Social Security Revisited

This essay was originally published in December 2000, updated in January 2005.

In the essay (Social Security by the Numbers,) we looked at Social Security from the point of view of the individual, specifically, “How much did I pay into it?” and “What can I expect to get out of it?” Our essay resulted in a number of comments and questions. In order to address these questions, we need to review Social Security in the aggregate, i.e. what does the whole program look like?


In Chart 1 we’ve plotted the following for each year since 1940:

1. The number of dollars workers paid into the program in taxes.

2. The number of dollars retirees received from the program in benefi ts.

3. The resulting assets in the “trust fund.”

As you can see, the program currently looks pretty good. There is close to $1.2 trillion in the trust fund, roughly three years worth of benefits. This looks impressive until you realize that most people receive benefits for 20 years, not three.

Most of you know that we don’t like to make projections, but in the case of Social Security, it’s pretty easy.

The benefi ts to be paid out each year will equal the number of retirees multiplied by the benefi ts promised to them.

The taxes paid in each year will equal the number of eligible workers, multiplied by the employment rate, multiplied by the withholding rate on their salaries up to a stated level.

Since nearly all the future retirees are already in the workforce and nearly all the eligible workers for the next 20 years have already been born, it’s fairly easy to project the taxes, benefits and assets for the next 20-40 years. In fact, our projections look just like the projections of the Social Security Administration (SSA). The projections are included in Chart 2.

Chart 2 shows why today’s retirees over the age of 65 don’t have a problem; there will be sizable assets in the program for the next 30 years.

But Chart 2 also shows why the children of today’s retirees, 30-40 year-olds, do have a problem: the program runs out of money in 2035.

Why has the program worked so far? When the Social Security program was initiated in 1937, the average life expectancy in the U.S. was less than 65 years. Eligibility for benefi ts was set at age 65 in the expectation that fewer than half of the workers would collect Social Security (because they wouldn’t live long enough). Furthermore, when the program started, there were a lot of workers paying into the program and few receiving benefits.

In 1945, the ratio of workers-to-retirees was over 40 to 1; in 1950, over 16 to 1; and in 1960, the ratio was 5 to 1.

Today, the worker-to retiree ratio is a little over 3 to 1. As the worker-to-retiree ratio fell, the SSA found it necessary to raise the tax rate from 2% in 1937-1949 (1% employee + 1% employer) to 6% by 1960 to 12.4% in 1990. The 12.4% rate remains today.

Furthermore, the SSA found it necessary to raise the level of wages on which the tax is paid from $3,000 in 1937 ($38,400 in 2004 inflation-adjusted dollars) to $87,900 in 2004.

We’ve plotted the applicable tax rate on Chart 3. We’ve also marked the years when the combination of tax rate and rate base first pushed the top payer over the levels of 2, 4, 6, 8 and 10 thousand dollars (all numbers inflation-adjusted). Note that Chart 4 is a continuation of Chart 3, simply with a change in scale.




Also, starting in 1984, the SSA started cutting the value of retiree benefits. In 1984 they started taxing benefits; they started taxing 50% of the benefit and now tax 85% of the benefi t. More recently, they’ve been raising the retirement age.

I was born in 1944; the age for me to qualify for full Social Security benefits is 66 years, not 65. If you were born in 1960, the age at which you will qualify for full benefi ts is 67 years, not 65.

Social Security benefi ts are calculated as a percentage of your qualifying pay prior to retirement. Currently, benefi ts are calculated at 90% of the first $627 of his/her average indexed monthly earnings; plus 32% of his/her average indexed monthly earnings over $627 and through $3,799; plus 15% of his/her average indexed monthly earnings over $3,779. Since the average wage earner today earns roughly $34,731 per year ($2,894 monthly), the average retiree is promised benefits a little over 45% of their pay. See Table 1 for details.




Three people each paying 12% in taxes can support one person taking 36% out in benefi ts. (Note that in 1960, five people each paying 6% in taxes could support one person taking 30% out in benefits.) The problem is that the ratio of workers toretiree will decline to just 2:1 by 2030.

At that point, the two workers would each have to pay 18% of their pay (nearly a 50% increase) into Social Security in order for one retiree to receive 36%. Some people believe this is a viable solution. I don’t. In the 1970s, I saw what happens when people are pushed into ever-higher tax brackets. At some point, they quit work. Even if the employees want to work – if the employer quits, the employees are out of work.

A second solution is to cut benefits. (In fact, I’ve suggested cutting the benefits of millionaires, but there aren’t enough millionaires to solve the problem.) Many retirees believe that their benefits can’t be cut. They believe they’re entitled to the promised benefts. But the Supreme Court (Fleming vs. Nestor, 1960) has ruled that we’re not entitled to the promised benefits. Some retirees were shocked when Medicare tripled the amount (from $10 to $30) of the co-pay for prescribed drugs. Folks, the rules on Social Security are set by the same people who make the rules on Medicare. If they believe it is necessary, they will cut your benefits.

But there is a third way to make Social Security viable for the next generation.

If a part of the taxes used to build the trust fund for the next 30 years could be invested to earn a reasonable rate of return, we could alleviate the problem. Some have suggested allowing people to invest part of their Social Security taxes in a Personal Social Security account. To me, a Personal Social Security account sounds a lot like an IRA (call it a PSA). So I took a look at my IRA to see how it has done.



Note: Within the above illustration, performance is net of all fees. Click here to see quarter-ending standardized returns for the Muhlenkamp Fund.

Chart 5 is a plot of my personal IRA from 1980 through 2004. The bottom line is the total dollars I’ve paid in, 24-years multiplied by $2,000 per year is equal to $48,000. The middle lines are calculated: they show the assets I’d have if I’d earned 2%, 6%, or 9% per year. The top line is what my account has actually done. Chart 6 simply extends Chart 5 out another 20 years to show a typical working lifespan of 44 years.

The Social Security Administration recently sent me a statement which said my promised benefit upon retirement is $21,924 per year. The IRS says my life expectancy at age 66 is 16 years. So the SSA expects to pay me $350,784 over my retirement years. We’ve marked that on Chart 6.

Alternatively, an annuity which promised me $21,924 for 16 years, would cost $240,000 at the start, if we assume an interest rate of 5%. We’ve marked that on Chart 6.



You’ll note that at contributions of $2,000 per year, the return has to be 6% or greater to reach $350,000 in 44 years but that it reaches $240,000 in 37 years. I’ve exceeded $240,000 in 24 years. The amazing thing is that the $2,000 per year that I put into my IRA is less than 30% of what I’ve paid into Social Security in the same period of time. It’s also less than 25% of what I’ve paid into Social Security to date. So just by earning a reasonable return on my investment (it’s been invested first in the Windsor Fund and later in the Muhlenkamp Fund), I will be able to fund an amount equal to my promised Social Security benefits with only 30% of the Social Security taxes. This makes 70% of my Social Security taxes available to someone else.

From the above data I reach several conclusions:

1. The 65+ year olds don’t have a problem, their children do.

2. Using Personal Security Accounts for a part of the taxes can help alleviate the problem.

3. We have a fairly short period of time (the next 10-15 years while the assets in the trust fund are building) to implement the PSA. After 15 years the window closes.

4. Politically, it will probably not happen soon enough unless those over 65 push for it.


Folks, our politicians know the numbers. They expect to hear complaints from young workers who know the numbers, but they fear a backlash from retirees who don’t know the numbers.

And they know that retirees vote in greater percentages than do younger people. Plus, the problem won’t come to fruition for 30 years, which is 5 to 15 elections away.

So, in order to solve the problems, it is necessary for retirees to insist to their congress people that they reform Social Security for the benefi t of their children and grandchildren.

We welcome your comments and questions.

— Ron Muhlenkamp