« USA's debt burden unchanged in ten years | Main | For the middle class, one less thing to worry about »


this is one heck of an idea, and one I hope to see politicians support.

The problem with this proposal is that the current SS system has some a lot of tilts and quirks in it.

First, the monthly benefit for seniors is 90% of $761 of Average Indexed Monthly Earnings (AIME), + 32% $761 to $4,586, + 15% over $4,586. AIME is based on your 35 best years, indexed to average wages for the year (not CPI). It considers only amounts subject to OASDI contributions, which are capped, currently at about $107,000/yr. ($9000/mo.).

The result is that lower paid workers receive a much higher percentage of their base in benefits. It may be far in excess of the amount they paid in. Conversely, a high earner will receive a lower percentage of earnings as his benefit, and his rate of return will be negative.

Duplicating this benefit structure in a defined contribution plan is difficult without resort to subsidies and penalties.

I don't think you are totally wrong, but the prime issues will always be the same. How much redistribution does the system require? Who will fund the payments? The workers through FICA? or general revenue? What happens to the employers contributions? Do they go to individuals?

I am going to cut this off here because it is too long for a blog comment.

Thanks for being concise. SS contributions are based on the first $x of earnings, above which a social safety net was decided should not apply. Benefits are structured accordingly, although the math might not work out to the nth decimal. In any case, nontransferable bonds could be structured to at least minimize the benefit difference, if not eliminate it. Private ownership of the asset is the point, because it replaces private dependence on the integrity of the political class.

"In any case, nontransferable bonds could be structured to at least minimize the benefit difference, if not eliminate it."

I don't want to be a wet blanket, but I believe that if you study the Social Security Handbook, an official government document that explains benefit calculations, you will find that you can only match the old age pension portion of Social Security with a defined benefit plan, if that plan subsidizes low earning participants and penalizes high earners.


My own belief is that in order to reform the system we need to decrease benefits, and squeeze them down for people who earn more than the mean. If you who want to understand the policy options that are being discussed, the CBO did an excellent study “Social Security Policy Options” in July 2010 laying out the options, their benefits, and their costs:


Help please: Who is the "you" in "you can only match..."?

Sounds as if you're describing a private entity, such as a corporation, small business, or sole proprietorship. Correct me if that's wrong.

Some positives:

1. Converts intragovernmental debt to public debt, which restores a modicum of integrity to the budget process & spending.

2. The U.S. isn't likely to see a spendable SS surplus for a long, long time, given current economic and demographic trends. Many (most?) politicians will be out of office by the time we ever do see one. So unless someone has an ideological problem with individual/private ownership of one's retirement resources, it's hard to imagine how this idea would get much political opposition.

One quibble though: such instruments wouldn't be "earned", per se, they would be purchased using the fruits of one's labor. As such, they are property. So it's difficult to see the moral justification for non-transferability. At a minimum, they should be inheritable, IMHO.

Help please: Who is the "you" in "you can only match..."?

Would the term policy planners help instead of the second person pronoun?

What I was trying to say is that converting Social Security, which is a defined benefit plan, into a defined contribution plan, such as the one you suggest, creates difficult issues of actuarial mathematics.

The foundation of this problem is that the benefit in Social Security bears no relation to when dollars were earned (other than by better or worse inflation adjustments). In a defined contribution plan the the oldest dollars are, because of compound interest, the most valuable.

In the current design of Social Security, the first important step in calculating benefits is determining the AIME (average indexed monthly earnings). For this purpose, the person who begins his career earning X dollars per month, steadily increases those earnings (remember this is indexed so we are think about real dollars, not inflation), until he makes 2X dollars, will have the same AIME as a person who starts by making 2X and steadily sees his earnings decrease.

The same AIME would get the same benefit. But, in a savings based system the first person, the increasing earnings person would have a smaller balance than the second example, because of compound interest. The only way to equalize them is to penalize number 2 or subsidize number 1.

The Optimist system is just, and so is the existing system, but they work differently. If I were a young person now, I would want to be enrolled in the Optimist system. But, people in their 50s and 60s might object and insist on staying in the old system. The transition between systems is fraught with challenges, and one of them is political, and created by demographics. There are more people in the old system, who are closed to retirement (i.e. the Baby Boomers), than there are in younger cohorts.

To me the only way to defuse this conflict, is to stretch it out and make the transitions generational. Benefit caps, retirement age increases, and formula changes will help. We should also look at things that will make current benefits less valuable, such as eliminating the partial exemption of benefits from income taxation, and increasing our reliance on excise taxes.

Thanks, I see what you are saying. But if the new SS bonds came with zero interest (except whatever it took to index for inflation), there wouldn't be a penalty or subsidy for 1 vs 2.

I assume the problem your proposal attempts to solve is to reduce the uncertainty of whether "promised/estimated" benefits will be paid when an individual reaches the designated retirement age. The proposal would require a change in existing law, and thus could be changed again at any time as the Federal Government deems necessary. So, in that regard, it may provide emotional reassurance, but it is questionalble whether it provides real contractural guarantees since an individual can not "sue" the government unless said government consents.

In the meantime, it does nothing to resolve the financial difficulty of the existing system and provides the individual with no freedom of action to try to improve returns or provide an estate for prospective beneficiaries. Given those concerns, I wonder how valuable this approach would really be.

We really have to decide at some point whether the purpose is to provide a defined benefit retirement or a welfare safety net. Only when that is resolved can we design the best approach to fund it.


Good points -- but the same problem with guarantees exists today with Treasury securities. They are viewed as the safest financial asset on the planet because everyone trusts the U.S. govt and its track record of never (except very recently) even coming close to defaulting, not because of any ability to sue the feds for defaulting.

The current debate is one of tweaking the status quo versus govt-mandated private saving (including stock market investing if the individual so desires). If that remains the choice, it's safe to rule out the latter and stop wasting valuable debate time on it. (Frankly, I'd rather hear the Republicans debate the best ways to spur long-run growth than waste time proposing a DOA social security "alternative.")

I think 30 years at no interest would be viewed as a penalty.

I guess it depends on whether ss is viewed as an investment or a safety net.

The comments to this entry are closed.