The Truth About Social Security

By image - January 2, 2005


The Social Security program is a topic of intense discussion and debate today due to President Bush's initiative to partially privatize the program.

Understanding the debate on Social Security, of course, first requires understanding the Social Security program.

The Social Security program is not a retirement account. Social Security is a program for paying benefits to retired and disabled citizens from contributions made by those who are currently working. Basically, dollars paid in today are paid out tomorrow. The money comes in and is then immediately sent out to beneficiaries; it's a constant cash flow. Social Security taxes generally bring in more money than is needed to pay out however, and this excess is used to buy government bonds in Social Security trust funds.

The name "Social Security" has some significance to it, because in reality the purpose of this program is not so much to provide a retirement for individuals, but rather it is intended to safeguard society from the problems caused by massive numbers of people who are unable to care for themselves financially.

It's an insurance program for society.

Prior to the development of the Social Security program, increasing numbers of elderly citizens were becoming incapable of caring for themselves. As this happened elderly homelessness, suicides, and medical crises increased. These conditions put a strain on society because there was no program designed to handle these problems.

Social Security is designed to ensure a minimum standard of living for those who pay into it. It is insurance. It is an insurance program for the individual who receives the benefits and it's an insurance program for the rest of society to make sure that our society is not burdened by large numbers of people who are incapable of providing for themselves.

Though the Social Security system was signed into law in 1935, the concept had really been around in America for quite some time. President Theodore Roosevelt was the first major politician to advocate a social insurance program during his 1912 Progressive Party presidential campaign, and the concept was first promoted in America in 1910 by Columbia University economics professor Henry Seager, who wrote:

As changing economic conditions are rendering the dependence of old people on their descendants for support increasingly precarious, so, on the other hand, new obstacles are arising to providing for old age through voluntary saving. . . The proper method of safeguarding old age is clearly through some plan of insurance. . . for every wage earner to attempt to save enough by himself to provide for his old age is needlessly costly. The intelligent course is for him to combine with other wage earners to accumulate a common fund out of which old-age annuities may be paid to those who live long enough to need it.

Indeed, a major benefit of the current Social Security system is that it uses a single common fund, instead of many separate individual accounts.

Social Security is designed to be an extremely reliable, low risk, low cost program just to make sure that everyone who works and contributes to the program will have the bare minimum they need to survive when they retire or in the event that they suffer a disability.

The Details of Social Security

What we call "Social Security" actually encapsulates several different programs with different funding sources and different accounts. There are four components to Social Security:

  • OASI (Federal Old-Age and Survivorship Insurance)
  • DI (Federal Disability Insurance)
  • HI (Federal Hospital Insurance)
  • SMI (Supplementary Medical Insurance)

Each of these four different programs have different accounts, different funding sources, different lists of beneficiaries, different financial obligations, and different costs for administration. The HI and SMI programs comprise what is known as Medicare.

All of the Social Security taxes are what is known as payroll taxes, which is to say that these taxes come out of wages. These taxes are automatically deducted from workers' paychecks and employers have to "match" the amount paid by the employee. Self-employed persons pay one sum that is equal to wage-earner plus employer compensation. In order to make things simpler, its easier to just discuss the "self-employed" tax rates because, effectively, these are the rates that are ultimately paid by everyone.

For 2003, the Social Security tax rates were as follows:

Program Tax Rate
OASI 10.6%
DI 1.8%
HI 2.9%
SMI Funded from premiums and general revenue

The amount of income that is subject to Social Security taxes is limited. Wages are only subject to Social Security taxes below a certain level… sort of. Taxes for the OASI and DI programs are subject to the same limit. Historically, HI had its own different limit, but in 1993 the upper limit for HI taxation was removed. This means that now all wage income is subject to the HI tax.

The discussion of privatization centers on the OASI program, which is generally what we think of as the retirement program, however, the biggest financial problems are faced by the HI and SMI programs because they are heavily impacted by the rising costs of healthcare.

There are many issues to discuss with all of the programs, but since privatization deals mainly with OASI this is the program that I will focus on.

In 2003 the upper limit for OASI and DI, known together as OASDI, was $87,000. This means that any wage income over $87,000 was not subject to the OASI or DI taxes.

Below is a graph showing the historical OASDI upper taxation limits in 2001 dollars.


As you can see, this limit has been increasing since the inception of the program. The tax rates themselves have also increased since the inception of the program.

When Social Security was first implemented in 1937 the OASI tax rate was 2%, and remained such until 1950 when it was increased to 2.25%. Below is a graph of the OASI tax rates since 1950.


As the tax rates and limits have increased, so have the benefits. Since its inception, the Maximum benefit payable to Social Security beneficiaries has increased as shown below. The benefit increases are calculated using the Average Wage Index. The Average Wage Index is based on all national wages, even wages beyond the Maximum Taxable Earnings. As income disparity increases in America, the Average Wage Index grows at a rate faster than the rate of inflation or the wages that are taxed by OASDI. The spike in the 1970s is known as "the notch" and was a product of a miscalculation that resulted in higher payouts to beneficiaries than was supposed to occur. 


In the early days after the implementation of Social Security, the ratio of workers to beneficiaries was very high. This was because most of the people paying into the program had not yet begun drawing benefits. In 1940 the number of workers to beneficiaries was 159.4 to 1, meaning that for every person receiving benefits there were 159.4 people paying into the program. By 1945 the ratio had already dropped to 41.5 to 1, and by 1970 the ratio of workers to beneficiaries had stabilized near its current level at about 3 to 1. In fact, from the mid 1970s through the 1980s the ratio of workers to beneficiaries was slightly lower than it is today. (see source for details)


The ratio of workers to beneficiaries is currently projected to decrease to 2 to 1 by 2030 and to 1.8 to 1 by 2076

As was stated earlier, Social Security taxes typically bring in more revenue than needs to be paid out, and when this happens that excess is put into trust funds. There is a separate trust fund for each Social Security program and each of the trust funds operates somewhat differently. The HI and DI trust funds are the funds that are currently projected to run out of funds first, however, as has been stated, the privatization program does not address these funds and focuses on the OASI program, which has the least problems. The OASI trust fund is currently projected to run out of funds anywhere from 2042 to 2075, at which point it is projected that 73% of its current projected obligations could be met directly from taxation alone.

OASI, DI, and HI trust fund ratios as a percentage of annual expenditures

The OASI trust fund is the major fund, with the largest amount of capital. The money for the trust fund is invested into low risk government bonds. In 1999 the rate of return for the OASI and DI trust fund investments was about 6.9%.

The revenue for OASDI comes from three major sources, taxation of wages, interest on the trust fund, and taxation of benefits. In 2003 83% of OASDI revenue came from taxation of wages, 13% came from trust fund interest, and 2% came from taxation of benefits.

Due to the simple nature of the Social Security system, administrative costs are very low. For the OASI program, administrative costs are under 1%. In 2003 the administrative cost of the OASI program was 0.6%. This administrative cost is less than half of the administrative costs for a typical private mutual fund.

One peculiar aspect of the Social Security trust funds is that because the trust funds are invested in special issue government bonds, the trust funds effectively represent loans to the federal government, and the interest on the Social Security investments has to be paid back by money raised from general taxation. This, however, is how all government bonds work, including those that are privately held by investors. Because of the huge size of the Social Security trust funds, however, when these bonds need to be redeemed on large scale to help pay for benefits, it will mean that general taxes will likely have to be raised to pay back the bonds.

Now, having laid out some general information about Social Security, we can move on to some analysis and discussion of the issues at hand.

source: A Summary of the 2004 Annual Reports

Social Security and the Economy

In many ways, the Social Security system is serving as a leading economic indicator. The problems evident in Social Security financing are in reality a product of the problems that all working-class people in America face.

What is not discussed about Social Security in the mainstream media is the economic environment and how changes in our economic environment since 1940 impact the Social Security system.

One of the many claims made about Social Security by those who are in favor of president Bush's privatization plan is the claim that the ratio of workers to beneficiaries is much lower today than in the past, and that this is part of the problem as to why Social Security is not sustainable in its current form. The following is an example of such a claim found in popular media:

The current system is unsustainable. Over the next generation or two, Social Security faces a projected shortfall of at least $3.5 trillion. One big reason: the ratio of workers to retirees is shrinking. Just after World War II, there were 16 workers supporting each retiree. Today, that ratio is 3 to 1. By 2030, it will be less than 2 to 1.

NAVARETTE- Social Security reality- 'Buddy, can you spare me $3.5 trillion?'

This is a dubious claim. As was noted above, the ratio of workers to beneficiaries stabilized at around 3 to 1 in 1970. It is true that the ratio of workers to beneficiaries is projected to be about 2 to 1 by 2030, but in fact this is not a dramatic decrease. The Social Security system has been performing just fine for the past 30 years with 3 workers supporting every retiree and the system is still bringing in more revenue than it pays out today.

What is not discussed in the article are the following facts:

Productivity has increased dramatically throughout the 20th century, as shown below:


So, while the number of people supporting beneficiaries is decreasing slightly over time, we are producing more value for every hour spent working than in the past as well.

This gets into a more fundamental discussion of economics. Forget about the money, forget about the percentages and the ratios, and let's get back to basics here.

Social Security is a way for current workers to provide a basic level of care for non-workers, specifically people who, themselves, have worked to help create our society and who have, themselves, paid to take care of the elderly and disabled before them.

What this means is that this is a system of redistribution. It's taking part of the value that a present day worker creates and it's giving it to someone else to help them live. It's the same principle that has been present in civilization throughout history. People never depended on saving enough resources throughout their life to live on once they became elderly; the current working generation has traditionally set some of their resources aside to help care for the aged and disabled.

Let's use a very basic example. Let's use the example of a worker at a lumber mill based on the GDP per hour information above.

Let's say that in 1970 a man worked in a lumber mill, and this man was able to cut 70 logs a day. In 1970 the Social Security tax rate was 6.3%, so every day 5 logs were subtracted from his pile and those logs were used to pay towards the living expenses of a retired person.

Together, 3.7 workers contributing 5 logs a day from their 70 log pile was enough to pay for the benefits given to a retiree: 18.5 logs a day.

By 1998 a worker at that same lumber mill was producing 107 logs a day and, due to the 15.3% Social Security tax rate, 16 logs were taken from his pile each day to give to a beneficiary. At that time, it took 3.3 people contributing 16 logs a day to makeup what was given to one retiree, 52.8 logs.

This represents an increase in the benefits received by retirees in 1998 over those in 1970, although the ratio of workers to retirees had gone down.

While compensation to beneficiaries had increased, the workers still had more logs for themselves also, because productivity had increased. In 1970 the worker had 65 logs to himself at the end of the day, and in 1998 the worker had 91 logs to himself at the end of the day.

Assuming that productivity continues to increase, and there is no reason to believe that it won't, there should be no problem with the number of workers per beneficiary decreasing…. Except…

Except the fact that our economic system doesn't really work this way.

The above example would apply if we had a purely worker based economy, but we don't, and if GDP was a measure of beneficial output, but it isn't. There are other major elements to the equation: profits (capital gains) and income distribution, as well as the fact that GDP only measures economic activity, not necessarily economic benefit. An increasing amount of American economic activity is what many would call, non-productive, in other words, things like producing junk mail or paying lawyers to settle divorce cases, etc.

Since the 1970s the share of national income going to labor has decreased. In 1979 the share of national income going to labor was 73.2%. By 2000 that figure had dropped to 70.5%, although it has come back up slightly to 71% in 2004 since the Stock Market bubble burst of 2001.

Not only has the share of national income received by holders of capital increased since the 1970s, but wage disparity has increased as well, with dramatic increases in high-end income. For example, between 1982 and 2002 the average compensation of corporate executives increased by a factor of 10. At the same time, median income for working males has been flat since 1970 in inflation-adjusted terms.





The IRS reported that total taxable executive pay between 1980 and 1995 increased from $109 billion to $307.6 billion in inflation-adjusted dollars. This is a 182% increase in taxable pay that far exceeded growth in either corporate revenues or net income. Revenues increased by 129.5% and corporate profits increased by 127% over the same time period. During this same time period, wages for the 80% of the working population that make up non-supervisory workers actually declined when adjusted for inflation.

source: CEO Compensation- A Problem That Just Gets Worse

It is important here to bear in mind that there is an upper limit on the OASDI tax, above which the tax does not apply. In 2000 that upper limit was $76,200. The figure below shows approximately what portion of labor income was untaxed by OASDI taxes in 2000. The portion in red is above the OASDI taxation limit and thus exempt from OASDI taxation.


As you can see, a large portion of the nation's labor income is untaxed by OASDI taxes. The Social Security taxes fall most heavily on the middle class. In 2000 the top 1% paid an average of 1.9% of their income to Social Security taxes, while the middle class, the 40th to the 80th percentile, paid an average of 10% of their income to Social Security taxes.

For example, the average wage income for the top 1% in 2000 was $595,000, of which only the first $76,200 was taxed by OASDI.

Understanding how labor compensation works is also important here. The capitalist system is based on the use of labor markets to determine wages, and this means that compensation is not directly related to productivity. Changes in wage distribution do not necessarily reflect changes in productivity distribution.

Both of these factors, the decrease in the share of national income going to labor and the increase in income disparity with a dramatic rise in high-end wages at the top of the income spectrum, have resulted in a shrinking "tax base", not because of the number of retirees per worker, but because a larger portion of national income is not subject to the OASDI taxes, either because it is above the limit or it is realized in the form of capital gains, i.e. profits.

At the same time, the way that benefits are calculated for Social Security is based on the Average Wage Index. The Average Wage Index is computed using all wages that are taxed by federal income taxes. That means that as income disparity increases, the average wage increases at a higher rate than the median wage or the rate of inflation. This means that the middle-class is called on to pay increasing benefits based on wages that the middle-class is not receiving. The graph below shows the growing divergence between the average and median wage.


We can now reapply these facts to our logging example.

Not only are workers as a whole today keeping a smaller share of the logs that they produce at work than they were in 1970, but a larger percentage of the logs produced are being distributed to a minority at the top end of the pay scale, above the limits of the OASDI tax limits.

In 1970 the worker at the lumber mill cut 70 logs a day. These logs all went into one big stack in the center of the lumberyard and no one kept a direct count of the logs that each worker produced.

When the worker took his job he agreed to do the work for 30 logs a day because other people agreed to do the job for 30 logs a day.

Since the overall return to labor at that time was 73%, or 49 logs out of 70, this meant that he was getting about 19 logs less than the total return to labor.

The other 19 logs went towards paying managers and other workers at the company.

Of the logs he was paid, about 12% were taxed away by the federal government for general taxes and 6.9% were taken for Social Security. This left him with 24.3 logs, and he contributed 2.1 logs to care for a retiree.

In 2000 a worker in that same lumber mill produced 107 logs a day, all of which went to the central company stack.

When he took his job he agreed to do it for 30 logs a day as well, because remember that in inflation- adjusted dollars median pay for male workers in the US has been stagnant since 1970, meaning that average male workers today are being compensated the same as they were in 1970.

Since the overall return to labor in 2000 was 70%, or 75 logs out of 107, this means that he was getting about 45 logs less than the total return to labor, again, these 45 logs represent a part of the return to labor that went to compensate mangers, etc.

Of the 30 logs he was paid, the federal government took 10% for general taxes and 15% for Social Security. This left him with 22.7 logs and he contributed 4.5 logs to care for a retiree.

If we take the 2.1 logs from 1970 and multiply it by 3.7 workers then we have 7.7 logs per retiree.

If we take 4.5 logs from 2000 and multiply it by 3.3 workers we have 14.85 logs per retiree.

The increase in compensation for Social Security beneficiaries between 1970 and today has come not from increases in productivity, but from increases in taxation, because OASDI taxes are income capped wage taxes and the increases in productivity in America since the 1970s have not been reflected in the incomes of ordinary working class Americans, who are the base of the Social Security system.

This is the real root source of the problem, it is the fact that a smaller and smaller portion of national income, i.e. what is produced, is actually subject to Social Security taxes because average workers are receiving a smaller and smaller share of the national income pie. This is why Social Security taxes have had to increase so much to pay benefits, even while productivity has dramatically improved over the years.

So, how does this relate to Bush's proposal to partially privatize Social Security?

Privatization of Social Security

If an increasing share of national income is going to capital then wouldn't privatization of Social Security take advantage of this? Yes and no. I am not fundamentally opposed to the development of a national investment system, in fact I have been advocating such a system for three years, as discussed in this article:  Understanding Capitalism Part I- Capital and Society

Despite that, there are still problems with president Bush's approach and with the arguments that are advanced in favor of his plan. One of the often heard statements in support of partial privatization of Social Security is that the average return in "the Stock Market" over time has been 10%, and that this 10% would be better for people who pay into the program than the "returns" currently offered by Social Security.

There are many issues with this line of reasoning. First of all, there is no real rate of return on Social Security, although it is often claimed that the rate of return from Social Security is around 2%, this number is actually completely meaningless. There is no real measurable rate of return for Social Security precisely because no one "owns" any shares in it. Paying into Social Security entitles you to a certain level of benefits, but there is no way to take all of the benefits in a lump sum or anything of that nature, so there is no real way to say that there is a rate of return. What you get out of Social Security depends on how long you live. If you die before you ever collect any benefits then you get nothing, and your dependants may or may not be able to get a part of your benefits. An average wage-earner retiring in 2003 will get back what they contributed to Social Security in 7.5 years. All payments from Social Security after that are in excess of contributions made.

If you die at age 70, after having only gotten a few years of benefits, then your "return" is less than if you live to 100 and collect Social Security benefits for 35 years.

That's really part of the point, again, of "Social" Security. We all pay in so that whoever lives a long time, or becomes debilitated due to injury, will have some means of support so that they don't become a burden on society. If you live to be 100 then you still get your checks, you don't have worry about running out of money. If you get diagnosed with terminal cancer and have 6 months to live then, unlike a private account, you can't get a big lump sum and blow it on gambling in Vegas. Again, it's a social insurance program, not a retirement account.

So, any discussion of "rate of return" from "your" Social Security investment is meaningless. As was stated earlier, the Social Security trust funds earn around 5% to 7% interest.

Secondly, as for the claim that the historic rate of return for the Stock Market is 10%, this rate of return has always fluctuated dramatically over time, as we have seen just recently with the Stock Market bubble of 2000 and 2001. When people are retired, these fluctuations could have major impacts on meeting needs.

That, however, is not even the major issue. The real fact is that, as investors will tell you, past performance is not indicative of future results. It doesn't matter if you are discussing specific stocks or the entire market, past performance is not an indicator of future results.

The 20th century saw an unprecedented level of growth. The population of the United States increased by 360% between 1900 and 2000. Over the next 50 years the US population is projected to grow by 47% to a whopping 400 million, but this rate of growth for the next 50 years will still be many times smaller than what occurred over the past century.

In addition to that, the United States of America was one of the leaders of the Industrial Revolution and remained as the only major industrial power in the world after World War II due to the destruction that affected almost every country in the world except America. Today we face much stronger international economic competition than ever before.

Additionally, the United States rose to industrial power during a time of unprecedented cheap energy costs, and as we are aware today, cheap energy will likely be a thing of the past for a least a few generations to come as oil supplies continue to dwindle and demand continues to rise. In the event of a major technological breakthrough a new source of cheap energy may replace oil, but there is no proven replacement yet.

On top of all that, the American dollar is continuing to decline in value and will likely never regain the dominance that it had during the last half of the 20th century because the adoption of the greenback as the international reserve currency was a product of post World War II treaties and agreements related to the Cold War. These conditions simply don't exist anymore.

These factors, and others, make predicting a 10% return on the Stock Market over the next 50 years a very dubious claim. The rate of return in the American Stock Market for the 21st century will likely be lower than it was during the 20th century for completely understandable fundamental reasons.

On top of these facts, any partial privatization of Social Security would inherently make the system more complex and increase the cost of operation.

While President Bush's endorsement of an "ownership society" based on stock ownership is certainly a good idea, the devil is in the details.

A major problem with partial privatization of the OASI program is that the OASI benefits are very stable and secure, and this security is a large part of their importance, not only to individuals, but as a stabilizing effect on the entire economy. OASI benefits provide for the basic needs of many retirees, without which many of them would be unable to meet their daily living expenses. This means that fluctuations in benefits due to Stock Market activity could have a major impact on the economy as a whole. 

The graph below shows 20th century business cycles in America. The implementation of Social Security in the 1940s was a major factor in the stabilization of the American economy in the 20th century.


OASI benefits should be maintained as they are because introducing increased risk for this type of system is very dangerous. This could certainly increase economic volatility and lead to extended recessions and depressions in the future because when a drop in the Stock Market does occur that would mean that many retirees, who would be dependant on the values of their privatized Social Security accounts, would immediately see diminishing benefits themselves, which could lead to major decreases in spending among the large retirement population, further deepening a depression, not to mention the problems caused by large numbers of retirees who may be unable to meet existing financial obligations, such as rent and utility bills.

Our Social Security system is a form of national economic diversification. Tying too much of our economy into the Stock Market decreases total national diversification and increases total national economic risk. 

In terms of a national investment system, I am in favor of such a program, however I think that it should be separate from Social Security. We should create a new national investment system, and not privatize the OASI portion of Social Security because it is filling the role that it is intended to fill, Social Security. Keyword security. What should definitely be done is that benefit increases should be based on either the rate of inflation or on median wages, instead of on the Average Wage Index. What could also be done, is that the cap on the OASDI taxes could be removed without increasing the maximum benefit payments to compensate. If this were done the overall OASDI tax rates could also be reduced. At the very least the DI portion of the OASDI tax should have its cap removed. This would make the DI program solvent beyond the term of current projections at least, if not indefinitely, and the DI program is a program that supports disabled citizens. Surely caring for this group should not be the sole responsibility of the middle class.

The problem with the idea of partial privatization of Social Security is that it does not address the root causes of the problems. The OASI fund is the most economically sound fund. The funds with the biggest problems are the HI and SMI funds. These funds, the Medicare funds, face immediate problems and privatization in no way addresses these problems. Medicare reform and healthcare reform are much more important than retirement benefit reform, and it was the Bush administration that failed to support the best opportunity to control costs within Medicare that has come along in a long time, and that was when Bush signed the Medicare reform bill in 2004 that forbids Medicare from negotiating lower prescription drug prices.

Furthermore, the projected distant future shortfall for the OASI fund is a product of a flaw in how benefit increases are calculated that is exacerbated by an underlying problem of the country, which is that income disparity is increasing and the middle class is being increasingly squeezed economically. This, of course, impacts the ability of the middle class to pay for the support of retirees.

The fact of the matter is that in a nation of increasing productivity there should be no need to extend the working time of citizens, or to increase tax rates, or to cut benefits in order to pay for the needs of retired workers and the disabled, even with a decreasing ratio of workers to beneficiaries. 

It is fallacious to say that as a society "we" can't afford to maintain retirement benefits. The fact is that "we" can, but the middle class cannot. "We" can afford to pay basketball players millions of dollars a year to play games. "We" could afford to spend $11.2 billion on cigarette advertising in 2001. "We" could afford to compensate the CEOs of the 365 largest companies an average of $8.1 million each in 2003.

It's not as though "we" can't afford to care for our elderly and disabled, its a matter of our priorities as a society however. When we say that "we" can't afford Social Security, what we are saying is that professional athletes receiving millions of dollars a year are more important to us than retirees.

If that's where people's values are, then that's where they are, but let's not pretend as though this is the result of an acute funding crisis; the problems with Social Security are a reflection of the problems within our entire society. The real solution for fixing Social Security is to fix the underlying problems.

See also:

Getting a grip on Social Security: The flaw in the system

Social Security Reform 2005

For more on the Bush administration's framing of the Social Security debate see:

Social Security- Bush's Lies vs. Reality

Move over Gingrich, Frank Luntz is the new Goebbels

This page has had Hit Counter views since 1/2/2005

Copyright 2003 - 2010  Website Launched: 5/22/2003  Last Updated: 9/28/2010  Contact: