As the population ages in the coming decades, growing expenditures for Social Security, Medicare, Medicaid and other public and private retirement security programs will put an unprecedented strain on the nation's resources. At the same time, one of the major engines of economic growth -- an expanding workforce -- will slow substantially due to the large exodus of older workers from the labor force and lower birth rates following the post-World War II baby boom. This combination of factors presents a distinct challenge for the economy in the future, which will be called upon to transfer a large and rising share of real resources from workers to retirees. It will be much easier to find these resources in a healthy growing economy than in a stagnant one. The best way to achieve that result is to increase national saving today. Greater national saving will provide the capital to finance investments, which in turn will increase the amount of goods and services each worker can produce.
It follows that a fundamental goal of Social Security reform should be to increase national saving. Without it, reform is a zero–sum game in which any gain for the Social Security system must come at the expense of the rest of the budget, the economy, and future generations. Raising future taxes for Social Security or accumulating stocks and bonds in new personal accounts will mean little if the economy has not expanded sufficiently to match the consumption demands of an older nation. It is the economy's capacity to grow over the next few decades that offers the greatest security to our aging population.
The imminent retirement of the post-World War II baby-boom generation will set off an era of extraordinary demands on the nation's workers. The Social Security actuaries project that the number of people age 65 or older will nearly double from 2000 to 2030, growing by 34 million. In contrast, the number of workers will rise by only 26 million, or just 17 percent.
The significance of this is reflected by the decline in the number of workers per aged person that is projected to occur by 2030. In 1970, there were 4.4 workers for every aged person. Over the subsequent 35 years, the growth in the number of workers was more than four times that of the aged population, keeping the ratio of workers to aged persons relatively constant. In contrast, the actuaries project that from 2020 to 2030, nearly the reverse will occur: the increase in the aged population will triple that of the working population, and by 2030, there will be only 2.6 workers for every aged person.
Food, clothing, medical care, and many other needed services cannot be stored in advance. To a large extent, therefore, the goods and services that society will consume in 2030 will have to be produced then. However, because of the slowly growing labor force, the economy is not expected to expand as rapidly as it has over the last half century.
Projections of future Social Security costs reflect those demands. As a share of gross domestic product (GDP), its expenditures are projected to rise from 4.3 percent today to 6.1 percent in 2030--an increase of almost 50 percent. Moreover, the demographic shift emerging with the baby boomers' retirement is not merely a temporary bulge in the proportion of elderly to non-elderly but a large step in a long-term shift to an older society caused by increasing longevity and persistently low birth rates. And because other sources of retirement benefits--pensions and health programs, for example--will also be affected, projections of Social Security spending portray a common pattern but only a fraction of the total costs of retirees for society.
The Looming Change in the Ratio of Workers to Retirees (in millions of people)
It has been suggested that if birth rates remain low, the rising costs of the aged might be offset by the lower costs of raising fewer children--the implication being that the overall dependency on workers may not be all that troublesome. Projections of the segments of the population that are not of primary working age (people under 20 and people 65 or older), considered together as a percentage of working-age adults, are not higher than when the baby boomers were in their youth.
But a stable ratio of dependents to workers does not mean that America's aging will impose only a small additional burden on tomorrow's workers. For one thing, the overall dependency ratio obscures the much greater cost of supporting each senior. Studies have shown that governmental income support and health care spending for the aged (by all levels of government) substantially exceed governmental spending associated with educating and providing services to children. As the percentage of the population that is elderly goes up, pressures will build on public budgets. Imposing a tax burden that rises with those costs could impair productivity-enhancing investments and incentives for people to work and produce more. Moreover, the level of dependency on working-age adults is not projected to fall when the baby boomers die as it did when they emerged from childhood.
Finally, aside from cost disparity, there is a profound societal difference between sacrificing to build for the future -- as we do when we spend on children -- and sacrificing to subsidize the living standard of the elderly. Both may be regarded as desirable social goals, but the former is largely an investment that bears returns over time while the latter is largely consumption. Thus, even within a stable overall dependency ratio, the shifting demographics we now face portend a spending burden that is both higher and more consumption-oriented than we experienced during the baby boom.
If the nation is to meet the rising demand for goods and services without imposing stifling tax rates, large deficits, or high inflation, it must invest more to spur productivity so that greater resources can be produced by fewer people. The best way to achieve that is through increased national saving -- public and private. When people and institutions save, they supply the capital to finance investments, which enhances productivity--it increases the amount of goods and services each worker can produce. Without creating new saving, legislative restructuring of Social Security is a zero-sum game.
With this in mind, various reform proposals aim to increase saving by moving from the current pay-as-you-go financing system to a partially funded system. The advantages of doing so are clear. A funded system means higher saving and hence higher productivity and higher national income.
In contrast, Social Security's current pay-as-you-go financing works against higher saving. In the first place, the program's widening cash deficits threaten to trigger a huge new run-up in the publicly-held debt starting in the 2020s. Moreover, many economists believe that Social Security's pay-as-you-go structure discourages household saving, and hence capital formation, because it promises households future benefit income while creating no real economic resources to generate that income. As a result, households put less into other (fully funded) forms of saving.
To one extent or another, almost everybody agrees that greater funding is an essential ingredient in Social Security reform. Indeed, much of the public seems to think that Social Security is already storing up a large trust-fund reserve. This reserve, however, does not constitute real funding. Any trust-fund surplus is immediately lent to Treasury, leaving Congress free to spend the money that it credits to the trust funds. But those credits are simply interest-earning Treasury IOUs whose sole function is to keep track of Social Security's authority to write checks. They constitute a claim on future tax revenues, not economic savings that can be drawn down to finance future benefits.
In an attempt to provide greater funding, some reform proposals would have the federal government buy stocks and bonds for the Social Security trust funds. Others, like that advocated by the President, would have workers buy them using Social Security taxes that have been diverted to personal accounts. As significant as those changes may seem, their likelihood of increasing national saving is largely uncertain.
The motivations for using private securities in reforming Social Security are predictably diverse, given that so much of the debate is about what the role of government should be. However, to the extent that the rationale is to increase saving and hence the well-being of future generations in their retirement, the effectiveness of such plans depends on the extent to which they are genuinely funded.
Earmarking government resources to acquire private securities--whether for government trust funds or personal accounts--may accomplish little. To make a difference, there must be genuine funding, i.e., new saving. It isn't enough to simply credit more Treasury bonds to the trust funds or to redirect existing payroll contributions into marketable securities, with or without personal accounts. Without new saving, a plan cannot increase the productivity of tomorrow's workers, and thus becomes an exercise in pushing liabilities from one pocket to another or from one generation to another.
The draw on society 30 years from now to meet the needs of a large retired population will not be any less simply because private securities could be liquidated then. By itself, using government resources to buy stocks and bonds, without other spending and tax changes, would not automatically lead to an increase in the nation's pool of investment resources.
At first glance, that conclusion may not seem logical: why wouldn't retirees' asset sales make them better off, or "richer"? The answer is that the value of the assets, physical or financial, will depend upon the size of the economy. The assets accumulated by future retirees (current workers) will boost their well-being only to the extent that those assets help the economy grow.
If the money to invest came from having a higher level of government borrowing, new money would not have been added to the nation's pool of investment resources. If it causes people to reduce their private saving because they feel more secure about receiving Social Security benefits, aggregate national saving could fall. In other words, the effect on national saving ultimately would depend on the economic decisions of future lawmakers and individuals, much of which cannot be predicted. While policymakers tend to have strong opinions about what the outcome would be, their positions are largely influenced by speculation about political and personal behavior -- not economic analysis.
Much of the recent Social Security debate has revolved around the question of how to establish claims on future resources rather than on how to increase saving that will produce those resources. The use of private securities to bolster the Social Security trust funds, the creation of personal accounts, the scheduling of future tax increases or suspension of scheduled tax reductions, and reliance on future borrowing by government are all means of financing that prescribe different ways of extracting resources from the economy.
They may have different distributional impacts among members of society (affecting who pays, who receives, and how much), but the method of financing does not directly answer the question of how to produce more. No matter how much nominal wealth can be traded or cashed in to produce given levels of retirement income, it is the amount of goods and services that can be furnished that will determine society's economic well-being. While ideological factors often cloud the debate over these options, the real issue is which is most likely to result in genuine saving. What legal, political, and fiscal incentives best ensure that resources are actually reallocated from the present to the future?
Perhaps the most persuasive case for adding personal accounts to the Social Security system is that it would put the funds involved beyond the reach of government to spend on other programs. If Congress tried to withhold or divert the funds from going into personal accounts, voters would likely object. But adopting personal accounts cannot be viewed as a substitute for the hard choices that are needed. In any true transition to a funded system, workers will have to pay more, retirees will have to receive less, or both. Reform plans that do not face up to this transition cost will not result in new net saving or a larger economy.
Another problem with the recent debate is that it tends to view Social Security in isolation from the rest of the government, the economy, and the other means by which people strive for retirement security. As a result, the focus is on where to get the money to close a gap between its future income and outgo.
If balancing the system's income and outgo were all that mattered, the simplest fix would be to enact a law allowing the Social Security trust funds to draw on the government's general fund--just give them more government bonds. However, there is no excess money in the general fund, and because federal deficits are projected for as far as the eye can see, none is expected in the future. That means there are two recourses in the future for making good on the future benefit promises: raising taxes or borrowing. Neither of those actions is likely to increase the size of the economy… and in all likelihood, they would impair it.
Achieving a long-range balance of Social Security's receipts and expenditures is not nearly as important for ensuring retirement security as is the need for the economy to expand. Tax increases that impede business investment, personal saving, and work may impair productivity. Running budget surpluses and using them to pay down the federal debt, constraining government expenditures that add to consumption, adopting policies that advance productive technology and investment in education, eliminating regulations that inhibit productivity, adopting tax measures that reward personal saving and work, and creating a funded system of personally owned accounts are the types of policies that are likely to have the greatest chance of spurring growth.
That said, even if the economy were to grow at greater than expected rates, it is unlikely by itself to solve Social Security's problem since the program's benefits are automatically tied to wage growth, meaning that if the economy grows faster than expected, future benefits will too. But economic growth will mean there is more productive capacity, allowing for a larger supply of goods and services. It will therefore lessen the burden of meeting those higher costs and the greater consumption demands of the aged generally.
Constraining the rise in Social Security and other entitlement expenditures are choices lawmakers can make to help stimulate that growth. Constraining those programs may seem counter intuitive--how can future retirees be better off if their benefits have been constrained? The answer is that to the extent the economy grows larger as a consequence of those decisions, the benefits paid in the future may retain or have even greater value. Social Security benefits that absorb 4 percent of GDP in an economy that has doubled may be richer than those absorbing 6 percent of an economy that has seen only modest expansion.
But why constrain Social Security and other programs for the aged? Why not hold down other government spending? The reason is that that is where the money is. Absent policy changes, within 30 years Social Security, Medicare, and Medicaid could absorb a share of the economy roughly equal to 80 percent of what the entire federal government spends today. Social Security will only be affordable in the future if the rest of government is affordable.
Genuine funding requires genuine resource trade-offs. To save more, we must temporarily consume less--at least until the productivity benefits of higher saving kick in. If lawmakers schedule future tax increases to pay for future consumption, it won't increase national saving today or increase the size of the economy tomorrow. If they divert current taxes into new personal accounts and borrow to make up the difference, national saving will not increase -- and may even decline if individuals save less in other areas because of the change. Taking such approaches is gambling that future lawmakers and individuals will make the necessary decisions to cut consumption. That's passing the buck. Actions are needed today to fuel the nation's growth engine, meaning that politically difficult steps need to be taken now.
Raising taxes today might be a tool to cut consumption and reduce the federal debt if one were convinced that lawmakers in the next Congress wouldn't spend the money or cut other taxes in response. Unfortunately, the track record in this regard is not promising. Slowing the growth of spending is uncertain as well, but probably more viable given the strains that projected entitlement growth will put on the tax base. Slowing the long-term growth of entitlement spending would help prevent escalating debt, ease the pressure for much higher taxes, and leave more resources available in the financial markets for investment. Creating new personal accounts and requiring people to fund them with personal income that would otherwise be used for current consumption would further enhance saving.
Economic growth will not occur without sacrifice. Governments, businesses and individuals can do two things with their money: they can buy things that are consumed immediately, or they can invest the money to enable them to produce more in the future. Simply put, there is no free lunch in achieving national saving… people can spend their money or save it. The same is true when making government policy.
 These numbers and all others herein are based on the so-called "intermediate" or central forecast of the 2005 Social Security trustees' report.