Chapter 13 Social Security

Social Security: The Individual’s View

Household incomes tend to vary quite a lot, but households like their consumption to be approximately constant over time. Households therefore use their savings to smooth out the variations in their income.We also discuss consumption smoothing in Chapter 12 "Income Taxes". For the purposes of understanding Social Security—both its problems and its reforms—we need to examine this idea of consumption smoothing more rigorously. Because Social Security is a program to provide for consumption in retirement, we must carefully lay out the decisions of a household over the entire lifetime of its members. By so doing, we can determine the likely effect of a promise of a transfer in the future on behavior today. Bear in mind that these transfers may be far in the future: for a 25-year-old worker, we are thinking about money that won’t be received for another 40 years or so.

As is often the case in economics, we start by looking at the simplest setup we can imagine. We do this not because we think we can answer every question with a simple model, but because we must make sure we understand the fundamentals of Social Security before we worry about the complexities. So, to keep things simple, we examine the life of a single individual—that is, think of the household as containing just one person. Hence we do not have to worry about multiple wage earners (who might be of different ages), and we do not have to worry about how to incorporate children (who grow up and leave the household) into our story. In this chapter, we use the terms individual and household interchangeably. We call our individual Carlo.

Carlo thinks about his income and consumption over his entire lifetime. Because he has the possibility of saving and borrowing, his income and consumption need not be equal in any given year. Carlo faces a lifetime budget constraintThe discounted present value of lifetime consumption must equal the discounted present value of lifetime income., however; in the end, his lifetime spending is limited by his lifetime income. The life-cycle model of consumptionA model studying how an individual chooses a lifetime pattern of saving and consumption given a lifetime budget constraint. examines Carlo’s decisions about how much to consume each year, given this budget constraint.

We begin with a simple numerical example. Suppose Carlo is 20 years old and very well informed about his future. He knows that he is going to work for 45 years—that is, up to age 65. He knows that, every year, he will receive income of $40,000, excluding Social Security contributions. He has to pay Social Security contributions on this income at a rate of 15 percent. Thus he knows he will pay $6,000 each year to Social Security, and his after-tax income is $40,000 − $6,000 = $34,000. After he retires at age 65, he knows he will receive a Social Security payment of $18,000 each year until he dies, 15 years later, on his 80th birthday (of heart failure, brought on by the exertion of blowing out all those candles).

To decide on his lifetime consumption and saving patterns, Carlo needs to know what his lifetime resources are. We know that, in general, a dollar today is not worth the same amount as a dollar next year—or 60 years from now—because of interest rates and inflation. We sidestep that problem for the moment by imagining that the real interest rateThe rate of return specified in terms of goods, not money. is zero. In this case, it is legitimate to add together dollars from different time periods.

So Carlo earns $34,000 per year for each of his 45 working years and obtains $18,000 per year for his 15 retired years. His total lifetime resources are as follows:

lifetime income = income during working years + income during retirement years = ($34,000 × 45) + ($18,000 × 15) = $1,530,000 + $270,000 = $1,800,000.

Over his life, therefore, he has $1.8 million to spend. Figure 13.1 "Lifetime Income" shows his lifetime income. His total lifetime resources are obtained by adding together the two rectangles labeled “Lifetime Income from Working” and “Lifetime Social Security Income.” The height of each rectangle gives his income, and the width of each rectangle gives the number of years for which he earns that income.

Carlo’s lifetime budget constraint says that his lifetime consumption must equal his lifetime income,

lifetime consumption = lifetime income.

If Carlo wants to keep his consumption perfectly smooth, he will consume exactly the same amount in each of his 60 remaining years of life. In this case, his consumption each year is given by the following equation:

$$\text{annual consumption}=\frac{\text{lifetime consumption}}{60}=\frac{\$\mathrm{1,800,000}}{60}=\$\mathrm{30,000.}$$

Figure 13.2 "Lifetime Consumption" shows Carlo’s consumption. The area of the rectangle is a measure of Carlo’s lifetime consumption since—as before—the height of the rectangle is his consumption per year, and the width is the number of years.

In each of his working years, Carlo earns $34,000 but consumes only $30,000. Thus he saves $4,000 every year. When he is 21 years old, he therefore has $4,000 in the bank. When he is 22 years old, he has $8,000 in the bank. By the time he retires at age 65, he has saved $180,000 (= 45 × $4,000).

During his retirement years, Carlo starts to draw on his savings. Social Security pays him $18,000, so he needs to take an additional $12,000 from his savings to have $30,000 in consumption. At age 66, therefore, he has savings of $180,000 − $12,000 = $168,000. For each of his retirement years, his savings are reduced by a further $12,000. After his 15 years of retirement, he has reduced his savings by $12,000 × 15 = $180,000 and dies at the age of 80 with exactly zero in the bank.

Figure 13.3 "Lifetime Consumption and Saving" combines Figure 13.1 "Lifetime Income" and Figure 13.2 "Lifetime Consumption" and shows Carlo’s income and consumption. The difference between income and consumption in Carlo’s working years is his saving. Notice the rectangles labeled “saving” and “dissaving.” One way of understanding his lifetime budget constraint is that these two rectangles must be equal in area.

Figure 13.4 "Lifetime Wealth Accumulation" shows his wealth over his lifetime. It increases from zero to $180,000 and then decreases again to zero. The fact that he ends his life with exactly zero wealth is just another way of saying that he exactly satisfies his lifetime budget constraint.

Social Security: The Government’s View

Now let us shift perspective and examine the Social Security system from the perspective of the government. The original intention was that Social Security would be (approximately) pay-as-you-go. Under a strict pay-as-you-go system, the inflows to the government in the form of tax revenues are exactly balanced by outflows to retired people. In any given year, in other words, the government takes money from those that are working and transfers all that money—not a cent more, not a cent less—to those who are retired.

Under this system the government does not maintain any kind of “savings accounts” for individuals: it taxes you when you work and transfers the revenues to retirees at the same time. The government promises to make payments to you after you retire, with these payments being financed by those who will then be working.

Let us pause for a moment here. We have to determine how to capture transfers across different generations in the economy in as simple a setup as possible. The easiest way to do this is to suppose that everyone in the economy is just like Carlo. That is, every working person in the economy earns $40,000 and pays $6,000 into the Social Security system. Every retired person receives a Social Security payment of $18,000 per year.

Let us further suppose that there is the same number of people of every age in the economy. In each year the same number of people is born, and—like Carlo—they all live to exactly the age of 80. Like Carlo, everyone works for 45 years (from age 20 to 65) and is retired for 15 years (from age 65 to 80). If we wanted to calibrate this roughly to the US economy (that is, make the numbers in the example a bit more realistic), we might suppose that there are 4 million people born every year. Since everyone lives to the same age, this means that there are 4 million 20-year-olds, 4 million 21-year-olds, and so forth, up to 4 million 79-year-olds. (This implies a total population of 320 million, which is close to the size of the actual US population.)

Having made these simplifications, it is a short step to realize that we might as well just suppose that there is only one person of each age. The basic structure of the economy will be the same, but the math will be much easier. (If you can prefer, though, you can multiply both sides of every equation that follows by 4 million.)

Given this demography, what do the government finances look like? Every year, the government collects $6,000 each in Social Security revenues from 45 working people, so that the total revenues are given by the following equation:

Social Security revenues = 45 × $6,000 = $270,000.

Meanwhile, the government pays out $18,000 each year to 15 people:

Social Security payments = 15 × $18,000 = $270,000.

You can see that we have chosen the numbers for our example such that the Social Security system is in balance: revenues equal receipts. A system like this one would indeed be pay-as-you-go.

The Effect of a Change in Social Security Benefits

Now, what would happen in this example if the government decided it wanted to increase Social Security payments by $3,000 per year? The total increase in payments would equal $45,000 since all 15 retired individuals would receive the extra $3,000. If the government is required to keep the Social Security system in balance, then it would also be obliged to increase Social Security contributions by $1,000 per worker (since there are 45 workers). How would Carlo (and everybody else like him) feel about this change?

Remember that Carlo had income before Social Security of $40,000 per year and had to pay $6,000 per year in Social Security contributions. Now he will have to pay $7,000 in contributions, so his income after tax is equal to $33,000. In his retirement years, however, Carlo will now get $21,000 a year instead of $18,000. His lifetime resources are therefore as follows:

lifetime income = ($33,000 × 45) + ($21,000 × 15) = $1,485,000 + $315,000 = $1,800,000.

Carlo’s lifetime resources are exactly the same as they were before. Of course, this means that Carlo would choose exactly the same amount of consumption as before: $30,000 a year. However, his saving behavior would be different. He would now only save $3,000 a year. At the time of retirement, he would have saved a total of $135,000. Over the remaining 15 years of his life, Carlo would draw on his savings at the rate of $9,000 per year, which—combined with his Social Security payment of $21,000—would ensure that he had $30,000 to spend in his retirement years. His saving and dissaving are illustrated in Figure 13.5 "Lifetime Consumption and Saving".

We have discovered a rather remarkable conclusion: the change in the government’s Social Security scheme has no effect on Carlo’s lifetime resources or lifetime consumption. From Carlo’s point of view, the change means that the government is saving more on Carlo’s behalf, and therefore he does not need to save so much for himself. Carlo’s saving declines by exactly the same amount as the increase in Social Security taxes ($1,000) per year; likewise, his dissaving declines by exactly the same amount as the increase in Social Security payments.

Another example is even more striking. Suppose there were no Social Security system at all. Then Carlo would receive $40,000 a year for 45 years but nothing at all in his retirement years. His lifetime resources would equal

lifetime income = $40,000 × 45 = $1,800,000,

which is again the same as before. To enjoy lifetime consumption of $30,000 a year, Carlo would save $10,000 in every working year and dissave $30,000 in every retirement year.

These numerical examples suggest an extraordinary conclusion: Social Security seems to be completely irrelevant for Carlo. No matter what the scheme looks like, Carlo has the same lifetime resources and same lifetime consumption. This is an amazing and perhaps even shocking finding. We have used some economics to analyze one of the most important government programs, one that is a source of constant scrutiny and debate (and, not incidentally, requires substantial resources for its administration). Not only have we found no reason to expect a crisis in Social Security, we have found that it is irrelevant.

Should we now pack up and go home, saying that “economists have analyzed Social Security and it is actually a nonissue”? We hope it is obvious that the answer is no. After all, all we have done so far is present a numerical example. The example suggests that Social Security might be irrelevant under certain circumstances, but it certainly does not prove that it is irrelevant in general.

This is how economists very often work. A simple numerical example can help us understand the operation of a complicated system like Social Security and can lead to some suggestive conclusions. Our next task is to determine whether the conclusion from our example holds more generally. We will first see that the result does not depend only on the particular numbers that we chose. We will then try to understand exactly where the conclusion comes from.

Household Budget Constraints

We first consider the budget constraints faced by an individual or household (remember that we are using the two terms interchangeably). There are two household budget constraints. The first applies in any given period: ultimately, you must either spend the income you receive or save it; there are no other choices. For example,

disposable income = consumption + household savings.

Households also face a lifetime budget constraint. They can save in some periods of their life and borrow/dissave in other periods, but over the course of any household’s lifetime, income and spending must balance. The simplest case is when real interest rates equal zero, which means that it is legitimate simply to add together income and consumption in different years. In this case the lifetime budget constraint says that

total lifetime consumption = total lifetime income.

If real interest rates are not zero, then the budget constraint must be expressed in terms of discounted present values. The household’s lifetime budget constraint is then

discounted present value of lifetime consumption = discounted present value of lifetime income.

If the household begins its life with some assets (say a bequest), we count this as part of income. If the household leaves a bequest, we count this as part of consumption. As in our earlier numerical example, we can think about the lifetime budget constraint in terms of the household’s assets. Over the course of a lifetime, the household can save and build up its assets or dissave and run down its assets. It can even have negative assets because of borrowing. But the lifetime budget constraint says that the household’s consumption and saving must result in the household having zero assets at the end of its life.

To see how this budget constraint works, consider an individual who knows with certainty the exact number of years for which she will work (her working years) and the exact number of years for which she will be retired (her retirement years). While working, she receives her annual disposable income—the same amount each year. During retirement, she receives a Social Security payment that also does not change from year to year. As before, suppose that the real interest rate is zero.

Her budget constraint over her lifetime states that

total lifetime consumption = total lifetime income = working years × disposable income+ retirement years × Social Security payment.

Our numerical example earlier was a special case of this model, in which

disposable income = $34,000, working years = 45, retirement years = 15,

and

Social Security payment = $18,000.

Plugging these values into the equation, we reproduce our earlier calculation of lifetime income (and hence also lifetime consumption) as ($45 × $34,000) + (15 × $18,000) = $1,800,000.

The Life-Cycle Model of Consumption

Economists often use a consumption functionA relationship between current disposable income and current consumption. to describe an individual’s consumption/saving decision:

consumption = autonomous consumption + marginal propensity to consume × disposable income.

The marginal propensity to consumeThe amount consumption increases when disposable income increases by a dollar. measures the effect of current income on current consumption, while autonomous consumption captures everything else, including past or future income.

The life-cycle modelA model studying how an individual chooses a lifetime pattern of saving and consumption given a lifetime budget constraint. explains how households make consumption and saving choices over their lifetime. The model has two key ingredients: (1) the household budget constraint, which equates the discounted present value of lifetime consumption to the discounted present value of lifetime income, and (2) the desire of a household to smooth consumption over its lifetime.

Let us see how this model works. According to the life-cycle model of consumption, the individual first calculates her lifetime resources as

working years × disposable income + retirement years × Social Security payment.

(We continue to suppose that the real interest rate is zero, so it is legitimate simply to add her income in different years of her life.) She then decides how much she wants to consume in every period. Consumption smoothing starts from the observation that people do not wish their consumption to vary a lot from month to month or from year to year. Instead, households use saving and borrowing to smooth out fluctuations in their income. They save when their income is high and dissave when their income is low.

Perfect consumption smoothing means that the household consumes exactly the same amount in each period of time (month or year). Going back to the consumption function, perfect consumption smoothing means that the marginal propensity to consume is (approximately) zero.With perfect consumption smoothing, changes in current income will lead to changes in consumption only if those changes in income lead the household to revise its estimate of its lifetime resources. If a household wants to have perfectly smooth consumption, we can easily determine this level of consumption by dividing lifetime resources by the number of years of life. Returning to our equations, this means that

$$\text{consumption}=\frac{\text{lifetime income}}{\text{working years + retirement years}}\text{.}$$

This is the equation we used earlier to find Carlo’s consumption level. We took his lifetime income of $1,800,000, noted that lifetime income equals lifetime consumption, and divided by Carlo’s 60 remaining years of life, so that consumption each year was $30,000. That is really all there is to the life-cycle model of consumption. Provided that income during working years is larger than income in retirement years, individuals save during working years and dissave during retirement.

This is a stylized version of the life-cycle model, but the underlying idea is much more general. For example, we could extend this story and make it more realistic in the following ways:

• Households might have different income in different years. Most people’s incomes are not constant, as in our story, but increase over their lifetimes.
• Households might not want to keep their consumption exactly smooth. For example, if the household expects to have children, then it would probably anticipate higher consumption—paying for their food, clothing, and education—and it would expect to have lower consumption after the children have left home.
• The household might start with some assets and might also plan to leave a bequest.
• The real interest rate might not be zero.
• The household might contain more than one wage earner.

Working through the mathematics of these cases is more complicated—sometimes a lot more complicated—than the calculations we just did, and so is a topic for advanced courses in macroeconomics. In the end, though, the same key conclusions continue to hold even in the more sophisticated version of the life-cycle model:

• A household will examine its entire expected lifetime income when deciding how much to consume and save.
• Changes in expected future income will affect current consumption and saving.

The Government Budget Constraint

The household’s budget constraints for different years are linked by the household’s choices about saving and borrowing. Over the household’s entire lifetime, these individual budget constraints can be combined to give us the household’s lifetime budget constraint. Similar accounting identities apply to the federal government (and for that matter, to state governments and local governments as well).

In any given year, money flows into the government sector, primarily from the taxes that it imposes on individuals and corporations. We call these government revenues. The government also spends money. Some of this spending goes to the purchase of goods and services, such as the building of roads and schools or payments to teachers and soldiers. Whenever the government actually buys something with the money it spends, we call these government purchases (or government expenditures). Some of the money that the government pays out is not used to buy things, however. It takes the form of transfers, such as welfare payments and Social Security payments. Transfers mean that dollars go from the hands of the government to the hands of an individual. They are like negative taxes. Social Security payments are perhaps the most important example of a government transfer.

Any difference between government revenues and government expenditures and transfers represents saving by the government. Government saving is usually referred to as a government surplus:

government surplus = government revenues − government transfers − government expenditures.

If, as is often the case, the government is borrowing rather than saving, then we instead talk about the government deficit, which is the negative of the government surplus:

government deficit = −government surplus = government transfers + government expenditures − government revenues.

Applying the Tools to Social Security

The life-cycle model and government budget constraint can be directly applied to our analysis of Social Security. Let us go back to Carlo again. Carlo obtains pretax income and must pay Social Security taxes to the government. Carlo’s disposable income in any given year is given by the equation

disposable income = income − Social Security tax.

Imagine that he receives no retirement income other than Social Security. Carlo’s lifetime resources are given by the following equation:

lifetime resources = working years × income − working years × Social Security tax+ retirement years × Social Security income.

Now let us examine Social Security from the perspective of the government. To keep things simple, we suppose the only role of the government in this economy is to levy Social Security taxes and make Social Security payments. In other words, the government budget constraint is simply the Social Security budget constraint. The government collects the tax from each worker and pays out to each retiree. For the system to be in balance, the government surplus must be zero. In other words, government revenues must equal government transfers:

number of workers × Social Security tax = number of retirees × Social Security payment.

Now, here is the critical step. We suppose, as before, that all workers in the economy are like Carlo, and one worker is born every year. It follows that

number of workers = working years

and

number of retirees = retirement years.

But from the government budget constraint, this means that

working years × Social Security tax = retirement years × Social Security payment,

so the second and third terms cancel in the expression for Carlo’s lifetime resources. Carlo’s lifetime resources are just equal to the amount of income he earns over his lifetime before the deduction of Social Security taxes:

lifetime resources = income from working.

No matter what level of Social Security payment the government chooses to give Carlo, it ends up taking an equivalent amount away from Carlo when he is working. In this pay-as-you-go system, the government gives with one hand but takes away with the other, and the net effect is a complete wash. We came to this conclusion simply by examining Carlo’s lifetime budget constraint and the condition for Social Security balance. We did not even have to determine Carlo’s consumption and saving during each year. And—to reiterate—the assumption that there is just one person of each age makes no difference. If there were 4 million people of each age, then we would multiply both sides of the government budget constraint by 4 million. We would then cancel the 4 million on each side and get exactly the same result.

We have gained a remarkable insight into the Social Security system. The lifetime income of the individual is independent of the Social Security system. Whatever the government does to tax rates and benefit levels, provided that it balances its budget, there will be no effect on Carlo’s lifetime income. Since consumption decisions are made on the basis of lifetime income, it also follows that the level of consumption is independent of variations in the Social Security system. Any changes in the Social Security system result in changes in the level of saving by working households but nothing else. As we saw in our original numerical example, individuals adjust their saving in a manner that cancels out the effects of the changes in the Social Security system.

The model of consumption and saving we have specified leads to a very precise conclusion: the household neither gains nor loses from the existence of the Social Security system. The argument is direct. If the well-being of the household depends on the consumption level over its entire lifetime, then Social Security is irrelevant since lifetime income (and thus consumption) is independent of the Social Security system.

The Baby Boom

During the period directly following World War II, the birthrate in many countries increased significantly and remained high for the next couple of decades. People born at this time came to be known—for obvious reasons—as the baby boom generation. The baby boomers in the United States and the United Kingdom are clearly visible in Figure 13.6 "The Baby Boom in the United States and the United Kingdom", which shows the age distribution of the population of those countries. If babies were being born at the same rate, you would expect to see fewer and fewer people in each successive age group. Instead, there is a bulge in the age distribution around ages 35–55. (Interestingly, there is also a second baby boomlet visible, as the baby boomers themselves started having children.)

Figure 13.7 "The US Baby Boom over Time" presents the equivalent US data for 1980, 1990, and 2000, showing the baby boom working its way through the age distribution.

As the baby boom generation makes its way to old age, it is inevitable that the dependency ratioThe ratio of retirees to workers.—the ratio of retirees to workers—will increase dramatically. In addition, continuing advances in medical technology mean that people are living longer than they used to, and this too is likely to cause the dependency ratio to increase. The 2004 Economic Report of the President predicted that the dependency ratio in the United States will increase from 0.30 in 2003 to 0.55 in 2080.Economic Report of the President (Washington, DC: GPO, 2004), accessed July 20, 2011, http://www.gpoaccess.gov/usbudget/fy05/pdf/2004_erp.pdf. Roughly speaking, in other words, there are currently about three workers for every retiree, but by 2080 there will only be two workers per retiree.

Social Security Imbalances

To completely understand both the current situation and the future evolution of Social Security, we must make one last change in our analysis. Although the Social Security system was roughly in balance for the first half-century of its existence, that is no longer the case. Because payments are calculated on the basis of past earnings, it is possible for revenues to exceed outlays or be smaller than outlays. This means that the system is not operating on a strict pay-as-you-go basis.

When the government originally established Social Security, it set up something called the Social Security Trust Fund—think of it as being like a big bank account. Current workers pay contributions into this account, and the account also makes payments to retired workers. Under a strict pay-as-you-go system, the balance in the trust fund would always be zero. In fact, in some years payments to workers are smaller than tax receipts, in which case the extra goes into the Trust Fund. In other year payments exceed receipts, and the difference is paid for out of the Trust Fund.

To be more precise,

tax revenues = number of workers × Social Security taxes = number of workers × tax rate × income

and

Social Security payments = number of retirees × Social Security payment.

If tax revenues exceed payments, then the system is running a surplus: it is taking in more in income each period than it is paying out to retirees. Conversely, if payments exceed revenues, the system is in deficit. In other words,

Social Security surplus = number of workers × tax rate × income − number of retirees × Social Security payment.

For the first half-century of Social Security, there was an approximate match between payments and receipts, although receipts were usually slightly larger than payments. In other words, rather than being exactly pay-as-you-go, the system typically ran a small surplus each year.“Trust Fund Data,” Social Security Administration, January 31, 2011, accessed July 20, 2011, http://www.ssa.gov/OACT/STATS/table4a1.html. Over the first half-century of the program, the Trust Fund accumulated slightly less than $40 billion in assets. This might sound like a big number, but it amounts to only a few hundred dollars per worker. The Social Security Trust Fund contains the accumulated surpluses of past years. It gets bigger or smaller over time depending on whether the surplus is positive or negative. For example,

Trust Fund balance this year = Trust Fund balance last year + Social Security surplus this year.

(Strictly, that equation is true provided that we continue to suppose that the real interest rate is zero.) If tax revenues exceed payments, then there is a surplus, and the Trust Fund increases. If tax revenues are less than payments, then there is a deficit (or, to put it another way, the surplus is negative), so the Trust Fund decreases.

The small surpluses that have existed since the start of the system mean that the Trust Fund has been growing over time. Unfortunately, it has not been growing fast enough, and in 2010, the fund switched from running a surplus to running a deficit. There are still substantial funds in the system—almost a century’s worth of accumulated surpluses. But the dependency ratio is so high that those accumulated funds will disappear within a few decades.

Resolving the Problem: Some Proposals

We can use the life-cycle model of consumption/saving along with the government budget constraint to better understand proposals to deal with Social Security imbalances.

We saw that the surplus is given by the following equation:

Social Security surplus = number of workers × tax rate × income − number of retirees × Social Security payment.

The state of the Social Security system in any year depends on five factors:

1. The level of income
2. The Social Security tax rate on income
3. The size of the benefits
4. The number of workers
5. The number of retirees

Other things being equal, increases in income (economic growth) help push the system into surplus.The effect of economic growth is lessened because of the fact that Social Security payments are linked to past earnings. Higher growth therefore implies higher payouts as well as higher revenues. Still, on net, higher growth would help Social Security finances. A larger number of the population of working age also tends to push the system into surplus, as does a higher Social Security tax. On the other hand, if benefits are higher or there are more retirees, that tends to push the system toward deficit.

The Uncertainties of Life

A century or two ago, if you were unlucky enough to fall into serious poverty, there was very little in the way of government help, even in the richest countries. You were likely to end up in the poorhouse (sometimes called a workhouse or an almshouse), where you obtained the bare minimum of shelter and food in exchange for grueling work. For those who were old and poor, the poorhouse was a place to die an ignominious death:

Numerous as are the old men’s homes, old ladies’ homes, and homes for aged couples that are supported by private charity, they are yet, as every worker among the poor knows, too few to meet the demand. Our almshouses are also practically homes for the aged poor. Some almshouse inmates became paupers before they were aged, but many of them led independent and self-respecting lives, and even put by something for the future while physically able to earn wages. When wages ceased, savings, if any were made, were used up or else lost in unwise investments, and at the end almshouse relief and the pauper’s grave were preferred to exposure and starvation.Henry Seager, Social Insurance: A Program of Social Reform, Chapter V—“Provision of Old Age,” 1910, accessed August 9, 2011, http://www.ssa.gov/history/pdf/seager5.pdf.

Social Security in the United States and other countries was set up largely to save old people from this fate.

Carlo did not face any of the problems suggested by the quotation. In Carlo’s world there was no uncertainty: working and retirement income were known at the start of his working life, and his dates of retirement and death were also known with certainty. Carlo had no risk of using up all his savings before he died, or of losing his money in “unwise investments.” But Carlo’s world is not the world in which we live. In practice, individuals face enormous uncertainty both about their lifetime income and their consumption needs in retirement.

The mere fact that we live in an uncertain world is not, in and of itself, a reason for the government to intervene. Private insurance markets might be available that allow individuals to purchase insurance to cover themselves against these kinds of risks. As an example, many people have disability insurance that they either purchase individually or obtain through their employer. Disability insurance means that if you are unlucky enough to suffer an accident or illness that prevents you from working, you will still receive income. It is also possible to purchase annuitiesAn asset that pays out a certain amount each year while you are alive and insures against the uncertain time of your death. (which are sort of a reverse life insurance): these are assets that pay out a certain amount each year while you are alive and allow you to insure yourself against the uncertain time of your death.

Early discussions of Social Security highlighted the insurance role of the program. During the Great Depression, it became clear that insurance provided through markets was woefully incomplete. Thus the government created a variety of safety nets, financed by public funds. Social Security was one of these programs. An early pamphlet on Social Security summarizes this view:

In general, the Social Security Act helps to assure some income to people who cannot earn and to steady the income of millions of wage earners during their working years and their old age. In one way and another taxation is spread over large groups of people to carry the cost of giving some security to those who are unfortunate or incapacitated at any one time. The act is a foundation on which we have begun to build security as States and as a people, against the risks which families cannot meet one by one.Mary Ross, “Why Social Security?” Bureau of Research and Statistics, 1937, accessed July 20, 2011, http://www.ssa.gov/history/whybook.html.

Financial sophistication has increased markedly since the 1930s, but insurance markets are still far from perfect, so most people agree that the government should continue to provide the insurance that private markets fail to deliver. As President George W. Bush’s Council of Economic Advisors wrote, “To protect against this risk [of living an unusually long time], a portion of the retirement wealth that a worker has accumulated must be converted into an annuity, a contract that makes scheduled payments to the individual and his or her dependents for the remainder of their lifetimes.”Economic Report of the President (Washingon, DC: GPO, 2004), accessed July 20, 2011, http://www.gpoaccess.gov/usbudget/fy05/pdf/2004_erp.pdf, p. 130. Once we acknowledge two things—(1) there is major uncertainty in life, and (2) insurance markets are lacking—we see a clear role for Social Security.

The Complexity of Optimization

There is another reason to think that our analysis of Carlo was much too simple. For Carlo, it was quite straightforward to determine his optimal level of consumption: all he had to do was to calculate his lifetime income, divide by the number of years of life that he had left, and he knew his optimal level of consumption.

We said earlier that the basic idea of this life-cycle model continues to hold even in a more complicated world, where incomes are not constant, real interest rates are not zero, and consumption needs may vary over one’s lifetime. If you have a PhD in economics, you even learn to solve these problems in a world of uncertainty.

Yet when one considers all the uncertainties of life, the problem certainly becomes very complex. Most individuals do not have PhDs in economics, and most people—even including those with economics PhDs—are not able to forecast their income and consumption needs very accurately. As a result, it seems likely that many people are not capable of making good decisions when they are thinking about consumption and saving over their entire lifetimes. As stated in the 2004 Economic Report of the President, “Some individuals may not be capable of making the relevant calculations themselves and may not be able to enlist the service of a financial professional to advise them.”Economic Report of the President (Washingon, DC: GPO, 2004), accessed July 20, 2011, http://www.gpoaccess.gov/usbudget/fy05/pdf/2004_erp.pdf, p. 130. Social Security can therefore be seen as a program that provides assistance to individuals unable to make optimal decisions on their own.That said, figuring payments under the current social security system is not easy either. To understand why, check out the information on benefits at the Social Security Administration website. “Your Retirement Benefit: How It Is Figured,” Social Security Administration, January 2011, accessed July 20, 2011, http://www.ssa.gov/pubs/10070.html.

In general, economists believe both that people are aware of their own self-interests and are capable of making good decisions. Economists tend to be suspicious of arguments that suggest that the government can make better decisions for people than they can make for themselves. At the same time, research by economists and psychologists suggests that individuals are subject to biases and errors of judgment in their decision making. And if government paternalism makes sense anywhere, then it is likely to be in the context of lifetime saving decisions. After all, we are not talking about deciding which kind of coffee to buy or what price to set for a product this month. There is no room for learning from your mistakes, there are no second chances, and the consequences of error are enormous. In life, you only get old once.

Distortions and Administrative Costs

The key arguments in favor of Social Security are therefore that it provides some insurance that may not be available through private markets and protects people in the face of their inability to make sound decisions when they are planning for the distant future. But just because there are some shortcomings of private insurance and annuity markets, we should not presume that government can do things better. Against the benefits of the Social Security system must also be set some costs.

First, any government program requires resources to operate. It costs about 1 percent of the benefits paid to administer the Social Security system. This is a direct cost of the program. Second—and more interestingly in terms of economics—whenever we have a government scheme that affects the taxes that people pay, there will be some distortionary effects on people’s willingness to work. Taxes lower the relative price of leisure compared to consumption goods, which may induce people to work less. Because Social Security imposes a tax on the incomes of working people, it distorts their choices. This is another cost of the Social Security system.

The Effect on National Savings

There is another effect of Social Security that is much more subtle. It reduces the savings of the nation as a whole. This means less capital and ultimately lower living standards. The intuition is as follows. When individuals save, they make funds available in the financial markets for firms to borrow. Thus saving leads to investment and a buildup of the economy’s capital stock. But as we saw, Social Security reduces the individual incentive to save. People don’t need to save if the government will provide for them in retirement. Furthermore, the taxes being collected by the government are not being used to finance capital investment either; they are being paid out to old workers.

A pay-as-you-go system thus tends to reduce overall national saving. In a fully funded Social Security system, this is not an issue, and indeed this is one of the most compelling arguments in favor of a gradual shift to a fully funded system.

Redistributive Effects of Social Security

Social Security redistributes income in ways that may not be desirable. After all, those who benefit the most from Social Security are those who live the longest. Thus the scheme effectively redistributes money from the unlucky people who die young to the lucky ones who live for a very long time. This is a politically charged argument, for life expectancy is correlated with poverty, race, and sex. The life expectancy of poor African American men is significantly lower than the life expectancy of rich white American women, for example. Social Security may redistribute resources, from poor African American men to rich white American women.

Complications

The world is much more complicated than our simple framework, and we need to make sure that our analysis has not left out some important feature of the real world that would change our conclusion. In this section, we briefly discuss some complications to our model. Some of these complications provide some additional reasons to support a Social Security system; others identify additional costs of the system. However, these additional costs and benefits are much less important than those we have already identified.

Why Do People Disagree about Social Security?

President George W. Bush’s suggestions for reforming the Social Security system encountered a lot of opposition and rapidly became a partisan issue in US politics. Yet it seems as if Social Security is a program that we could analyze completely and carefully using the tools of economics. Why is a basic economic program such as Social Security so politicized?

Some people, of course, will view any proposal from the perspective of politics. There are undoubtedly people who supported President George W. Bush’s proposals not on their merits or demerits but just because they support the Republican Party. Likewise, there are surely Democrats who opposed the president’s proposals simply because they came from a Republican. But leaving such extreme partisan viewpoints aside, there are still good reasons why reasonable people might have different opinions on Social Security:

• People differ in their assessment of the importance of market failure in insurance markets. A key argument for Social Security is that private markets do not permit people to insure themselves against the risk of poverty in old age. Insurance and annuity markets do exist, so some people argue that this failure of markets is no longer very significant. At the same time, it requires financial sophistication to take advantage of these markets. Many people do not have the expertise to use these markets or access to financial professionals who could advise them.
• People differ in their beliefs about whether individuals can make good decisions about lifetime consumption and savings. Economists generally think that individuals are the best judges of their own well-being. As a consequence, economists are suspicious of arguments that suggest that the government knows better than you do how you should make your own private decisions (such as how to manage your money). However, the decision making required for lifetime financial planning is very complicated, and the consequences of error are so severe, that many economists nonetheless think that failures of individual decision making are a good reason to support Social Security.
• People differ in their beliefs about how much government should be involved in people’s lives. Some people are, in general, philosophically opposed to significant government involvement in individual decisions. Even if individuals make poor decisions about their lifetime consumption and savings and end up poor, these people would argue that individuals should bear the consequences of their own mistakes, and government should not bail them out. Others tend to the view that government has a critical role to play in protecting the unfortunate and unlucky.
• People have different views about fairness and equality. Some people have the view that an important function of government is to protect the worst off in society and to redistribute some resources from those who are relatively rich to those who are poorer. Such people tend to be strong supporters of programs such as Social Security because it protects those who, through bad luck or poor decisions, would otherwise end their lives in poverty. Others disagree, saying that government should not be involved in redistribution of resources. They also point out, as we observed earlier, that Social Security, by its very nature, benefits those who live for a long time, so it is not a good deal for groups with lower life expectancies.

Beyond Social Security

You may have heard in the news that discussion of the need to reform Social Security applies to other government programs. In particular, if a part of the Social Security program is a growing imbalance in the age distribution, then other programs that support transfers to older people are potentially in trouble as well.

A leading example of this is the Medicare program.You can find information about this program at Medicare.gov: http://www.medicare.gov/default.asp. This program provides health care to the elderly. A second example is Medicaid, which is also a publically funded program, administered at the state level, to provide health care; this program is intended to provide assistance to poor people.“Medicaid Program—General Information,” US Department of Health and Human Services, June 16, 2011, accessed July 20, 2011, http://www.cms.hhs.gov/MedicaidGenInfo. These programs, like Social Security, entail large outlays by the government. In his testimony in June 2008 to the Senate Finance Committee, Peter Orszag, the director of the CBO, stated the following: “The Congressional Budget Office (CBO) projects that total federal Medicare and Medicaid outlays will rise from 4 percent of GDP [gross domestic product] in 2007 to 12 percent in 2050 and 19 percent in 2082, which, as a share of the economy, is roughly equivalent to the total amount that the federal government spends today. The bulk of that projected increase in health care spending reflects higher costs per beneficiary rather than an increase in the number of beneficiaries associated with an aging population.”“The Long-Term Budget Outlook and Options for Slowing the Growth of Health Care Costs,” Congressional Budget Office, June 17, 2008, accessed July 20, 2011, http://www.cbo.gov/doc.cfm?index=9385. This quote contains two key ideas. First, it seems likely that outlays for these two programs will be growing rapidly over the next 50 or so years. From the CBO projections, the share of spending on Medicare and Medicaid grows while the share of spending on Social Security is basically constant after 2020.This comes from figure 1 of the following testimony: “The Long-Term Budget Outlook and Options for Slowing the Growth of Health Care Costs,” Congressional Budget Office, June 17, 2008, accessed September 20, 2011, http://www.cbo.gov/doc.cfm?index=9385. Second, in contrast to Social Security, the problem is not only demographics. Instead, as noted in the testimony, a significant part of the increased cost of these programs comes from the increases in treatment per individual, rather than the number of individuals.

Thus as you use the tools provided in this chapter to ponder Social Security, keep in mind that other programs have similar budgetary challenges. Long-term solutions are needed either to finance the projected increase in outlays or to reduce the costs of these programs.