this reason, the creation of personal accounts is not expected to have any detrimental effect on financial markets or on the overall economy.

Comprehensive reform that includes personal accounts would permanently eliminate the unfunded obligations of the current system. Ultimately, that is the standard by which any legislation to strengthen the Social Security system must be held. To achieve this, the long-term growth in annual Social Security outlays cannot be greater than the long-term growth in program-generated receipts.

Projections regarding various proposals' fiscal effects and their impact on beneficiaries will be based on analyses provided by the non-partisan actuaries and other technical experts at the Social Security Administration.

Let me paraphrase what I have said repeatedly on this blog. In principle, I think this is pretty close to the mark. It will not be exact Ricardian Equivalence, but people will choose to hold some bonds. Stock prices will not get totally out of line with bond prices.

More to the point they are making, I think they also have the right idea about the transition cost. First of all, the transition cost simply brings forward the future obligations that under the current system we will not be able to pay.

So this (from the first paragraph quoted above) is, at least as a first approximation, correct:

While the creation of personal accounts would increase Federal borrowing, they would also reduce the Federal Government’s future spending obligations by an equal amount....It is misleading to measure only the near-term impacts.

Yes, I realize that this is a bit heroic. It's a first approximation. That is why I went on to qualify my stance a bit further.:

I'd like private accounts, but I'd like them done correctly. The more gradual the transition, the more the transition cost can be spread out. The sooner we start, the more gradual we can afford to be.

If anyone at the White House is reading this, you can have that line free of charge.

Good discussion today from Dave at macroblog and pgl at Angry Bear. For full effect read Angry Bear first, then macroblog with the comments from pgl (and me).

The current Social Security surplus makes the General Fund deficit look smaller. We have been, in other words, raiding the "trust fund." In that narrowly defined sense, Social Security is pay-as-you-go. In that sense, the system will go bankrupt in a few decades due simply to the collision of demographics and economics. In that sense, payroll taxes will only pay 73% of benefits in 2042 according to the statement I recently received in the mail.

All of this assumes that the rules don't change. Of course, we all know that the rules will change somehow, sometime between now and 2042, one way or another.

Dave asks:

First, does anyone seriously believe that "repudiating" the accumulated social security surpluses by altering benefit rules would be viewed as the same sort of thing as refusing to pay off debt held by foreign governments or the public? Do you believe, in fact,that it would create even a perceptible ripple in the deep waters of our financial markets? If you answer "no," then aren't you admitting that the the world perceives the trust fund as little more than an accounting gimmick?

My answer is "No," and I therefore admit it's an accounting gimmick. (If you've been following my posts, you knew that!) Foreign governments and the public in their role as bondholders wouldn't bat an eye. The public in their role as workers and retired persons would scream political bloody murder. And I think you can differentiate those roles. Politicians know this too, and it's what will drive them to do something, sometime, to preserve the benefits.

Fully funded politically and morally, not economically. Not in the narrow sense, at least.

Now for the punchline. Does it matter that the funding of our benefits is based on political trust? Probably not so much. The "Trust Fund" is a "useful fiction" to perpetuate the idea that there will be something for you. It's up to presidents and congresses between now and 2042 to work out the details, but there will be something. The voters understand that a lot better than they understand the economic distinction between fully funded and pay-as-you-go. They may not articulate it, but I think they understand it at some level.

In my comments at macroblog, I mentioned Ricardian equivalence, a gauntlet that, once thrown down, pgl (of Angry Bear) quickly picked up.

The Ricardian Equivalence point of William is intriguing. RE makes the following implicit assumption. There is no game of chicken ala Sargent & Wallace's devastating writing about Reagan fiscal policy where all agents have a good idea of how the budget will be eventually balanced. But suppose we have this Reagan-Greenspan 3-card monte (to take AB's excellent post on this). Workers were led to believe they were prefunding their SS retirements and not paying higher employment taxes. The Kudlow and Luskin investor class, however, realized they were getting permanent tax cuts to be later paid for by robbing the lockbox. Workers do not curtail consumption while the investor class increases its consumption. And ole - consumption goes up which seems to contradict the simple version of RE. Were workers irrational to think Reagan was telling the truth? I guess you believe so.

Wow. This is just asking for a nice theoretical model. Maybe later. Anyway, he's got it about right. Insofar as payroll taxes affect the worker class and investor class differently, pure RE will not hold, at least with regard to the savings decisions of the workers and investors.

My allusion to RE was actually a little more nuanced however, and goes back to the idea of the trust fund as a useful fiction. (Again, read my comments at macroblog.) Suppose the government raids the trust fund. Where does that money come from? Who does it belong to? The question is key to whether it's a pay-as-you-go or fully funded system.

If they are required to honor the social contract (defined benefit structure), I don't see that it matters. With payroll taxes that are less than perfectly correlated with the defined benefits, I don't see how it matters who you say you are borrowing from. This seems a lot like a form of RE, though it is a little different from the usual notion. It's a generational sort of RE--which generation are you taxing to provide defined benefits for a given generation? Are you taxing the present generation of workers or did you tax the previous generation of workers years ago and borrow from them to finance last period's spending?

The trust fund is a social contract. In that sense, it's fully funded politically and morally. But I see nothing in the accounting structure of Social Security to suggest that it is fully funded in the strict economic sense.

But I think that a form of Ricardian equivalence may in fact blur the line between the two concepts due to the defined benefit structure and the social contract. I realize that I've laid out these ideas rather crudely in this post, but I think that there is something there worth chewing on as the debate continues.

After all, there's got to be some explanation for the fact that there is disagreement over this. If the two types of systems are hard to tell apart in modern political application, that might explain it. I will continue to refine this idea, and I welcome comments.

Paul Krugman returns to the simple numbers of Social Security and makes a good point against the proponents of privatization. It's a good point because it is technically correct. It's also not enough in my opinion to end the debate.

Here's what he says:

Schemes for Social Security privatization, like the one described in the 2004 Economic Report of the President, invariably assume that investing in stocks will yield a high annual rate of return, 6.5 or 7 percent after inflation, for at least the next 75 years. Without that assumption, these schemes can't deliver on their promises. Yet a rate of return that high is mathematically impossible unless the economy grows much faster than anyone is now expecting.

And later,

In the long run, profits grow at the same rate as the economy. So to get that 6.5 percent rate of return, stock prices would have to keep rising faster than profits, decade after decade.

The price-earnings ratio - the value of a company's stock, divided by its profits - is widely used to assess whether a stock is overvalued or undervalued. Historically, that ratio averaged about 14. Today it's about 20. Where would it have to go to yield a 6.5 percent rate of return?

I asked Dean Baker, of the Center for Economic and Policy Research, to help me out with that calculation (there are some technical details I won't get into). Here's what we found: by 2050, the price-earnings ratio would have to rise to about 70. By 2060, it would have to be more than 100.

What is left for the privatizers?

They can rescue their happy vision for stock returns by claiming that the Social Security actuaries are vastly underestimating future economic growth. But in that case, we don't need to worry about Social Security's future: if the economy grows fast enough to generate a rate of return that makes privatization work, it will also yield a bonanza of payroll tax revenue that will keep the current system sound for generations to come.

Alternatively, privatizers can unhappily admit that future stock returns will be much lower than they have been claiming. But without those high returns, the arithmetic of their schemes collapses.

It really is that stark: any growth projection that would permit the stock returns the privatizers need to make their schemes work would put Social Security solidly in the black.

And finally,

And I suspect that at least some privatizers know that. Mr. Baker has devised a test he calls "no economist left behind": he challenges economists to make a projection of economic growth, dividends and capital gains that will yield a 6.5 percent rate of return over 75 years. Not one economist who supports privatization has been willing to take the test.

But the offer still stands. Ladies and gentlemen, would you care to explain your position?

It's hard to find anyone predicting 6.5 percent in real returns over the next 75 years or even 50 years. A more conservative 3 to 5 percent would be more reasonable. Also, Krugman is entirely correct that if we did get the kind of growth that would push stock prices up 6.5 percent, the Social Security "crisis" would essentially go away. So far, so good for Krugman's argument.

I want to focus on this sentence:

Without that assumption, these schemes can't deliver on their promises.

"That assumption" being the 6.5 percent real stock return. For the last couple weeks, I have become increasingly convinced that this might be the case. And I think that there is a pretty simple way to address this. It's an idea that has been implicit in some of my recent posts on the subject.

Just lower the cutoff age for participation in private accounts. Remember, during the transition, the current workers need to "pay the freight" as Dave (macroblog) puts it. Intuitively, this is why you need to have high returns to make the privatization work. The returns have to be high enough to pay the frieght that they already owe to the retired generation AND replace some of the benefits that they would have otherwise received from the next generation of workers when they retired.

Krugman and others might very well be right about the numbers not working for the current proposal being floated around out there without having implausibly high stock returns. But that need not kill off the idea of private accounts. We just need to phase them in more slowly (perhaps very slowly), and with a lower cutoff age for participation.

So I reiterate: I think it would be very useful to have a discussion of the mathematics of phasing in private accounts, beginning with new 18 year old workers if necessary.

When Social Security was first instituted, it took the form of a pay-as-you-go system so that benefits could be paid out right away, before building up the vast reserve that a fully funded system requires. (Remember Ida Fuller?) We're not in that position today. The transition does not have to take place so quickly. If the present system can remain solvent for at least a couple more decades, we can take our time in phasing in a fully funded system. If that means beginning with new workers entering the system, then so be it. It will take longer to see positive results, but it should hold harmless anyone who is working today or retired today.

And I don't think that a slower, more gradual transition would require such high returns because there would be less freight to pay.

I'd like private accounts, but I'd like them done correctly. The more gradual the transition, the more the transition cost can be spread out. The sooner we start, the more gradual we can afford to be.

Angry Bear reaches back into the archives for an idea he had about a year ago.

The idea:

Here's a proposal I'd like to see analyzed: means-tested matches, on a sliding scale, to Roth and Traditional IRA contributions, with EITC recipients getting some amount match-free.

And his comment today:

Amidst all of the talk over privatization, I've yet to hear much along these lines. But this is a form of "ownership society" I could support -- as long as (1) it is in addition to Social Security, (2) it is funded out of the general fund and not Social Security taxes, and (3) it is funded in a revenue-neutral fashion by, e.g., partially reinstating the estate tax.

I'm inclined to think that this sort of thing could turn out to be a little expensive, and if it were implemented I think there would eventually be a discussion about reducing Social Security benefits as people increasingly rely on these accounts.

But I'll agree with him on one thing. I would like to see it analyzed.

Back in 1998, Timothy Cogley at the San Francisco Fed described a transition to a fully funded system. When I have time, I might compare and contrast this with the President's plan or with my ideal plan. In this short "weekly letter" format, Cogley leaves out some of the details, but he does indicate that equities would be part of the system. Something along these lines could be workable, even though there are still a lot of blanks to fill in. More on this later. In the meantime, study your (recent) history.

PGL, whose posts I frequently read at Angry Bear, was kind enough to read my blog and comment on my last Social Security post. This weekend, I was preoccupied with a combination of work and college basketball, but I have been meaning take this up again. You can read the original post, with the comment here.

Basically, my point is that if bonds (like TIPS) were available for people to put into their private accounts, they probably would. Private accounts would become diversified. Then, Krugman's argument that there would have to be some fool out there willing to sell stocks to those private accounts loses some of its bite.

PGL writes that if people already have balanced portfolios, including the bonds implicit in the Trust Fund and their 401(k)s and they suddenly receive the Trust Fund, they will not buy stocks. The implication is that they would buy bonds, effectively replacing the bonds in the Trust Fund and keeping their risk/return ratio the same.

Our logic is pretty much the same. If Social Security was fully funded, I would absolutely 100% agree. But it's not. So I don't. Not 100% anyway. I have to ask myself this question. Suppose the government gave me a $1000 refund on my FICA tax for this year and offered me two options. 1) Give it back to Social Security and forget the whole thing ever happened or 2) buy a $1000 government bond (TIPS, so I don't have to worry about inflation) and keep rolling it over until I retire.

I choose door number 2. In a heartbeat. In other words, I don't think that the bonds implicit in the Trust Fund are the same as a bond in my hand (or my private account). If I did think that they were the same, I wouldn't care--option 1 and option 2 would be equivalent. (UPDATE/CLARIFICATION: It's not by a wide margin that I choose door number 2, but I would choose it. The implication that has for PGL's comment is that while I don't agree 100%, I think he's close. Close enough that there should be more discussion on this point.)

All of this makes me think that perhaps the best fix for Social Security would be a very gradual transition to a fully funded system restricted to holding government bonds. The outcome of a pay-as-you-go system like we have is determined by demographics and Congress. The former we have a fairly good handle on; the latter is unpredictable. After all, Congress switched from price indexing to wage indexing in 1977. Bush wants to switch us back. Both represent efforts to curb the growth of benefits (first in a high inflation environment, second in a low inflation environment). Nothing says they couldn't change it again, or do something else. Pay-as-you-go encourages both political gamesmanship and sloppy government accounting.

Ok. That said, the debate is heating up. I'm generally pretty open to well thought out ideas for privatization (though I do not endorse any specific proposals yet). And I think the wage/price indexing debate should be kept separate and thought out very carefully before we do anything. I realize that separating the two issues might make it harder to enact the reform that the Bush administration seems to want. (And I accept that.)

And because I think a good debate is important, I am grateful for PGL's comment, and for his recent posts at Angry Bear here and here. Also deserving mention are posts by Angry Bear and The Lowest Deep. The last of these is worth a careful look if you are really into this debate.

Krugman in today's New York Times. Here's the core of his argument.

The whole scheme ignores the most basic principle of economics: there is no free lunch.

There are several ways to explain why this particular lunch isn't free, but the clearest comes from Michael Kinsley, editorial and opinion editor of The Los Angeles Times. He points out that the math of Bush-style privatization works only if you assume both that stocks are a much better investment than government bonds and that somebody out there in the private sector will nonetheless sell those private accounts lots of stocks while buying lots of government bonds.

So privatizers are in effect asserting that politicians are smart - they know that stocks are a much better investment than bonds - while private investors are stupid, and will swap their valuable stocks for much less valuable government bonds. Isn't such an assertion very peculiar coming from people who claim to trust markets?

There's a pattern emerging here. Like Roger Lowenstein, Krugman is appealing to the argument that Bush-style privatization involves giving up government bonds to pick up stocks. Krugman is simply looking at the other side of the transaction, wondering if there is anyone out there stupid enough to give up the stocks and buy the less valuable bonds.

Remember, the Social Security "trust fund" owns the bonds, individuals don't. So don't be fooled by the wording. Krugman, of course, knows this. What he's really worried about is the impact on government finance when the trust fund's demand for bonds gets smaller. (You can say this a number of ways--the "cost of the transition to privatization" for example.) This has been a common theme of Krugman's recent columns, and this is just another way of putting it. But at the heart of the matter, this is as much a political problem as an economic one.

The whole discussion ignores the fact that prices will adjust.

Consider this thought experiment. There is a private investor (call him a market maker or a broker, if you want) who holds a stock and bond portfolio he thinks is optimal. He is just indifferent between buying stocks and bonds at current prices.

Then, privatization happens. Workers want to buy stocks to put into their individual accounts. Simultaneously, the trust fund reduces its demand for bonds. Our private investor (or broker) helps the market to clear, selling the stocks for a higher price to be compensated for mopping up the excess supply of bonds (whose price has fallen).

Does the story end here? Not necessarily. If the price of stocks goes up too high the workers might not want to buy as many. Those workers might even want to buy some bonds (at lower prices the bond yields have gone up). The worker's portfolio might even start to resemble that of a well diversified investor.

Am I oversimplifying? Maybe a little. But I maintain that a lot of the concerns that people have about privatization would be less of a problem if workers could hold stocks and bonds in their individual accounts.

We definitely have to get away from this stock vs. bond comparison. Currently, you don't hold the bonds, the governement sells them to itself. How the government will deal with a little less demand from itself is a question that will have to be answered someday even without privatization. Wouldn't it be better to try to figure that out gradually over time when there is plenty of time and we're not at a crisis point?

UPDATE: Brad DeLong also thinks Krugman is too harsh, for somewhat different reasons. And he's with me on diversified portfolios. DeLong concludes:

The problem is that bad portfolio choice and administrative costs can easily eat up all of those gains. Private accounts should be placed in well-diversified portfolios, should not be altered in response to fads, and should be administered at extremely low management fees. Thus it is with a sinking feeling that I learn that Fidelity Investments thinks it has a dog in the private accounts fight. The kind of private accounts that raise Fidelity's stock price--and thus the private accounts that the Bush political machine has told inside supporters will be coming--are not the kind of private accounts we need.

This brings up an important question that has received scant attention (so far we've just been attacking or defending the very idea of privatization). Who will run the accounts? I hope it's more than Fidelity. I hope there is some independent oversight. I hope, I hope...

Point taken.

Via Tyler Cowen at Marginal Revolution:

One proposal for social security reform involves heavy government investment in equity markets. In a previous post, I questioned how much this would capture in the way of superior returns. My second worry is whether such a scheme would avoid politicization. After all, can we trust government as a large shareholder? Will government start controlling corporations for short-term political gain? Should our government have owned Enron and Philip Morris?

Brad DeLong has suggested government equity ownership along the lines of the Federal Reserve. An impartial panel of experts would make decisions concerning ownership, voting rights, and portfolio investments.

He then goes on to list his concerns.

So... Does anyone remember how Social Security reform movement fizzled out last time?

Brief synopsis: Republicans and some Democrats (most notably Daniel Patrick Moynihan and Bob Kerrey) began looking for a way to shore up Social Security. Partial privatization and individual accounts were gaining popularity. Being more worried about the deficit, Clinton was against this at first. But, as the late 90s arrived and the deficit became a memory, Clinton (and Gore) reversed course somewhat. Towards the end of his presidency, Clinton was actually in favor of a proposal to invest the Social Security "trust fund" in the stock market. No private accounts, just let the government invest in the stock market.

The concerns voiced by many Republicans and assorted other free marketers then were much like those that Tyler Cowen voices in his post today.

Was Clinton's embrace of that type of reform a clever attempt to muddy the waters and derail the whole thing? You decide. It happened pretty late in his term. And if Gore had won in 2000, he never would have gone through with it. Anyway, if that was the idea, it worked. Gore had a little explaining to do when he ran for president and advocated a "lock box" but let go of the idea of investing in the stock market. But in the end, I don't think that's what hurt him.

If there is a serious proposal on the table to let the government hold equities as a part of Social Security reform, there is a very real possibility of history repeating itself. This bears careful watching.

Once more for the record, I don't think that Social Security is in "crisis." I've said so before. However, the system will need some kind of adjustment in the next few decades. Almost no one disputes that.

So wouldn't it be nice to take prudent steps now that might have a long run payoff? (One doesn't have to think that the system is in crisis to agree to that.) I'm not sure that the plan that is being discussed is exactly what I want. My ideal plan would probably be more cautious in some ways and more daring in others. I would just like to preserve the insurance aspect of Social Security while improving the rate of return to younger workers and leaving the return to retired (and nearly retired) people unharmed.

But for now, chew on this NY Times Magazine article by Roger Lowenstein.

Even though I disagree with much of it, and it has some problems (see below). I did learn a thing or two. Read it, but read it critically.

Here's an amusing part, even though it hurt a bit:

Politicians and other commentators tend to speak about these long-range trends, or at least about Social Security's finances, with an air of precision. This is almost amusing, since few economists can predict the swings in the federal budget even a year in advance.

I pointed out in an op-ed a few years ago that the cumulative forecast error in the federal budget from about 1997 to 2001 (based on a 1997 forecast) was almost half as large as the annual budget itself. The difference between the Social Security Administration's optimistic and pessimistic forecasts can be almost as bad. Forecasting is hard. We all know that. That shouldn't stop us from trying, though.

Then there's this:

Conservative economists say the figure [how many people would be below the poverty line if they didn't have Social Security] is irrelevant: if Social Security didn't exist, people would save more. This may be true of economists, but what about the rest of us? The argument illustrates the ideological agenda of those who favor privatization: they want to change people's behavior.

That paragraph is confused. Start at the end. Those who favor privatization want to change people's behavior. (Imagine! I mean, certainly no other policy advocate ever wanted to (gulp!) change behavior!) But... the only people who would save more if Social Security didn't exist would be the economists. And yet that doesn't stop policymakers from trying to coerce the non-economists to conform to their desired behavior.

That just doesn't make sense.

Mr. Lowenstein has a low opinion of our (the economists') ability to forecast, but apparently he thinks we would be the only ones smart enough to save more for retirement if Social Security were not there.

The Lucas Critique is like the Law of Gravity--you don't have to know how and why it works for it to affect you. (I should put that on a bumper sticker!)

One rationale for privatization is that workers would get a better return on their money in Wall Street securities than with Social Security's dowdy old Treasuries.

I don't think I'm taking this out of context. (It's on page 8 if you want to see for yourself how it fits into the whole article.) He seems to be doing one of two things, either he's totally missing the boat concerning how Social Security works or he's comparing "privatization" with "fully funded Social Security." Lowenstein seems quite knowledgeable about Social Security, so I don't think it's the former. Yet, the article certainly doesn't hold up a fully funded system where workers actually have claim to government bonds as an option. Truthfully, I'd take the return on "dowdy old Treasuries" in a fully funded system over pay-as-you-go Social Security any day. I even made a post a few days ago that said I would like to see TIPS as a safe choice for private accounts.

Much is being made about the Bush administration's proposal to shift from wage indexing of social security benefits to price indexing. See, for example, here.

And there is the now famous Peter Wehner memo that has been posted on many blogs, including Brad DeLong. One paragraph reads:

It's worth noting that wage indexation was not part of the original design of Social Security. The current method of wage indexation was created in 1977, under (you guessed it) the Carter Administration. Wage indexation makes it impossible to "grow our way" out of the Social Security problem. If the economy grows faster and wages rise, this produces more tax revenue. But the faster wage growth also means that we owe more in Social Security benefits. This has produced a never-ending cycle of higher tax burdens, even during periods of robust economic growth. It is the classic case of the dog chasing his tail around the tree; he can run faster and faster, and never make any progress.

Given all the discussion of the shift to price indexing, I thought I would look at how wage indexing came about. The memo is correct. It did happen in 1977. The following is from the New York Times, May 10, 1977 (page 55):

Inadvertently, as it turned out, Congress in 1972 created what virtually all analysts regard as an overly generous inflation-adjustment formula for calculating the initial benefits of newly retired persons. It takes account of both wage inflation and price inflation.

Breaking that link is called "decoupling." Mr. Carter proposed that Congress do this by eliminating price rises from the calculation of initial benefits and by using a wage-ratio formula that would maintain the ratio of benefits to final pre-retirement earnings at the present 45 percent. Thereafter, benefits would escalate with the Consumer Price Index, as they do now.

Without decoupling, the Administration said, by the year 2020 some retired persons might be drawing benefits at a 60 percent ratio. Some might draw more than 100 percent of their earnings in their last working year.

And then there is the September 10, 1977 New York Times (page 8):

In addition the Republicans also proposed a new formula for calculating initial benefits of retired persons that would undo the over-compensation for inflation that Congress adopted in 1972. All sides agree that this formula must be changed.

...

Joseph A. Califano Jr., the Secretary of Health, Education and Welfare, criticized the Republican decoupling formula on the ground that it would lead to initial retirement benefits in the future 6 percent below what the present formula would produce.

The Republicans have said as much. They maintain that such an adjustment is fair because the 1972 formula led to an increase in benefits that was 6 percent greater than the increase in the cost of living.

However, the Republicans contend that no one who retired before the formula was changed would suffer a reduction in benefits, nor would any future beneficiary have to accept less than he would have been entitled to under the pre-1972 formula.

So, yes it was during the Carter administration, but it was bi-partisan. And truthfully, in looking at the news accounts of the time, people were much more concerned about the changes in the payroll tax rates than the change in the indexing method. The bottom line is that between 1972 and 1977 inflation caught Social Security between a rock and a hard place. Wage indexing was a way to reduce the burden on the system caused by inflation. It worked. Then the pendulum swung back in the other direction. Today, we find ourselves in an equal, but opposite situation.

In any case, wage indexing was not a plan to expand benefits. Quite the contrary, it was a plan to slow the growth of benefits in a time of high inflation.

Tyler Cowen at Marginal Revolution has a defense of the Bush proposal to freeze benefits in real terms. I'm not sure I'm ready to, as he puts it, "push the 'yes' button" on this just yet. That said, a compromise might be in order here. The formula does not have to be all or nothing in either direction. Honest folks on both sides should take a look at how to moderate the increases in benefits without reverting to what some might see as Draconian cuts. As Tyler suggests, the big problem is medical care, and there are probably better ways to solve that problem than by growing Social Security. I'll give it a cautious "yes" with a heartfelt plea for compromise.

My only problem with the change is the extent to which it may affect workers not yet at retirement, but who have read their annual statement from Social Security and used it in their retirement planning. I would be much more comfortable with an indexing change that affects new workers or very young workers so that accurate expectations may be formed. Ideally, if the private accounts are only phased in for the younger workers, I would start the indexing change there as well. Expectations matter. Start reform with the young, and if it works, expand it over decades.

Dave at macroblog posts an excellent response to my question on voluntary private accounts for Social Security. He concludes his post with the following:

The rub, of course, was that the results from this type of exercise depend -- sometimes critically -- on the assumptions that are made. Modesty (or self-defense) requires me to conclude that making a definitive judgment about the true return to the typical private account, for example, is just a bit beyond the rank of your standard well-meaning policymaker. The solution, so it seems to me still, is to make participation voluntary, and let people sort themselves into the plan that they deem best. If our guess is right, we will see a systematic sorting of younger people into the privatized system, and the pay-as-you-go world will slowly fade away. If they don't, well maybe the critics of privatization are right.

There will, of course, be an element of trial-and-error in all of this. Those in the privatized system still have to help pay the freight for the existing pay-as-you-go liabilities, and it will not be immediately apparent what tax rates settings will do the trick. That will be revealed in time, and adjustments will have to be made. But being a good conservative commentator, I live by a simple creed. Give the market a chance. It will probably give you the right answer.

Well said. The part about those in the privatized system paying the freight for the existing pay-as-you-go liabilities (including those young workers joining the system who choose pay-as-you-go over private accounts) is exactly what I'm concerned about. The more people who opt out of private accounts, the more freight there is to pay. I have confidence in the ability of the government's actuaries to work this out through a combination of the Law of Large Numbers and trial and error, but I think a split system would add significantly to the cost of administering the system as well as lowering the overall returns to the participants.

If the idea of voluntary accounts is to give people a choice over where to put their contributions (with the existing system being the safe choice), I would take a different approach. Give people the option of putting their contribution into TIPS (Treasury inflation protected securities). Risk would be extremely low, and the return would almost certainly dominate that of the pay-as-you-go system. If the government could guarantee that the return from TIPS would dominate what they would have received in the old system, we could dispense with the old system right away at least for brand new workers contributing for the first time.

I think this is the right kind of debate to have. I am looking forward to seeing more specifics from the administration and more discussion of the finer points.

UPDATE/CLARIFICATION: When I say "dispense with," I don't mean the whole system, of course, since the administration is only suggesting that part of the payroll tax would be allowed to go into the private accounts. In effect, what I want is for that portion of the payroll tax to be permanently and totally detached from the pay-as-you-go system.

UPDATE/CLARIFICATION: I'm assuming about a 1.8% real return on pay-as-you-go Social Security as suggested by Gokhale and Lansing. Real yields on TIPS less than that over a long period would be pretty unlikely--if it ever happened, we'd have more problems than just Social Security.

I see that the Bush administration's Social Security proposal makes the personal accounts voluntary.

The President favors voluntary personal accounts as part of a comprehensive solution to give younger workers the option to save some of these payroll taxes. Personal accounts give younger workers the opportunity to receive higher benefits than the current system can afford to pay, and provide ownership, choice, and the opportunity for workers to build a nest egg for their retirement and to pass it on to their spouse or their children.

Those who do not choose to have a personal account would continue to draw benefits as Americans have long done from the Social Security program.

Personal accounts will provide Americans who choose to participate with an opportunity to share in the benefits of economic growth by participating in markets through sound investments.

See... This is what happens when economic policy goes through the political sausage maker. "Voluntary" sounds so nice and politically correct. It's less threatening, perhaps.

But having voluntary private accounts makes it sound like you're comingling a fully funded and a pay-as-you-go system, doesn't it?

Will someone please explain to me how today's young people who choose to stay in the existing pay-as-you-go system can expect to someday receive the same level of benefits that are paid today if some in the generation after them are in a fully funded system?

Higher taxes, I suppose. How does this fix anything?

And doesn't this make the actuarial accounting of the system more nightmarish than it already is?

I'm generally not opposed to doing something to improve the system. (see previous post) However, the voluntary aspect of the proposal on the table sounds like trouble to me. I have to this point seen no evidence that my concerns have been thought through by the administration.

Any thoughts?

Yes and no. Consider these two claims:

Could the right sort of adjustment/modification/reform of Social Security in the next few years make future retired persons better off? Almost without question, yes.

Is Social Security on the verge of becoming the nation's biggest fiscal problem if it's not fixed in this presidential term of office? Definitely no.

Even the AARP might agree with my first claim, although the latest out of the White House sounds like they might take issue with my second claim. In any case, these two claims are fairly well accepted by economists, policymakers, and pundits on both sides of the debate. So what should be done?

Let's get one thing straight. I would not be in favor of any reform that would mean a reduction of benefits for today's retirees. I don't think anyone would favor that. Hence, it won't happen. Today's retirees need not fear. Whatever reform takes place, it will (possibly) help future retirees (today's kids and those not yet born--I fear that I may even be too old to benefit, more on that later).

I also would leave the disabilty and SSI benefits pretty much the way they are. What we are talking about here is the basic retirement aspect of Social Security.

I tend to agree with macroblog on this one:

I fully concede that you cannot take any old private-account scheme off the shelf and claim that it dominates the current system. But I think it should also be conceded that it is conceivable that a privatized system might dominate, if properly constructed.

And then he recalls some research that he and Jagadeesh Gokhale did on the issue.

If we allowed people below some critical age the opportunity to shift to private accounts, while at the same time taxing them to pay promised liabilities to those who remain in the system, would it be in their interest to do so? Our answer: yeah, maybe.

I've been a supporter of their (Altig and Gokhale's) idea since before there was a macroblog. I read their research from the Cleveland Fed when it first appeared in the mid 1990s. Read here for one of their articles much like the Cato piece referenced above. Read here for one by Gokhale and Lansing from about the same time. The latter has nice historical graphs, but it is getting a little dated.

Articles like these convinced me about 9 years ago that Social Security could be fixed if we really wanted to. We just haven't wanted to. In 1996, Altig and Gokhale reckoned that transitioning everyone under the age of 43 to a privatized system could potentially be Pareto improving. In the 1997 Cato piece, they put the cutoff at age 32. At the end of the summary of that piece, they say if we wait until 2011, only those under the age of 20 could move to the new system. We're halfway there. As I suggested above, I think it's too late for me.

The reason why the window is closing on this type of reform is that in the next few years the Social Security surplus will begin to shrink. Because Social Security contributions have been added to the government's general fund for some time, the disappearance of these funds will be noticed. In order to support the benefits of retirees and those near retirement (all those baby boomers), we will need to draw on more of the current contributions (from younger and younger workers). Once the Social Security surplus is gone (2018, by their own estimate), the transition becomes more difficult and costly. I don't see how anyone can look at the figures and the charts and fail to see that the time do something that would be Pareto improving is sooner rather than later.

Good solid economic growth with some fiscal responsibility will extend the life of the system for decades even if we do nothing. We don't have to do anything (except maybe raise the retirement age by one year every decade or something similar if life expectancies keep rising). But if it is possible to do something Pareto improving, I think we should do something.

But I also concede that you can't just take any old privatization scheme off the shelf and say that it will be a Pareto improvement.