What’s Social Security worth to you?

Here’s another way of looking at Social Security, courtesy of a new report from the National Academy of Social Insurance: If you bought an insurance company annuity to create an income stream equal to your Social Security benefits, it would run you about $225,000.

That assumes a 65-year-old retiree collecting the average monthly Social Security benefit for 2007 of $1,045. The annuity’s payments would rise with inflation (similar to Social Security) and continue paying a surviving spouse after the retiree died (ditto).  We haven’t checked the report’s math, but its logic seems reasonable.

Of course, the value of your benefits will depend on many factors, one of which is whether you start to collect at age 62, at your “full” retirement age, or as late as age 70. We explored some of the pros and cons of that decision in this recent article in Consumer Reports.

If you’d like to weigh in on the value of Social Security, or any other retirement-related topic, visit our new retirement forum.

Comments

[Bill Woessner]

OK, so Social Security is worth $225K. The next logical question to ask is, how much do you have to pay for it? Let's run the numbers, shall we?

The median personal income for people over the age of 15 is $28,567. The Social Security tax rate is 12.4%. From age 15 to 65 is 50 years. $28,567 * 12.4% * 50 = $177,155.4. Now that includes absolutely no real interest - just keeping up with inflation. If we apply a little Newton's method, we can figure out the real rate of return. Turns out it's 0.92%. And that's assuming no benefit cut or tax increase. Pathetic.

Forget about investing in stocks. People would do better to take their Social Security taxes and put them in a savings account. Even with the recent Fed funds rate cut, you can still get 4.4% APY. Subtract out inflation and you get 1.16% real return. I'll take that over 0.92% any day. In the scenario above, it translates to an additional $15K at retirement.

Oh, and it has the added benefit of not plunging the country in to massive debt. Right now, Social Security represents an unfunded liability of roughly $10 trillion. If you think Bush's deficit spending is bad, it's only half of the deficit incured by Social Security. Medicaid, of course, is far worse (~$60 trillion), but that's another topic.

[Paul Van de Water]

Bill’s calculations are way off the mark. According to the Social Security actuaries, the internal real rate of return for a worker with medium earnings reaching age 65 this year ranges from 1.86% to 4.38%, depending on marital status and spousal earnings. More important, though, it doesn’t make sense to think of Social Security as a savings account. You certainly wouldn’t think about the rate of return on your automobile or fire insurance. Social Security provides life and disability insurance when you’re still working. It also provides inflation-protected income in retirement that continues for as long as you live—protection that you just can’t buy in the private market.

[Bill Woessner]

"According to the Social Security actuaries, the internal real rate of return for a worker with medium earnings reaching age 65 this year ranges from 1.86% to 4.38%"

Those figures might have been accurate... BEFORE the Social Security tax rate was jacked up to 12.4% (of course, that's also before the full retirement age was jacked up to 67, but that's another matter). Those who paid Social Security taxes before 1990 did not pay the full 12.4% that we pay today. In fact, if you started with Social Security all the way back in 1937, you had the luxury of only paying 2%.

Regardless, I can mathematically prove why 0.92% is the correct value. The median salary for people age 15 and older is $28,567. With a tax rate of 12.4%, that translates to $3,542.31 in taxes every year for 50 years. All we need to do is plug that in to the annuity formula:

F = D * (e^(rt) - 1) / r

where F is the future value, D is the annual deposit, r is the interest rate and t is the length of time. This is assuming continuously compounded interest. You could substitute discretely compounded interest but it won't dramatically change the outcome. For our numbers we get:

F = 3542.31 * (e^(.0092*50) - 1) / .0092
= 224888.16

Due to rounding, I'm off by $111.84. Close enough for government work.

If you think it's unfair to start at age 15 (which it's not; 15 year olds pay Social Security taxes at the same rate as everyone else), we can up the ante to age 25. The median income for those 25 and older is $32,140. The rate of return here is 1.63% as proved by:

F = 3985.36 * (e^(.0163 * 40) - 1) / .0163
= 224787.93

Again, rounding causes the difference of $212.07.

"More important, though, it doesn’t make sense to think of Social Security as a savings account."

Then why do the Social Security actuaries do it?

[Ken Buffin]

Bill Woessner invites us to "let's run the numbers, shall we?" and then proceeds to act as the WE and produces THE numbers based on assumptions and scenarios selected by himself. His assumptions, scenarios and choice of numbers are highly selective and do not capture the broad range of potential results that could emerge from utilizing more realistic alternative assumptions, scenarios and numbers. He begins by assuming an entrant to the workforce at age 15 who contributes to Social Security for a full 50 years until age 65; this is a far from realistic profile of the typical American worker who is likely to be still in high-school at age 15 and quite probably attending college well into his or her early-twenties. He then adopts a single number of $28,567 as the median personal income for individuals in his calculation and assumes that income is payable as a level amount over his chosen period of employment of 50 years; again this does not represent a realistic pattern of career earnings for a typical American worker. He then proceeds to double this hypothetical worker's contributions that are deducted as FICA payroll taxes at the rate of 6.2% on covered earnings and takes full credit for the employer's taxes making up a contribution rate of 12.4% to arrive (arithmetically incorrectly) at an amount of $177,155.40 (the correct arithmetical result is $177,115.40). He then "applies a little Newton 's method" to produce an annual rate of return of 0.92% to equate his chosen stream of 50-year contributions to the amount of $225,000 as the value of Social Security benefits according to the National Academy of Social Insurance. (It turns out incidentally that the correct result for this computation is actually 0.94%).

A previous response to Bill Woessner by Paul Van de Water has pointed out that the actuaries at the Social Security Administration have produced a much more realistic range of values for the internal rate of return for workers reaching age 65 based on scenarios and assumptions that more closely track the reality of the work experience of the average American worker and this range is from 1.86% to 4.38%. Now, it's a free country and we are all free to express our opinions and to make up our own facts based on our own chosen assumptions and scenarios, but given the choice of having Bill Woessner or the Social Security actuaries run THE numbers, I believe that most rational Americans would prefer to choose the latter as someone WE can trust to "run the numbers". Accordingly, I would like to go on record as politely declining the request of Bill Woessner to "let's run the numbers, shall we?" No thanks, Bill, I don't trust you to choose realistic assumptions and realistic scenarios or in fact to do the arithmetic correctly.

Bill Woessner also makes a reference to another topic with an unqualified statement that "right now, Social Security represents an unfunded liability of roughly $10 trillion." It is not specified as to whether this reference is to an infinite horizon projection or a closed group projection. I know the opinion of one expert who believes that the first type of projection is ridiculous, bizarre and designed to mislead and that the second type of projection is simply not meaningful in terms of the established financing arrangements for Social Security.
While the official Trustees' Report for Social Security explains that the projected income and assets for Social Security are adequate to meet projected outgo until 2041, potential deficits may arise subsequently in the absence of changes to restore the solvency of the system beyond 2041. A key question is how far out beyond 2041 should projections be made to quantify the potential range of future shortfalls in the adequacy of Social Security revenues to meet scheduled benefit payments. It has been conventional practice to consider 75 years as the limit for which credible projections can be made and to quantify the extent of the uncertainty associated with them by projecting alternative scenarios on a high-cost and low-cost basis (as well as in recent years using stochastic model projections) to capture the extent of divergence in the range of plausible outcomes. No one is denying the existence of potential future requirements to strengthen the solvency and sustainability of the Social Security system; but whenever it's time to run the long-term Social Security projection numbers, please let's explain them carefully and rationally in non-sensational terms so as not to mislead the public by providing deliberately distorted and politically-motivated misinformation thus creating public hysteria about the viability of Social Security. The U.S. Social Security system is providing an extremely valuable socioeconomic benefit and service to millions of Americans. It is an institution that we can justifiably be proud of. It is above all a demonstration of national community spirit, risk-sharing and concern for our fellow Americans, particularly the elderly, the disabled and the many children, widows and others among us who deserve our cooperation and support to alleviate poverty and provide a degree of dignity and income support in old-age and in other difficult circumstances.

[Candy]

Well, Paul disability insurance is much cheaper than SSI. I don't know of another "insurance" policy that people are forced to pay for their entire working lives. When you buy an auto policy you are allowed to shop around and you know what it will pay out. How is SSI inflation protected? The government does not use real inflation to calculate increases in SSI. Plus there is no guarantee of benefits, my old SS card actually stated "There are no guarantees of benefits". It is a tax and subsidy, vote buying scam. It has been from the get go.

Seeing how the stock market has averaged a return of about 10% during every 10 year period. Even during the depression years the market saw gains. Subtract real inflation from that and you are much better off.

[Knows mathematics]

Two points about the math above in Bill Woessner's post.

1) Wages are not the value of the last week worked. Over my years paying into SS my wage has increased from $600/year to (over the SS cap)/year. I have paid less than $100K into the system @6.2% of salary even though I have exceeded the cap since 1992. The actuarial value of my SS at full retirement at current annuity rates is close to $500,000 and I am in the highest bend point which means that I get less as a % of salary than anyone else. So most other people get a greater bang for their buck than I do from an annuity argument.
2) If the employer contribution is considered the employee contribution, the wage must be adjusted such that the 12.4% is the same level of contribution as before the math mistake. You can't just use the employer contribution to inflate your calculations on the same salary. The employer contribution does not come out of the salary listed as one's gross income. If you argue that, without SS, the salary would increase, you need to increase the salary to support the argument. This is the argument made when people try to call the employee contribution 12.4%. Otherwise you must use 6.2% because that is all you paid from your taxable income. And, as I said, that 6.2% must be on each years real taxable salary and not on a current median income.

[Bill Woessner]

Ken, it's a pretty clear sign that your argument is lacking in substance when you have to focus on a $40 error on a $177K number(0.02% relative error) which is pretty clearly the result of a transcription error. But I digress.

What I have done is apply a 1st order approximation. Personally, I think it's a very reasonable approximation to take the median personal income and apply it over the working career. People earn less when they first start out and more when they're close to retirement. That's why it's called the median. I use the median because the mean is heavily influenced by extremely high wage earners. If you don't like starting at age 15, that's fine. I reposted with the calculation starting at age 25.

You also complain about including the employer half of FICA. But the employer portion is paid on my behalf. It is paid because I work. So I think it's eminently fair to include the employer portion. And really, the employer/employee division of FICA taxes is just a shell game. Employers consider their half of your payroll taxes as part of your compensation. If we wanted to be transparent with payroll taxes, we would "shift" them 100% to the employee. But 15.3% looks a lot worse than 7.65% and people might get upset.

There is the issue raised by "Knows Mathematics" as to the actual percentage. It's a good point. The Social Security tax is 12.4% of your salary but it is NOT 12.4% of your compensation. If your employer kicks in 6.2% of that, it really raises your compensation to 106.2% of your salary. So to be fair, you could say that the tax is 11.68% of your total compensation. But in the computation I did above, I based everything off of median salary, so I stand by using the 12.4%.

Also, Ken, your 0.94% rate of return is based on annual compounding whereas my 0.92% rate of return is based on continuous compounding. I apologize for not stating that assumption in my initial post. However, I did state it in my follow-up post. Furthmore, you validated my comment that changing the compounding period does not dramatically alter the result.

At this point, I won't say anything else about the math. You're nitpicking at it, which is a pretty clear sign that you have no real argument against it. Besides which, there's a much more important matter at stake. You seem willing to just accept whatever the government tells you. Social Security has a vested interest in making itself look good so as to perpetuate itself. My interest is in doing what's best for me and my family. I can understand you not trusting me; you don't know me from Adam. But putting such blind faith in government makes is downright dangerous.

Finally, I will yield the point on a 1.86% to 4.38% rate of return for people retiring this year. But if you accept that, then it immediately follows that the rate of return is significantly lower for people entering the workforce this year. It should be obvious since the tax rate is higher and benefits have been reduced by pushing back the retirement age. And with tax hikes and/or benefit cuts looming, the situation is just going to get worse. How bad does it have to get before we say enough is enough?

[Ken Buffin]

Bill, there is a much larger point at stake here. You take the view that your interest is “doing what’s best for me and my family” and that seems to lead you to look at the Social Security system as a personal savings and investment account with its value determined by a hypothetical rate of return on both your and your employer’s payroll taxes. Social Security is not a system of personal savings and investment; it a system of Social Insurance. Many people look differently at Social Security and understand its value as an insurance program with a sense of community-spirit, social solidarity, and concern for their fellow Americans. They pay Social Security taxes and take pride in seeing the system pay insurance benefits to widows and children and the disabled and poor, with the amount of these benefits exceeding the value of the contributions paid by or for them. You also seem to have missed the main point in my previous comments. What is wrong with your approach is the inappropriate choice of assumptions and scenarios. I very clearly stated my opinion that the actuaries at the Social Security Administration have produced a much more realistic range of values for the internal rate of return based on scenarios and assumptions that more closely track the reality of the work experience of the average American worker. I also stated that it’s a free country and that we are all (including you) free to make up our own facts and assumptions. However, you are quite mistaken in your view that I seem willing to just accept whatever the government tells me. I researched and analyzed the social security actuaries’ scenarios and assumptions very carefully before deciding that they are more appropriate and credible than yours.

[Bill Woessner]

I hear what you're saying about Social Security being an insurance program. But if I think about it as such, it makes me even more angry because Warren Buffett collects Social Security. Why is the 2nd wealthiest man in the United States collecting benefits from a system designed to protect widows, children, the disabled and the poor? I want to help the poor, Ken, I really do. But I don't want to line Warren Buffett's pocket.

Furthermore, if Social Security is insurance, then it is absurdly expensive insurance. As a responsible parent, I bought life insurance and disability insurance after my son was born. The life insurance is worth 20 times my salary and it costs about 2% of my salary. The disability insurance would basically replace my entire income and I pay only 2.75% of my salary for that. Both of these policies are worth an order of magnitude more than Social Security, yet cost far less.

If you want to help the poor, then Social Security is not the answer. With Social Security, you get what you pay for (and you pay dearly for it). The answer to helping the poor is welfare. Consider this. For less than we currently pay for Social Security, we could completely eliminate poverty in the United States. According to 2005 census data, there would have been 55,369,000 Americans living in poverty if not for government assistance. Give each of them enough money to raise them above the poverty line ($9,570 for 2005). Maximum cost: $530 billion. Total cost of Social Security in 2005: $564 billion.

Note that $530B is just the maximum cost. The real cost is likely to be half that much. For starters, not everyone below the poverty line has 0 income. If you assume the distribution of income among those below poverty is uniform (it's not, but it's closer than assuming everyone has 0 income), then the estimate goes down by a factor of 2. Also, the upper bound doesn't factor in people living together. The poverty guideline for 2 adults living together was $13,078, not $19,140.

It seems to me that Social Security has no clear goal. If the goal is to provide for retirement, we need to provide a competitive return on investment. If the goal is to provide life insurance and disability insurance, we need to have premiums that are competitive with the open market. If the goal is to help the middle class, why do we make them pay for it? If the goal is to help the poor, we could do much more with much less. If the goal is to garner votes... well, maybe Social Security works pretty well.

[Mortgage Claims]

It seems to me that Social Security has no clear goal. If the goal is to provide for retirement, we need to provide a competitive return on investment. If the goal is to provide life insurance and disability insurance, we need to have premiums that are competitive with the open market. If the goal is to help the middle class, why do we make them pay for it? If the goal is to help the poor, we could do much more with much less. If the goal is to garner votes... well, maybe Social Security works pretty well.