Paying for It

There aren’t many mysteries about the financial challenges posed by the aging of America’s population. While little consensus exists on how to shore up Social Security, there is widespread understanding that the system will be in deficit within a decade of the first baby boomers’ retirements, which start in 2008. The Medicare financing outlook is even bleaker; the federal health-insurance program for the elderly is already in the red even as a costly new prescription drug benefit is being implemented. Front-page stories about corporate pension plans that go belly up or are cut back, at the same time that retiree health-benefit programs are curtailed, add to the general anxiety.

But perhaps the biggest concern Americans should have about their retirement system is the sheer inertia that has prevented the nation from addressing its problems. For more than two decades, we have known about the demographic challenges facing Social Security. We knew before prescription drug benefits were added to Medicare coverage that the system was in trouble. It makes for a sad spectacle indeed that we enjoy the rare advantage of being able to see the future with clarity yet are unwilling to act.

Meanwhile, other countries have started to address some of the same challenges, and they have done so with greater inventiveness and determination than the United States has shown. The list of pioneers ranges from the familiar example of Chile to the less noted examples of Sweden, Germany, and Canada. All offer lessons from which America can learn.

By some measures, America’s aging problem is relatively minor compared with what other developed countries face. For every retired person in America, there are currently about four working people. (Australia, Canada, and the United Kingdom have similar ratios.) In Japan, the ratio is closer to three workers per retired person; in Italy, it’s down close to two, and Germany is not much better off than that.

The demographic future also looks at least as favorable for the United States as for any other developed country. The retirement burden on American workers is not expected to be any greater in 2030 than it already is today in Germany or Italy. By that year, Germany’s burden is expected to be twice and Italy’s 2.5 times America’s. Italy will have only one active worker per retiree. Other developed countries, such as Switzerland and the Netherlands, will be in better positions than that but will still face bigger burdens than the United States.

Perhaps the most important lesson to be learned from abroad comes from countries that are less reliant on pensions to provide income security to older people. They do not rely as much on pensions for the simple reason that many older people in these countries are still working. Japan presents an especially interesting case. It has the most rapidly aging population in the developed world, but its retirement burden in 2030 is expected to be attenuated because older people in Japan tend to work later into life than their counterparts in most other developed countries. The average retirement age for Japanese men is nearly 67.

The undeniable fact is that many people, especially in the world’s developed economies, are retiring at ages when they could still be highly productive. An astonishing 38 percent of Italians in the 50-to-54-year-old bracket are already out of the labor force, and hardly anyone in Italy between 60 and 64 still works. Granted, Italy is an extreme case; even in Sweden, workers are less likely to retire early. But the trend toward early retirement is widespread.

The most significant influence on when the majority of people leave work behind is the structure of the retirement system: An earlier “normal” retirement age or more generous benefits for early retirement lead, predictably, to more retirements. In Iceland, 81 percent of the population between the ages of 60 and 64 is still in the labor force, largely because of incentives in the retirement system that encourage people in that age bracket to continue working. Even in the United States, 47 percent are still economically active during those years. In Europe, the comparable numbers range from 22 percent in Germany to only 14.5 percent in France.

In Iceland, the average man works until age 67. In France, his counterpart retires at 59. Such variations help explain the range of costs associated with different retirement systems. In France, the average remaining life expectancy for a male at the typical retirement age is 20.5 years. It is 13.7 years in Iceland. All else being equal, a male retiree in France will cost about 60 percent more in retirement benefits than one in Iceland simply because of the longer duration of retirement.

Pension payouts are not the only cost of retirement. By withdrawing their labor from the economy, retirees also slow economic growth, making it harder to underwrite retirement costs. If a country has a growing elderly population that does not contribute to national output and workers have an incentive to retire in their fifties, the particular method of financing retirement—pay-as-you-go, like Social Security, or fully funded retirement accounts—doesn’t matter a great deal. There will, in any case, be trouble down the road.

Most government-run retirement systems around the world operate similarly to the U.S. Social Security system: Benefits paid to current retirees are financed on a pay-as-you-go basis, out of revenues from current workers. There’s little or no “money in the bank.” Beneath this level of general similarity, however, there are some significant differences in how public pension systems are structured, how large a role private pension plans play, and, most important for our purposes, how willing policymakers have been to make the hard choices needed to ensure the survival of these systems.

One little-appreciated difference between the American system and most others has to do with its basic architecture. Most other countries have two-tier publicly financed retirement systems. The United States has a single tier.

In most of those other countries, virtually every citizen is entitled to a basic first-tier benefit upon reaching retirement age. In some cases, this benefit comes in the form of a universal “demogrant”; in others, recipients face income or assets tests that may reduce the size of the benefit or, in some cases, eliminate it.

On top of this basic benefit, these countries provide a separate, second tier of retirement benefit that is proportional to a worker’s preretirement earnings. The second-tier benefit usually accounts for the largest share of the total retirement pension.

What’s important is that the first tier is explicitly funded by general tax revenues, not by earmarked employer and employee contributions. As a result, the first-tier basic benefit seems to voters less like “their” own money, less like something they are entitled to. And that makes it politically easier for policymakers to adjust this part of the retirement system.

In the United States, the whole Social Security benefit is based on earnings, but it includes elements of the first- and second-tier benefits provided in other countries. Low-wage workers receive larger benefits relative to their earnings than those with high earnings, an arrangement that hides the character of the implicit basic benefit in our system. Yet workers with higher levels of covered earnings receive larger absolute benefits, which links payouts to covered earnings levels, albeit indirectly. Virtually all benefits are financed through payroll tax contributions. The structure of the U.S. system has led people to believe that they have “paid for” their benefits even though the benefits are highly subsidized.

Such perceptions matter a great deal. In Canada, for example, legislators were able several years ago to impose a means test on the basic pension through the income tax system, reducing benefits for higher-income retirees. Australia, too, has cut back its basic benefit, though with a twist. All Australians are still entitled to a relatively generous basic benefit, set at roughly 25 percent of the average worker’s earnings, but it is subject to a means test that reduces recipients’ benefits as their earnings and assets increase. In the past, Australians nearing retirement age did everything they could to avoid losing out under the means test—going on shopping sprees, taking round-the-world voyages, or doing whatever else it took to jettison some of their wealth. As a result, Australia in the early 1990s implemented a mandatory retirement savings program designed literally to force workers to save so much that many would fail the asset and income tests.

There has been a lot more innovation around the world in the second tier of public pensions. These pay-as-you-go programs are usually financed by contributions from both employers and employees, and they have traditionally promised workers a future “defined benefit” linked to their earnings during their working years. Social Security is such a system. Americans of working age are reminded of this every year when they receive a statement in the mail from the Social Security Administration showing, based on their contributions so far, how big a Social Security check they can expect in retirement. Such defined-benefit programs are in the deepest trouble, and reforms are under way from Canada to Japan and Sweden.

Some countries have shifted their retirement programs to a more fully funded basis—meaning they are now putting “money in the bank” to pay for future benefits. In the late 1990s, for example, Canada raised the contribution levels of employers and employees and created a separate, non-government board to invest the the money in stocks and bonds, seeking higher returns. The United States has been reluctant to pursue similar policies because of political concerns over how the government might manage what could become the largest pool of investment capital in the world.

Traditional defined-benefit programs guarantee people a certain fixed payout in the future. But several countries have moved toward a bigger role for “defined-contribution” plans. In such systems, the ultimate payout is not fixed in advance but is determined by the level of return on the invested money. In some countries, employers select the defined-contribution plan provider. In others, individuals choose from a series of authorized private-fund managers. The investment risk in both cases is largely borne by the individual account holder, although some systems provide a guaranteed minimum return of some sort. The theory here is that at retirement, workers will simply receive benefits in accordance with what they have contributed.

Since the 1980s, a number of countries have even adopted defined-contribution plans as the primary element of their second tier. The example that has received the most attention is Chile, which radically reformed its pension system in 1981. In essence, the Chileans transformed their severely dysfunctional pay-as-you-go system into private individual retirement accounts. The accounts are mandatory, fully funded, fully vested, and portable. Workers must contribute 10 percent of their annual earnings to their retirement accounts and choose where to invest their savings from funds offered by highly regulated, specialized, private-fund management companies. Workers also must purchase term life insurance and disability insurance from their pension managers. All told, the package comes to about 13 percent of gross pay. At retirement, participants have a retirement nest egg whose size will vary depending on how the investments have performed.

Two main criticisms have been leveled against the Chilean system: The fund management companies have extracted very high fees, and large numbers of Chileans are not covered because they are self-employed or too poor, or for other reasons. However, the system has been reasonably successful, with many workers enjoying more secure retirement prospects than they did under the old system. A number of other Latin American countries have followed the Chilean example.

Roughly a decade after Chile acted, Australia embarked on an equally radical reform. Beginning in 1992, employers were obliged to contribute a share of workers’ pay (now nine percent) to privately operated retirement funds. As a result, most workers will accumulate enough assets so that they will not qualify for Australia’s means-tested first-tier basic pension until much later in their retirement (as they exhaust their savings) than under prior policy. Australian voters accepted this change because there would not be significant reductions in basic pensions for many years and because those affected would have a bigger pot of personal savings at retirement.

Sweden created supplemental private accounts without much fuss during the mid-1990s as part of a larger reform of its pension system; it now requires that workers contribute 2.5 percent of covered pay to individual accounts invested through government-approved investment managers. At the same time, Sweden made a revolutionary change in the traditional pay-as-you-go system that still provides the main pension benefit for Swedes. Under the new setup, workers continue to pay contributions to the system, and they amass retirement accounts—even though in reality their contributions are quickly paid out to current retirees. But as the change is phased in, workers’ accounts will be treated as what they really are: virtual or “notional” accounts, with no assets backing them. The accounts are bookkeeping devices. As a result, there is no defined benefit. Benefits are not determined until a worker reaches age 61, when they will be set in a two-step process.

In the first stage, a basic benefit will be determined on the basis of the group’s remaining life expectancy. As life expectancy increases in the future, monthly pensions will decrease to keep the total lifetime value of pensions relatively constant. In the second step, the benefit will also be adjusted so that each age group’s expected lifetime benefits will be covered by what is anticipated in worker contributions during the course of its retirement period. This latter demographic adjustment will keep the system in balance as the number of retirees grows relative to workers. The result will almost certainly be a reduction in benefits, as well as much less strain on the Swedish economy.

The individual accounts Sweden mandated in the 1990s were not hugely controversial. Political leaders   educated the public about the need for reform, and the accounts were part of a larger overhaul. The system’s alterations were to be implemented in the future on a gradual basis. It is remarkable that attempts in the United States to shift toward such supplemental personal accounts in our Social Security system, most recently by the Bush administration, have proved so controversial, denounced as somehow being an abrogation of the “social” character of the U.S. system. The accounts that most American supporters have talked about range from 2 percent to 4 percent of earnings, not much different from the 2.5 percent Sweden has legislated.

Policymakers in Germany and Japan took a long, hard look at the Swedish example several years ago when they grappled with their own painful reforms. Neither was willing to move to Swedish-style notional accounts, but in both cases they adopted provisions for their defined-benefit systems that mimic the demographic adjustment the Swedes will use in determining future benefits. German and Japanese leaders took that route, most observers agree, in part because they thought it would be easier to get people to accept reforms if it were not so clear what the implications of the changes would be.

Americans are now experiencing—unwittingly, for the most part—the effects of just such a delayed reform undertaken in 1983, during Ronald Reagan’s presidency. Congress raised the age at which Social Security benefits would be paid, but it deferred implementation until 2000. As a result of the reforms, people retiring this year will need to wait until eight months after their 65th birthday to stop working if they want to receive full benefits; people who are now 46 or younger will need to wait until their 67th birthday. There is a lesson in the fact that this delayed reform provoked virtually no protest in 1983 or when it began phasing in 17 years later.

Private pensions are the other big element involved in thinking about how to pay for retirement. Such pensions loom larger in the retirement landscape of some countries, such as the United States, than others, but it seems clear that their role is likely to grow everywhere. As a rule, where public pensions are rich, private pensions tend to be spare, and where public plans are small, private pensions tend to be more substantial. In the past, it made little sense for a German employer to provide a generous pension, for example, since the country’s public pension already allowed workers to retire at virtually the same standard of living they enjoyed while working.

But the remorseless logic of demographic change has led even the Germans to try to strike a new balance between public and private pensions. As part of a controversial package of public-pension tax increases and benefit cutbacks in 2000, the Social Democratic government of Chancellor Gerhard Schröder created tax incentives designed to encourage employers to establish voluntary supplemental retirement savings plans for workers and thus to take some of the pressure off the public system. Schröder’s successor, Angela Merkel, recently declared that private pensions must be increased and that it will be necessary for Germany to push the normal retirement age up to 67.

Because Americans as a whole are not as dependent on the national pension, the United States enjoys some advantages in addressing the challenge of an aging society. Shoring up Social Security should be a much more manageable task than it has been in countries such as Germany and Sweden, where retirees are heavily dependent on public pensions. And mandating private pensions or savings programs, as many countries have already done, should not be as disruptive or expensive as it might be in countries where employers and workers do not already have a lot of experience with voluntary private plans. Requiring that all workers have supplemental accounts to bolster the national pension, as Sweden has done, would seem to play to the generally successful American experience with 401(k) plans.

What remains an obstacle is the fear that such accounts will unduly expose workers to the risks of the stock and bond markets. There are risks in financial markets, and anyone who argues otherwise is misleading us about the choices we face. At the same time, many Americans have forgotten the painful 1977 Social Security reforms that reduced benefits for the “notch babies” (people born between 1917 and 1922) by as much as 20 percent relative to prior law. If U.S. workers aren’t required to save some added amount of their pay in individual accounts, the Social Security benefit reductions that loom in the future probably will be far more widespread. If America doesn’t address this problem soon, the eventual cutbacks will likely have to include some people on the verge of retirement, if not those already retired. Every path has risks, but the risks are greater in doing nothing until the onset of a crisis.

If the United States does not take to heart the lessons that some other countries have learned, it will be forced to repeat the unpleasant experiences of those that refused to act until there was no alternative. The Germans discovered that they had no choice but to reduce pension benefits, not just for future retirees but for existing ones. The Japanese learned that legislators confronted with such a crisis may come to blows on the floor of parliament. The future that looms before the United States is neither a blur nor a mystery. Its outlines can be seen with all the clarity of an actuarial table, and so can the choices.