by reader coberly


We often hear claims that the “rate of return on investment” of Social Security is less than what a person could realize by investing “in the market.” So much so that some authors claim that Social Security has created a “legacy debt” that amounts to a “backward transfer” of wealth from future generations to past generations.

This claim is as bizarre as it sounds. No future taxpayer will be giving money to his long dead grandparents. But it is also fundamentally wrong headed. Social Security is not an investment club in which returns paid to some subscribers result in less money available to earn interest to pay other subscribers. Social Security is an insurance policy, and the costs and “returns” of an insurance policy must reflect the real world costs of the insured event. They cannot be a simple expression of an arbitrarily chosen interest rate.

Calculations of “rate of return” usually leave out important complications like fees, taxes, inflation, and market variations. And they never refer to the insurance function whereby Social Security protects you from disability and your family from your early death. And no consideration whatever is given to the accidents of health or employment that could leave you unable to save enough through market investments to pay for even a marginally adequate retirement.

And finally, these comparisons always ignore the fact that as an insurance policy, there is no one “rate of return” that adequately describes the facts of Social Security. My purpose here is to make a beginning to addressing that lack of understanding. Ultimately there are too many complications for a short essay to address all the possibilities, but here is a start:

I’ll explain the method below, but here are the results.

An employee earning a below average wage of 26000 per year (adjusted for inflation and average real wage growth), in order to get the benefits Social Security will guarantee to him and his wife, would have to invest his 6.2% payroll tax and get a steady 12% return on investment per year every year above taxes and fees.

That same employee without a wife would have to get 10%.

A self employed person, investing the entire 12.4% payroll tax, would have to get 8.5% if married, 6.3 % if single.

An employee earning the average wage of 44,000 per year, adjusted, would have to get 10.4% married, 8.5% single; self employed married 6.7%, self employed single 4.5%.

An employee earning 70,000 per year, adjusted would have to get 9.2% married, 7.3% single; self employed married 5.5%, 3% single.


I assumed an average wage equal to the “average wage index” given in Table V.C1, page 98 of the 2007 Trustees Report, for each year from 1975 through 2009. This is nearly equivalent to starting at $8600 per year in 1975 and increasing 5% each year after that. For the low wage worker, I assumed 60% of the average. For the high wage worker I assumed 160% of the average. In each case I assumed an investment each year equivalent to either 6.2% of the wage for “employee” or 12.4% for the self employed.

I calculated benefits based on the average adjusted wage, which is the same as the final wage in this exercise, using the bend points in Figure V.C1, page 100 of the 2007 Trustees Report. I assumed the retiree would be able to buy an annuity using all of his savings plus return on investment calculated above. I assumed that annuity would earn 8% and pay the Social Security calculated benefit plus 3% inflation adjustment each year for a life expectancy of 15 years following retirement. I assumed that a married retiree would receive an extra 50% spouse allowance.


Since I looked at only the required 35 years for the Social Security calculation, some would argue that I neglected the money the employee would have earned, and paid taxes on, over a greater number of years. On the other hand, since early wages tend to be lower in relation to the average wage than later wages, by holding the wage constant, as a function of the average wage, I very likely overestimated the earnings from the early years, those that would have the greatest increase due to compounding. I did not count wages earned by the spouse, but neither did I count taxes and fees, possible inflation surges and market dips. Nor did I consider that actual life expectancy is even now somewhat longer than 15 years after retirement.

The largest source of contention, I suspect, would be whether or not the employer’s share of Social Security taxes would be available for the employee to invest if there was no tax. My view is that while it is reasonable to guess that they would be, obviously for the self employed, probably for high earners with bargaining power, or unionized employees with bargaining power, it is extremely unlikely they would be for low wage workers who have no bargaining power. And it is even more unlikely that if those taxes magically turned into a 6% wage hike for the low wage earner, that he would save and invest them. A worker making 400 dollars per week is not going to invest 50 dollars on the market even if you give him a 25 dollar raise. In any case according to the law, and in historical fact, the employer’s share does NOT come out of the employee’s earnings. I prefer to deal with reality rather than the woulda shoulda coulda of even the best economists’ imaginations.

And the whole point of Social Security is to provide retirement security for those workers who, for whatever reason, end up after a lifetime of work without enough money saved to pay for a retirement that includes a roof and groceries. Since this could be you… even now… if you are currently enjoying high expectations, instead of obsessing about what you “could have” earned “if only,” let me suggest you pour a glass of good wine and rejoice that you are going to at least get your money back adjusted for inflation. And all in all it is better to pay a little for insurance you don’t “need” than to not pay for insurance you end up wishing you had. In fact, if the universal mandatory automobile insurance requirements are a guide for thinking, there is some reason to suppose that it is better for you that other people have insurance, even if you yourself will never need it.


A low wage employee with a stay at home wife would have to get more than 12% “rate of return” from investing his Social Security tax to get the same retirement income that Social Security guarantees him. This is based on current tax and benefit schedules and historical data that reflect roughly a 3% inflation rate and a 2% real income growth. An unmarried, self employed, average earner would have to earn 4.5% every year on his savings of 12.4% of his income in order to get the retirement income that Social Security guarantees him. The rate of return is highly dependent on income level, with rates of return being much higher for people whose lifetime earnings fall below average. But even a single, self employed, high earner will at least get his taxes back adjusted for inflation. Social Security is insurance, not an investment, but its “rate of return” is quite competitive with real world investments.
by reader coberly