A Modest Proposal

This proposal is for a new retirement system. This is not a reform of the current Social Security system, but rather a complete replacement of it.

How would it work?

The basic idea is this: At birth every child who is a natural born citizen citizen¹ would have the government put $8,000 into a retirement fund. The money in that fund would be invested in an index fund such as an S&P 500 or Russell 3000 index fund (I would prefer a Russell 3000 fund). This takes advantage of both the high rate of return of these index funds and the power of compounding when started at the earliest possible age. For example, if the future average real annual growth in the fund is 6%, then for every $1,000 invested there would be, in inflation adjusted terms, ~$44,000 65 years later.

What would be the amount of the annual benefit?

The amount of the annual benefit would be determined by the fund administrators. The administrators would set the amount based of the total amount in the fund relative to the number of recipients. That would make this system a cross-breed between a traditional company pension fund, which would be called a defined benefit plan, and a retirement system based on IRAs and 401ks, which would be called a defined contribution plan. The plan administrators should be somewhat conservative in setting the amount of the annual benefit so as to allow for downturns in the economy when the total principal amount in the fund would decrease. At regular intervals the plan administrators would review the health of the fund and increase amount of the payments as appropriate.

Why an index fund?

A 2010 study showed that an S&P 500 index fund outperformed 99.4 % of all mutual funds. Historically the Russell 3000 index has modestly outperformed the S&P 500 index. So it stands to reason that a Russell 3000 index fund would outperform a slightly larger percentage of all mutual funds.

What would be the expected future annual growth rate?

Good question. The Russell 3000 index went from 888.89 on January 3, 1995 to 6792.93 on January 3, 2017. Inflation for the same period was 61.57%. This results in an annual real growth rate 6.19%. Since January 1, 2013 the index has  increased by more than 65%, more than twice the average previous rate, with no significant change in the economy to justify such a large increase. Now, if you include the period since 2013 in the long term average you get an annual growth rate of over 8%. If you don’t include 2013 you get the annual rate of 6.19% referenced above. 6.19% seems a little low while 8% is definitely too high, especially when you consider that the average PE ratio for S&P 500 is more 24. Since the long term average for the S&P 500 PE ration is around 16 it would seem the the stock market is currently overpriced. So I will make an educated guess and say 6.5%, a little above 6.19% and well below 8%. The reality is that there is so much uncertainty about the future that any such estimate would have a huge error bar associated with it.

Why a fund and not an individual account?

It seems reasonable to assume that even smaller a percentage of individuals who are not financial managers would be able to outperform an S&P 500 index fund. An individual account which would allow trading would simply be an excuse for the banks and/or Wall Street to syphon off 1% of the principal in the fund annually. I did consider individual account without a trading option which would pay benefits based on the principal amount in the account, but I realized there was a problem with that idea. Consider two children. One is born on December 9, 2007, the day the stock market peaked just before the housing bubble broke. The other is born on March 9, 2009, the day the market bottomed-out following the bubble. At that point the first child’s principal amount would have declined to $4,292.37. From that point on they would have exactly the same growth in their accounts, but the first child would end up with only 43% of the annual benefit of the second child. In fact, it would be even worse since, if they both retire at the same age, the second child would benefit from an additional 15 months of growth in the fund. It would be very unfair to have such a large difference in the benefit received based solely a person’s date of birth. The only way I can see that would allow for people retiring at the same age receiving the same benefit is for all of the initial money to go into a fund.

Inheritability

One of the key features of this proposed system is that the benefit would be inheritable. But instead of inheriting the annual payments, the heirs would inherit the amount of principal necessary to pay that benefit. I would propose that the benefit would only be inheritable by the children of the person receiving the benefit. That would mean that not only would everyone covered by this plan would have a decent income in retirement, buy also that the second generation of people covered by this plan would inherit a reasonable amount of money from their parents.

Funding of the new system

In the long term the funding of this proposed system would be much easier than the current system, requiring an employee payroll tax between .6 and .8 per cent with no employer contribution needed. Now, when I say long term, I mean long term – 65 to 70 years. This is because of the need to continue funding the current system until the last group of people eligible for the current system reach retirement age. At that point the payroll tax can be gradually reduced for employees and gradually eliminated for employers. But that is for the long term. For the short term there is a very real funding problem since the first people entering the new system will not enter the work force, and thus start paying the payroll tax, for some number of years.

I can think of a number of ways of solving this initial funding problem.

One is to have the Treasury loan the money to the new trust fund, to be repaid as the people in the new system start paying the payroll tax. This has the advantage of requiring no money other than that coming from the employee payroll tax. It has the disadvantage of diverting money away from the current OASI trust fund, which already has funding problems to resolve.

Another is to increase the payroll tax by .5%, .a .25% increase for both the employee and employer. This is probably not a good idea in a sluggish economy and is also probably a political non-starter.

The third way is to initially fund the system from general revenue. This is the option which I would prefer. This would divert no money away from the current OASI trust fund, which would mean that any funding problems with that trust fund are problems they would have anyway. The obvious problem is that this would be a new spending program. What would be unusual about this spending program is that there would be a fairly definite date in the future, albeit 65 years in the future, when the spending would end.

Transition from the current Social Security to the new one

There are actually two transitions required.

The first is a bureaucratic transition, new forms, procedures, etc. While a non-trivial problem, it is readily resolved. Bureaucrats know how to do this sort of thing and it should not be a significant problem.

The second is the financial transition. If the public funding option mentioned above is implemented, then the financial transition is trivial. The funding problems for the current OASI trust fund would be no greater than they already are. If any funding option for my proposed new system is implemented which diverts any payroll tax money away from the current system, then the funding problems for the current OASI trust fund would be greater, probably significantly greater. So while there would be no legal linkage between the new and old systems, there would be financial and political linkages.

What about other savings for retirement?

The only other retirement savings program I would keep would be the 401K program and then only to the extent that there is a company matching amount. All of the rest of the current retirement programs, including the 401K contributions beyond the amount matched by the company are really primarily beneficial to people who have sufficient income that they should not need any government assistance in their savings effort. I would encourage people to save for retirement beyond what this program would provide, it’s just that I don’t think the government should be involved in since such involvement has a strong tendency to end up benefiting the financially better off portion of the population.

Note: I have another proposal coming up which would interact with and almost certainly cause changes to this proposal. I will discuss those interactions and changes when I post the other item.

  1. The Supreme Court has never ruled on what is meant by the phrase ‘natural born citizen’. I think it is obvious, a natural born citizen is anyone who is a citizen of the United States by reason of their birth, that is, either their father or mother is a U.S. citizen or they are born in the United States.