Social Security Reform Around the World: Lessons from Other Countries

Social Security reform is one of the most prominent domestic policy issues in the United States. The U.S. is not alone in facing the daunting challenges posed by its retirement security program. The gross implicit unfunded debts of retirement security programs in developed countries (that is, the amounts they owe current pensioners and workers) far surpass the official explicit federal debt. The same is true of the net implicit liabilities, adjusted for projected future rights and contributions.

• The United States' gross public pension debt is more than 100 percent of Gross Domestic Policy (GDP).
• Public pension debts in Italy and Japan are more than double their official federal debts and the two combined exceed 240 percent of GDP.
• France and Germany shoulder public pension debts four times the size of their official federal debts.

Paying this debt will put a huge strain on public treasuries and on these economies more broadly. Under traditional systems, either benefits will have to be cut below scheduled levels, or contributions will have to rise substantially.

Recognizing the problem, 20 countries have made funded private retirement accounts (PRAs) for 80 million workers part of their mandatory retirement security programs. Many other countries are about to do the same. The reformed systems in these countries have many differences, but share certain characteristics:

• The private retirement accounts represent a partial shift from a pay-as-you-go system to a prefunded system.
• They represent a partial shift from defined benefit (DB) plans to defined contribution (DC) plans under which retirement income is determined by contributions plus investment earnings.
• The account balances, and their contribution to national saving and economic growth, increase through time.
• Contributions to the private accounts are compulsory for new workers and sometimes for current workers as well; the investments are regulated; and the private accounts are integrated with the remaining public program.
• Future benefits to retirees under the new social security system come from two places the public DB and the annuity from the individual's own account. Part of the pension is financed by assets in the accounts, rather than by contributions coming from young workers.

Reformers in the United States can take several lessons from the experiences of other countries around the world. In constructing their systems of funded private accounts, these countries have grappled with problems that we will face, as we decide how to shape our system. Their solutions are not necessarily right for us, but they do show us that solutions exist and they provide us with a menu of tested options.

Lesson 1: Private sector control over personal retirement accounts is more profitable than a centrally controlled pension reserve. The average privately managed pension fund around the world earned a large positive rate of return, that far exceeded inflation and wage growth, during the last 30 years. By contrast, the average publicly managed pension fund lost much of its capital during the same time period. This occurs for two reasons

• Public managers are often required to invest in low-interest government securities, as is the case with the U.S. Social Security system, or are pressured to make politically motivated investments.
• Hidden and exclusive access to centralized pension reserves makes it easier for governments to run larger deficits or spend more wastefully than they could if they had to rely on a more accountable source of funds.

Competitively managed private pension plans are more likely to invest in diversified portfolios and to resist political manipulation.

Lesson 2: Individual accounts can be created in a way that minimizes administrative fees.

• In Chile, administrative costs relative to assets have fallen substantially as assets have grown, and now amount to less than 1 percent of assets per year for an average worker who contributes for 40 years. This is lower than the fee paid by the average mutual fund investor in the US.
• Most Organization for Economic Cooperation and Development (OECD) countries, like Switzerland, Australia, Denmark and the Netherlands, use group plans with an employer and/or union choosing the investment manager, which often produces administrative costs that are far lower than those in Chile.
• Bolivia used a competitive bidding process to choose two investment managers, thereby cutting out most marketing costs.
• In Sweden, pension authorities established a maximum fee schedule for fund managers and mandated central collections and record keeping. Administrative costs are projected to be less than 0.5 percent of assets per year in the long run.

Lesson 3: Personal retirement accounts, if structured properly, do not involve undue risk. Reforming countries typically reduce risk by 1) encouraging diversification of pension fund investments; 2) guaranteeing absolute or relative returns; and /or 3) instituting a pension floor or other benefit from tax-financed sources that supplements the personal account.

• Chile and other Latin American countries started with a list of quantitative regulations over permissible fund investments but they have gradually liberalized.
• Switzerland requires a nominal return of at least 4 percent of assets, over the worker's lifetime with an employer.
• Several Latin American countries provide a minimum pension guarantee; Argentina provides a flat (uniform) public pension to all eligible workers; Australia sets a floor through a means and asset-tested old age pension.

Lesson 4: Reformed systems can continue the redistribution of income. In some cases, the reformed systems redistribute income from high earners to low-income earners better than traditional pay-as-you-go programs, which are biased against people with shorter life expectancies.

• Argentina's flat pension (about 25 percent of the average wage to all workers with at least 30 years of contributions), disproportionately benefits low-income workers. Chile's minimum pension guarantee goes mainly to low earning women.
• Australia's means and asset-tested public old age benefit goes to the bottom 2/3ths of workers in the income distribution.
• In Switzerland, a public benefit that is almost flat, financed by a payroll tax that has no ceiling, accompanies the personal accounts, so the net result is a very redistributive old age pension that also reduces risk.

Lesson 5: Reform involves transition costs. Borrowing temporarily to transition to a funded system does not increase the size of a country's total public debt. Instead, it transforms a hidden implicit debt into a transparent explicit debt. Because this pension debt stops growing when personal accounts are created, eventually it can be paid off.

All Latin American and Eastern European countries have funded their personal account systems by moving some portion of the workers' contributions from the public system to the private, in what is known as the carve-out approach. This creates a transition cost. But in the course of reform, the total pension obligation of the government has actually been reduced in almost every country. By contrast, most OECD countries have avoided the transition cost by mandating additional contributions. They can use this approach because they started with a modest public benefit and contribution rate.

Currently, some 80 million workers in 20 countries have access to personal retirement accounts. These countries include Chile, the United Kingdom, Switzerland, Denmark, the Netherlands, Argentina, Colombia, Peru, Bolivia, Mexico, Uruguay, Australia, Hungary, Kazakhstan, Poland, Latvia, Sweden, Hong Kong, El Salvador and Croatia (roughly in the order in which they adopted the plans).

Macedonia, the Dominican Republic, Kosovo and even China have passed reform laws, which they are now in the process of implementing. Other countries are moving in that direction. Interestingly, the United States is not yet on this list.

[page]

¹The presidential election of 2000, followed by the establishment of the President's Commission to Strengthen Social Security, put Social Security reform on the political agenda in the United States. In its final report, the Commission proposed reallocating part of the Social Security payroll tax to individual accounts that are invested in the financial markets – a recommendation that was welcomed by some and attacked by others. The debate is likely to accelerate over the coming year, and significant changes in our current system may be forthcoming.

"Over the past two decades, 20 countries have adopted private funded accounts to partially replace their pay-as-you-go systems."

The United States is not unique in its willingness to pay attention to this issue. Over the past decade, many countries around the world have instituted major structural reforms. This study places the U.S. debate in an international perspective. It explains the problems that have led countries to reform, outlines some of the commonalities and differences in those reforms, and considers their applicability to the United States. The most important commonality is the shift toward greater prefunding in accounts that are privately managed, as an important part of the mandatory Social Security system.

It may come as a surprise to many to learn that, compared with other developed countries that have not yet reformed, the U.S. Social Security system is relatively healthy. Yet it is clearly not sustainable at its current contribution and benefit rates, since cash outflows will exceed inflows in just a few years. The challenge we now face is how to cover this gap in a way that:

• Is fiscally balanced and contains mechanisms to keep it sustainable in the long run;
• Is good for economic growth, thereby increasing the welfare of both old and young;
• Is administered in a cost-effective way;
  Avoids excessive risk – both political and financial market risk; and
• Distributes its costs and benefits equitably.

The experience of other countries may help us find an answer to that challenge.

Cross-sectional analysis shows that public spending on formal pension plans increases exponentially as populations age. [ See Figure I.] In developing countries today, only 2 percent to 3 percent of gross domestic product (GDP) is spent on old age security. But in many industrialized countries this figure already exceeds 10 percent and will grow still higher in the years ahead. Because of its young demographics, U.S. expenditures on old age security are now only 6 percent of GDP, but this number will escalate when the baby boomers start to retire. Public spending on pensions and health (primarily Medicare) combined is already almost double that amount. Programs for the old are by far our largest civilian public programs, and they are destined to grow further.

"Ten more countries are considering the adoption of personal retirement accounts."

With such large sums involved, how this money is generated and spent can affect the entire economy. It influences the quantity and productivity of labor and capital and therefore the size of the GDP pie. For example, high payroll taxes for old age pensions may discourage employment or work effort among the young; and subsidized early retirement may reduce the supply of experienced labor, which will be especially harmful as populations age. In contrast, pension plans that accumulate retirement funds in advance can help to increase long-term national saving, productivity and growth. Also, with such large sums involved and so many people dependent on old age programs, it is important to structure the programs so that they remain fiscally sustainable, even as external conditions such as life expectancy change in unexpected ways.

Increasingly aware of these broad effects, countries have been reforming their systems to have beneficial effects on the economy as well as to provide a more secure old age income. They have done so by a structural reform that adds a funded privately managed defined contribution component to their existing mandatory Social Security systems. The establishment of individual accounts with part of the Social Security tax, which has been proposed in the United States, is a typical form that such a structural reform might take.

[page]

Over the next 30 years, the proportion of the world's population over age 60 will nearly double, from 9 percent to 16 percent. This is due to a sharp drop in the birth rate and an increase in life expectancy. As Figure II shows, almost 30 percent of the population in industrialized Organization for Economic Cooperation and Development (OECD) countries will be over age 60 by 2030. The population of the United States is still young compared with those of most European countries, but the proportion of people over age 60 in the U.S. will increase from 17 percent to 28 percent in the next 30 years.

[page]

"The percentage of the world's population over age 60 will nearly double over the next 30 years."

Most old age security systems established by governments in the past were financed by payroll taxes on a pay-as-you-go (PAYG) basis. This means that the contributions made by today's workers are used to pay the pensions of those who have already retired. The pension benefits are defined according to a formula based on the worker's earnings and years of service. These systems, including the U.S. Social Security system, have helped many old people live out their senior years with dignity and security. However, these systems also generate many problems. The problems were small when the systems were immature and small, but they can no longer be overlooked as the systems mature and grow. The problems concern sustainability, growth and equity.

Sustainability. Most public attention has focused on the sustainability problems. Pay-as-you-go systems are very sensitive to demographic change and system maturation. Under a pay-as-you-go system,

C = B/S = B*D, where

C = the contribution rate (as a percentage of wages) required to balance the books in a given year,

B = the average benefit (as a percentage of average wage),

S = the support ratio (the number of workers per retiree) and

D = the dependency ratio (the number of retirees per worker).

As the system matures and populations age, which is currently happening almost everywhere, the dependency ratio grows and the revenues are eventually insufficient to pay the promised benefits. Then either B must go down or C must go up.

For example, suppose the benefit ratio is initially 40 percent of the average wage. When the support ratio is 3.3, the required contribution rate is 12 percent. This is approximately where we are now in the United States. But when the support ratio falls to 2 due to increased longevity and decreased birthrates, the contribution rate must rise to 20 percent or the benefit ratio must fall to 24 percent, or some combination of the two must occur. That is roughly the direction we are heading in, under the current system. If we reach 2041 without prior reforms, changes approaching this magnitude will become unavoidable.

"Pay-as-you-go systems don't work when the number of elderly grows faster than the number of workers paying to support them."

Parametric changes in the system – raising the contribution rate or the retirement age, reducing the pension, modifying the indexation formula – could keep it solvent as demography changes. However, these changes would be very difficult for workers and retirees to swallow and very difficult for politicians to make. What is more, they would have to be made not once, but repeatedly, as changes continue to occur. For example, every time the trustees of the U.S. Social Security system have assessed its fiscal soundness, they have found that people were living longer than had been expected at the previous assessment and therefore some adjustment to the benefit formula or contribution rate was needed to keep the system solvent. In addition, some demographers expect longevity improvement to accelerate, which would also accelerate the required adjustment for Social Security.

Growth. Pay-as-you-go systems also have negative effects on economic growth:

• High and rising payroll taxes for pensions (exceeding 25 percent of wages in many countries) may increase unemployment;
• Early retirement on actuarially unfair terms (often below the age of 60) reduces the supply of experienced labor;
• Giving benefits to the first generation of retirees and continuing to provide public annuities to future generations may discourage private saving; and
• A large implicit pension debt accumulates, and the need to pay off this debt may crowd out the government's ability to provide other important public goods.

In the United States, the payroll tax is relatively low, the retirement age is relatively high and early retirement is penalized. But many economists consider our national savings rate too low, and we have not used our old age security program to raise it.²

"Pay-as-you-go systems often restrict economic growth by discouraging saving and encouraging early retirement."

Contributions that workers make to pay-as-you-go systems are justified by the promise of pensions when they retire in the future. Unfortunately, no assets back up these promises. Thus, a large unfunded implicit pension debt accumulates – the present value of the benefit promises that have been made to current workers and pensioners. It is an implicit debt because it is not written down or legally binding, but it is a debt in the sense that ethically and politically most of it must be paid. In every industrialized country, this implicit pension debt exceeds the explicit debt (bonds). Usually the unfunded pension debt is more than 100 per cent of GDP and in some cases more than 200 per cent. Figure III shows some examples. Fortunately, the United States is at the low end on this score because our benefits are quite modest and our population is still quite young. But as the baby bulge moves through the system, our pension debt will increase. Most countries would be horrified by an explicit debt that exceeds 100 percent of GDP. But these same countries act as though they are unaware that they are building a much larger implicit debt. Future generations will have to pay off this debt.

"In all developed countries, implicit public pension debts dwarf the traditional public debt — often they are twice as large."

Equity. Empirical evidence has cast doubt on the equity of many traditional systems. In many countries, rich people gain at the expense of the poor. Rich people live longer, collect benefits for more years and, because of steeper age-earnings profiles, they often collect higher pension benefits per dollar contributed. Although the United States has a relatively progressive benefit formula, this tends to be offset by the greater life expectancy of the rich – a disparity in life span that has been increasing. Under the defined benefit formula in effect in most countries, workers who retire early receive benefits that are subsidized by taxes paid by those who work longer. This is not as big a problem in the U.S., but our system, too, produces fewer goods and services when workers retire before the age of 65 – as has been increasingly the case. The U.S. has other surprising and questionable winners and losers. For example, nonworking spouses of high-earning breadwinners are big winners, taking out of the system much more than they ever put in. On the other hand, unmarried heads of households and dual career families are big losers.³

"More than 80 million workers worldwide have the opportunity to invest in private retirement accounts as part of their social security systems."

The biggest redistribution in the pay-as-you-go systems of virtually all countries stems from the payment of generous benefits to the first generation to retire under these systems. These retirees contributed small amounts for only part of their working lives and received relatively generous benefits for their entire retirement. The money paid to them prevented the buildup of funds in the accounts of younger workers, and left us with a large unfunded debt. While most of us might be glad to pay these redistributions to low earners who suffered through the depression and war, high earners benefited as well. In fact, the high earners benefited more. Today's young and middle-aged workers and future generations of workers, including many low earners who will get a low return on their Social Security contributions, will inherit this debt.⁴

[page]

To avoid these dangers in the future, 20 countries (from Latin America, Europe and the Asia-Pacific region) have reformed by incorporating prefunded, privately managed retirement accounts into their mandatory Social Security systems. [See Figures IV and V.] In most of these countries, somewhere between 6 percent and 12 percent of payroll is contributed to the funded plan. We can learn from the methods they have developed for managing the funds and handling the concomitant issues of risk, cost and equity.

The experience of the last 20 years shows us that structural pension reform is difficult but possible – even in democracies and welfare states. These structural Social Security reforms have certain commonalities, but also many differences – as a result of different initial conditions and political economies. The most important commonalities are:

• A partial shift from pay-as-you-go to prefunding, with private management of the funds;
• A partial shift from defined benefit plans to defined contribution plans;
• Separate arrangements or "pillars" for the poverty-prevention part of the old age system (the public pillar) and the retirement savings part (the private pillar); and
• Government regulation over the private pillar.

The term "privatization" has been applied to these Social Security reforms, but it would be more accurate to call them public-private partnerships, since each sector plays an important role – the public sector regulating and providing a social safety net, the private sector investing the funds.

The reforming countries also exhibit important differences, chief among them the nature and size of the public pillar and the question of who chooses the investment managers in the private pillar. This paper focuses on funding and how the funds have been managed.

Why Prefunding? Most analysts now agree that some prefunding is desirable for the sustainability of a pension system and for the broader economy. It is good for the finances of the system because a given contribution rate will support a higher expected benefit rate under prefunding than under pay-as-you-go.

• The rate of return in a pay-as-you-go system is (approximately) the rate of wage growth plus the rate of population growth, both of which raise the payroll tax base that finances benefits. With a stable population and a wage growth rate of, say 2 percent, this yields a 2 percent rate of return to contributions.
• In contrast, the rate of return in a funded system is the return on investments, which historically has been more than 5 percent in real terms for a mixed portfolio of stocks and bonds.⁵

"Funded systems give workers higher rates of return on contributions than pay-as-you-go systems."

This means the individual will get a larger pension from his contribution to a funded system. Thus if the support ratio is 2 and the target benefit rate is 40 percent of the average wage, a pay-as-you-go system requires a 20 percent contribution rate, but a funded system in which the funds earn a 5 percent real return requires only a 7 to 8 percent contribution rate. So a funded system shrinks the burden on younger workers and avoids the peak tax rates that would be required under a pay-as-you-go system as the population ages.

Prefunding is also good for the broader economy because it can help to build and mobilize long-term national saving. Saving, in turn, facilitates capital accumulation. Moreover, empirical analysis suggests that saving that is committed for the long term, as is the case for retirement savings, is especially productive.⁶ It increases the size of the GDP pie that will later be available for people to consume. If savings are suboptimal to begin with, due to public or private myopia or a tax wedge between social and private returns, the increase in future consumption is valued more than present consumption forgone by the savings. Everyone can then be made better off. Thus, mandatory Social Security saving can be an important ingredient of a long-run strategy for increasing productivity and output, enabling the standard of living to remain high when the ratio of retirees to workers increases. Long-term national saving can of course be increased in other ways, such as paying down the national debt, but retirement accounts is one of the most effective ways.

"Mandatory retirement saving can be used to increase a country's productivity and output."

However, prefunding mandatory retirement accounts will increase national saving only if it does not crowd out other private savings or increase public dissaving. Regarding the first point: if workers believe that a prefunded system is more credible than a pay-as-you-go system, they may save less on a voluntary basis for their own old age (or borrow more for current consumption), thereby offsetting some of the increased mandatory saving. In the United States, since few people save voluntarily, this offset is likely to be small. Regarding the second point: if the buildup of pension reserves relaxes fiscal discipline or if the government finances the transition solely through issuing additional bonds, this will mean that increased public deficits have absorbed the increased personal saving. The choice of debt finance versus other means of financing the transition will be discussed further below.

Why Defined Contribution? In a defined contribution plan, the worker contributes according to a fixed schedule (e.g., 3 percent of wages per year) and these contributions, together with the investment returns earned, eventually turn into retirement income. This creates a very close link between contributions and benefits. It is designed to discourage evasion and has few labor disincentives.

Most important is the impact on early retirement. Generally, defined contribution accounts are turned into annuities or other forms of gradual withdrawals when the worker leaves the labor market. For those who retire early, the annual annuity or withdrawal will be relatively small, because it must cover many years. For those who work longer, the annual benefit will automatically be larger, if it is adjusted on actuarially fair terms. Those who retire early bear the cost of their early retirement in the form of lower accumulations and benefits rather than passing the costs on to others and undermining the financial viability of the scheme, as occurs in most defined benefit plans. This provides an incentive for continued work, which increases the nation's labor force and productive capacity. The conversion of savings into annuities or other withdrawals on actuarially fair terms, which is characteristic of funded defined contribution plans, is thus good for the economy. It is also good for the sustainability of the Social Security system, particularly as longevity increases, since it provides an automatic mechanism for encouraging workers to raise their retirement age – but each individual makes his or her own choice. This avoids the need for a series of collective decisions about retirement age, which are difficult for politicians to make.

"Returns to privately managed funds outpace returns to centrally controlled accounts."

Why Private Management of the Funds? While many analysts now agree on the benefits of funding, a big issue in the United States concerns how these funds will be managed. Who should choose the investment managers and how much should we constrain worker choice? Practically all countries that have adopted structural reforms in recent years have chosen private control over the funds – believing that this maximizes the likelihood that economic rather than political considerations will determine the investment strategy and will produce the best allocation of capital and the highest return on savings. In some Organization for Economic Cooperation and Development (OECD) countries (e.g., Switzerland, Australia, Denmark and the Netherlands), for historical reasons, employer and/or union representatives choose the investment managers for employees of an entire company or for all the members of an occupational group. In Latin America and Eastern Europe (e.g., Argentina, Chile, Mexico, Hungary and Poland), workers choose the investment managers for their own individual accounts in the retail market. And more recently an "institutional market" approach has been adopted in countries such as Bolivia and Sweden. Under this approach, fees are negotiated centrally in order to gain for numerous small accounts the low fees that large institutional investors usually pay. In all these cases, the investment managers are private companies, ultimately chosen by private rather than public sector agents.⁷

"Private control means economic, rather than political, factors will control investment decisions."

These countries have all been influenced by empirical data that show publicly managed pension reserves around the world earn low returns, far below the rate of return on bank deposits or the growth of per capita income, and frequently lose their principal as well. [See Figures VI and VII.] This is largely because public managers have been required to invest in low-interest government securities, make loans to failing state enterprises or invest in other politically motivated investments. Moreover, with publicly managed funds politicians are subject to pressures to raise benefit payments rather than to invest (for example, this happened in the early years of the U.S. Social Security system). The hidden and exclusive access to these funds also makes it easier for governments to run larger deficits or to spend more wastefully than they could if they had to rely on a more accountable source of funds. Some economists believe this happened in the United States beginning in the 1980s, as the Social Security trust fund built up and was used to finance the growing public deficit in a nontransparent way.⁸

"Government-owned, centrally controlled pension funds make it easier for governments to run larger deficits or spend more wastefully."

In contrast, as can be seen in Figure VIII, competitively managed funded pension plans are more likely to be invested in a mixture of public and corporate bonds, equities and real estate, thereby earning a higher rate of return. They reap the benefits of investment diversification, including international diversification, which enables them to increase their yield and reduce their risk. They build constituencies that help them resist political manipulation. They spur financial market development by creating a demand for new financial instruments and institutions, especially important in middle-income countries. In Chile, the systemic reform implemented 20 years ago has made financial markets more liquid as the number of traded shares on the stock market and their turnover increased; has created demand for the equities of newly privatized state enterprises; has encouraged the emergence of information disclosure and credit-rating institutions; has expanded the variety of financial instruments including indexed annuities, mortgage and corporate bonds; and has improved asset pricing. These developments have played a particularly important role in explaining Chile's rapid growth rate since it started its new Social Security system.⁹

"Privately managed pension funds earn a higher rate-of-return than public funds without incurring additional risk because their holdings are more diversified."

Some critics of private management have argued that the United States is different from the African and Latin American countries that misallocated and dissipated their publicly managed pension funds. They argue that we should prefund, and the funds should be invested in securities markets, but that the government should manage the funds and choose the investment strategy, in order to benefit from scale economies and financial expertise. Canada and Ireland are now experimenting with a centrally managed trust fund that will be invested in a diversified portfolio, hoping that they have built in legal safeguards that will protect them from political manipulation and will make them the exception to the rule. It is too soon to tell whether they will succeed.

The United States does indeed have governance procedures and trustee laws that would prevent gross abuses of a publicly held trust fund. However, we also have pressure groups, intensive lobbying and campaign contributions that influence policy formation. If funds were publicly managed, we could imagine these pressures being brought to bear on issues such as: Which companies, industries and indexes should be the focal points for investment? Which products (e.g., tobacco, abortion pills) should be prohibited for investment? Should the funds be used to prop up the stock market when it is falling or to temper it when it is soaring? Should antitrust actions be started and profit-reducing regulations be imposed on companies in which the public fund has major investments? Should the government as regulator provide insider warning to the government as investment manager, before starting regulatory actions that might hurt stock prices?

"If public funds are invested in the private markets, it is reasonable to expect that political considerations will interfere with prudent investment choices."

Such pressures on fund allocations will not maximize the financial returns to the system or the productivity of capital in the economy. They could also lead to wasteful spending of resources to influence these allocations. Public control of Social Security funds would concentrate a lot of market power in one large investor and could lead the government to play a large and largely unhealthy role in the governance of corporations in which it has invested. Moreover, it would be extremely tempting for the government to use these funds as a hidden source of deficit finance, enabling it to spend more and tax less in the short run. But this means that, in the long run, the funds would not be saved – the payment of pensions would remain an increasing liability of future taxpayers. For all these reasons, private decentralized management of the funds seems to be in the best interest of the pension system, the broader economy and the general polity. In the U.S. context, this would imply individual accounts.

[page]

While prefunding, with private control over the funds, may be beneficial for the sustainability of the system and for economic growth, most analysts agree this should be done in a way that keeps administrative costs low, avoids undue financial and political risk and distributes its costs and benefits equitably.

Are Administrative Costs Too High? The biggest criticism of individual account systems concerns the high administrative cost and fees that have developed in some countries using the Latin American (retail market, worker choice) model. The advantage of private over public investment is the likelihood that it will produce a better allocation of capital, earn higher returns for the fund and spur higher economic growth for the economy. However, if decentralized systems charge high administrative fees, they reduce net returns and pensions. In Chile and most other Latin American countries, fees are front-loaded. This means that workers pay a one-time fee on new contributions rather than an annual fee based on assets. Specifically, this one-time fee is about 2 percent of wages, or 15 to 25 percent of new contributions, in virtually all cases. Ultimately, this makes the pensions 15 to 25 percent lower than they would have been in the absence of these costs. While expenses related to investments and record keeping are inevitable in any system of retirement savings, almost half of these total costs are due to marketing expenses that may be avoidable.

"Experiences in other countries show that personal accounts can be structured in a way that keeps administrative fees low."

To understand the impact of one-time fees on net returns, it is helpful to convert these one-time charges on contributions into their equivalents in terms of annual charges on assets, which is the way most mutual fund charges are assessed in the United States. Obviously, for accounts that have small accumulated assets (young workers with few years of contributions), this fee will be high relative to assets. However, for accounts that have built up substantial assets over the years, the fee will be small relative to assets. Simulations show that if the current fee schedule is maintained, the average Chilean worker who contributes for 40 years will pay the equivalent of less than 1 percent of assets per year.

To put these numbers into perspective, these charges are somewhat less than the fees and expenses found in the average U.S. mutual fund in which millions of individuals invest on a voluntary basis. So the perceived benefits far exceed the costs for many people. Moreover, they are not excessive in comparison to less expensive systems that produce much lower gross and net returns (e.g., publicly managed reserves in Singapore and the U.S. Social Security trust fund). Nevertheless, these costs and fees are a source of concern in mandatory systems where everyone is forced to participate and pay. And they are a particular problem in the early years of a system, when accounts are small.

Investing: The Retail Market, Worker-Choice Method. Most Latin American and Eastern European countries use the retail market, in which investment managers enter the industry and try to attract individual worker-savers. This is much like the mutual fund industry in the United States. Some analysts believe that administrative costs would be lower under a group plan with decisions made by an employer or union. Such group plans may be better positioned to benefit from economies of scale, greater financial expertise and lower marketing costs.¹⁰

"The average Chilean worker who contributes to a personal account over a 40 year period will pay less than 1 percent of assets per year in fees under the current fee structure."

Investing: The Group Plan Model. Most OECD countries use group plans – with an employer and/or union choosing the investment manager – and this often produces lower administrative costs. While these plans used to be defined benefit plans, under the new system employers have been shifting to defined contribution plans to avoid bearing the defined benefit risk. When employers or union representatives make the investment decisions while workers bear the risk, as in a defined contribution plan, such arrangements can open the door to financial abuse and principal-agent problems. For example, employers might choose investment managers or strategies that benefit them even if this results in lower returns for their workers. For this reason, we would expect to see greater worker choice – a shift toward the retail approach – in these OECD countries.

Investing: The Institutional Model. A third alternative draws on the experience of the institutional market, in which large investors (company defined benefit plans, foundations, endowments) face much lower rates. The reformed systems in Bolivia and Sweden represent attempts to use or mimic the institutional market to achieve lower fees in their mandatory systems by aggregating numerous small accounts into large money blocs and negotiating a group rate. This alternative may be especially desirable in (1) small countries whose markets cannot support many pension companies efficiently due to scale economies, (2) countries with undeveloped financial markets that want to attract investment expertise and minimize startup costs, and (3) countries with low contribution rates.

In Bolivia, an international competitive auction process was used to select two pension funds to run its mandatory private pillar. Although initially assigned, workers will soon be given the choice between them. This competitive bidding process has resulted in much lower costs than in other Latin American countries.

In Sweden, the pension authorities established a maximum fee schedule for managers of mandatory retirement accounts. Workers are permitted to choose among the large number of mutual funds, but the money is moved in large blocs, records are kept centrally, and funds do not even know the names of the individuals who invest with them – an attempt to avoid sales commissions. The allowable fees, again, are much lower than those charged in Latin America or Eastern Europe or by the mutual fund industry in Western Europe or the United States.¹¹

Along similar lines, the voluntary Thrift Saving Plan for U.S. federal employees chooses its asset managers on the basis of competitive bidding. Workers have a choice of five (recently increased from three) portfolios, specializing in stocks, bonds and money market instruments, respectively. All these portfolios are benchmarked to broad market indexes and use passive investing, which is much cheaper than active management. Total cost is 11 basis points (0.11 percent of assets) per year.

The sidebar describes methods proposed by the President's Commission to Strengthen Social Security to keep administrative costs low in the United States.

[page]

Risk and uncertainty are inevitable in all old age security plans, given the long time horizons that they cover. One of the concerns most frequently voiced concerning individual accounts deals with financial market risk – the possibility that investments will have a poor outcome for some workers. However, it should be borne in mind that there are risks of a different kind associated with public pay-as-you-go defined benefit plans.

In a pay-as-you-go system, as conditions change in unexpected ways, a collective decision is made about how to adapt to those changes. For example, suppose the old live longer and the dependency rate rises. Who bears the resulting costs for the system? We don't know the answer until the situation arises. If, on the one hand, retirees are politically strong, as they are in the United States today, they will be able to retain their benefits for many years, while workers pay more. If, on the other hand, workers are politically strong, benefits will be cut in order to keep payroll taxes stable. It is also possible, in the case of a major political upheaval (as in the former Soviet Union, the former Yugoslavia or China), that the rules on which workers counted will change altogether. In all these instances, workers and retirees face the risk of changing conditions that will greatly affect their current or future well-being. But the risk is political risk.

"Stock market risk can be mitigated over the long-term by diversification."

In contrast, in a private funded system, workers and retirees face financial market risk – the market may not perform as well as we hope or believe it will. Generally, the investment contract spells out how benefits will be affected, so the locus of risk is better defined before conditions change than in a public defined benefit scheme. A variety of risk-sharing options are available. They range from arrangements where the entire loss is borne by the worker investors (as in stock market mutual funds) to arrangements where the entire loss is borne by a guaranteeing institution (as when insurance companies guarantee a fixed rate annuity) and to complex risk-sharing contracts that afford workers some downside protection if they agree to forgo some potential upside gain (as in "collars" or variable annuities with floors). In general, the more risk the worker- or retiree-investor bears, the greater the return he can expect and the greater the loss he can incur.

Risk cannot be avoided, but it can be reduced and the rules for sharing it can change. What are some reasonable principles that might be considered for mitigating risk and sharing it efficiently?

Reducing Risk through Diversification. As a starting point, given the great uncertainty faced by old age security programs stemming from their long time horizons, the best protection against risk is broad diversification of income sources. This is achieved by a system that has publicly and privately managed components: financing based both on wages and returns to capital, on both mandatory taxes and voluntary savings, and investments both in stocks and bonds. In this sense, adding a private investment component to a system that is exclusively public and pay-as-you-go probably reduces risk overall by diversifying the sources of total retirement income. If the public source fails due to political risk, the private source may remain, and vice versa.

Within the funded pillar, financial risk can be further reduced:

• By regulations that require benchmarking to a broad market index rather than concentrating on a narrow sector;
• By making periodic contributions rather than one lump sum (dollar-cost averaging);
• By holding for long time-horizons and withdrawing gradually, rather than trying to time the market; and
• By investing in a broad range of securities, both in domestic and international markets, to minimize country-specific risk.

"Most reforming countries have restricted the portfolio choices to reduce risk."

History tells us that attempts to avoid market risk through fixed income securities imply a low nominal return and therefore introduce the risk that pensions will not meet replacement rate targets. Investing gradually, holding stocks for the long term and withdrawing gradually protects the equity premium against short-term volatility by diversifying through time. While the future may differ from the past, holding a diversified stock portfolio for 20 years or more has always resulted in a net gain that exceeds the return from a pure bond portfolio.¹² Diversifying across sources of retirement income, types and locations of financial instruments, and dates of investment and withdrawal seems to be the most effective way to reduce risk in a very uncertain situation. [ See sidebar ]

Risk Control Strategies in Reforming Countries: Diversification and Guarantees. What have reforming countries actually done? Frequently they have restricted the portfolios in which pension funds may invest in order to rule out highly volatile and concentrated portfolios. Latin American and most continental European countries include explicit portfolio limits, while the Anglo-Saxon countries typically rely on the more discretionary "prudent man rule."¹³ The latter approach seems to perform better than the former, but it may be difficult to implement in countries without well-developed rules of law and prudent man precedents.

"Most reform plans in the United States would allow workers to choose from a limited number of broadly-diversified portfolios."

To shift part of the remaining risk away from the worker, many countries require guarantees of absolute or relative rates of return by the pension funds and/or include a minimum pension guarantee by the government. Switzerland requires a minimum return of 4 percent nominal over the workers' tenure with a given employer. This constraint may, however, lead to overly conservative investments, as rates of inflation and nominal interest rates fall. Chile penalizes funds that earn a return that is 50 percent lower or 2 percentage points lower than the industry average, whichever comes first. (The averaging period, initially one year, was increased to three years, and further easing is now under consideration.¹⁴) This rule has been accused of leading to investment herding among pension funds, as each fund tries to look very much like the others. Rather than having a choice of different points on the risk-return frontier, stemming from differing asset allocations – as would be the case in a well-functioning financial market – workers have the much less meaningful choice among companies that provide the same asset allocation and risk-return mix.

"In many Latin American countries, the government provides a minimum pension guarantee, designed to keep retirees out of poverty."

In many Latin American countries, the government provides a minimum pension guarantee that retirees will not fall into poverty. In Mexico, current workers are permitted to return to the old pay-as-you-go system upon retirement if this yields a better pension. One might anticipate that this would lead to a moral hazard problem – workers buying overly risky portfolios, knowing they will benefit from the prospect of gains but are protected from downside loss. The Mexican authorities have avoided this moral hazard problem by greatly limiting the funds' choice of investment strategies: At least 65 percent of all assets must be invested in government bonds (as of 2000, the funds were still 99 percent in bonds), and international investments are proscribed. Since workers have no real choice of portfolios, moral hazard is avoided, but the flow of funds through the pension funds to the financial market and the private sector is also avoided. Thus, guarantees reduce risk for the worker but they also introduce new costs, potential moral hazard problems and investment restrictions.

[page]

A third key issue concerns the distribution of the costs and benefits of a Social Security system that includes individual accounts. Different systems affect distribution of income across generations, between workers and retirees at a given point in time, between full-career versus part-career workers, between the two genders, and among various socioeconomic groups. For purposes of this discussion, we focus on only two types: redistributions to low earners to keep them out of poverty and "perverse" redistributions to high earners.

What Have Countries Done? As already discussed, pay-as-you-go systems are not as progressive as we might like to believe – features such as longer lifetimes and steeper age-earnings profiles often make high earners the chief gainers. Even in the United States, which has a supposedly progressive benefit structure, recent empirical studies suggest that lifetime redistributions do not benefit the poor, on the average.¹⁵ Because many traditional systems are both inefficient and inequitable, an opportunity exists to improve both outcomes. However, it is still an open question whether or not the reforms have succeeded in improving equity. Closer examination suggests that the devil is in the details and some of the results are surprising.

"Pay-as-you-go systems often benefit high-earners more than low-earners — just the opposite of what many expect."

For example, workers are eligible for Chile's minimum pension guarantee (about 25 percent of the average wage) after 20 years of contributions, meaning that the government tops up the benefits of these workers to the guaranteed point if their own accumulation does not suffice. The main beneficiaries will be low earners who worked only 20 years – disproportionately females with limited labor market attachment – while workers who remain in the formal sector for a full career are unlikely to receive this subsidy. This benefit is well targeted toward the poor but may encourage them to withdraw from the formal labor market after they reach the 20-year point.

In contrast, Argentina pays all workers with at least 30 years of contributions a flat benefit of about 25 percent of the average wage (plus an additional 1 percent for every year above 30 up to 45 years). This benefit structure encourages and rewards continued formal labor market participation. The main recipients are workers who have spent most of their adult lives in the formal labor sector. In sharp contrast to Chile, women and other low-income transient workers are unlikely to qualify.¹⁶

Switzerland's public pillar is earnings-related and hence appears less redistributive than Argentina's. However, the benefit schedule is very compressed and the payroll tax that finances it is levied on all earnings (that is, there is no ceiling on taxable earnings), which makes it quite redistributive toward the poor. In fact, low earners get such a high replacement rate from the public pillar that they are not even required to contribute toward the private pillar, thereby increasing their take-home pay as well. [See the sidebar for ideas on ensuring equity in the United States.]

[page]

If a country with a pay-as-you-go pension system switches to a multipillar system that includes a funded component, some of the contribution usually is shifted to the funded pillar. For example, the proposals of the Social Security Commission would shift (or in the parlance of Washington, "carve out") amounts of 2 to 4 percentage points from the total payroll tax into the funded pillar, the individual accounts. This creates a temporary financing gap between the remaining pay-as-you-go revenues and the revenues needed to cover the current obligations of the old system. Some other revenue source must be found to cover this short-run financing gap, in addition to the long-run preexisting financing gap of the old system. This short-run gap due to the reallocation of the payroll tax to the accounts is known as the transition cost problem.

Countries that finance their funded pillar by adding an extra contribution rather than diverting money that was originally slated for the pay-as-you-go pillar (an "add-on" as opposed to a "carve-out") do not face this problem. For example, most OECD countries with multipillar systems started with modest pay-as-you-go pillars and financed their funded pillars by mandating additional contributions, thereby avoiding the transition financing gap. If the United States chose a small add-on instead of a carve-out, we too would avoid the transition financing problem. An add-on also has the advantage of helping to increase national saving. For these reasons, one of the Commission's plans included an add-on. The downside to this strategy is that workers who participate have to pay more.

"Transitions costs arise from the need to meet obligations that already exist."

All Latin American and Eastern European countries have used the carve-out approach and therefore have faced the transition cost problem. How did they finance the transition? Which of these methods would be most applicable to the United States? Because of the fungibility of money, it is difficult to answer these questions precisely. That is, if government debt and taxes both rise, it is difficult to know how much of the increased debt, versus taxes, was used to finance the pension transition. Knowing that would require knowing the counterfactual – exactly what would have happened otherwise – and unfortunately we do not. Compounding this problem is the fact that, even without a diversion of contributions to the funded pillar, all these systems were or would soon be in financial distress because their future obligations exceed their incoming revenues under the old system. The term "transition costs" properly applies only to the additional gap created by the carve-out.

While we cannot give precise numbers, we can describe more generally the five strategies that countries have used:

• Making the carve-out relatively small and keeping some workers in the old system so that most of the contribution continues flowing into the pay-as-you-go pillar;
• Downsizing the benefit obligations of the pay-as-you-go pillar, particularly for young workers, expecting the growth of the funded pillar to restore these benefits;
• Applying state-owned assets or budgetary surpluses to offset the pension debt;
• Borrowing temporarily to spread the burden of transition costs across generations; and
• Using general revenues (higher taxes, lower government expenditures) to repay this loan over time.

Each of these methods and its applicability to the United States is discussed below. Each has different effects on income distribution and national saving which must be evaluated. The important thing to remember is that transition costs arise from the need to meet obligations that already exist. The transition costs diminish as these old obligations are paid off. In a viable program, in the long run the drop in new pay-as-you-go obligations exceeds the drop in revenues, so "transition costs" become "transition gains." Moreover, if the new system enhances economic growth by increasing long-term saving, labor supply and productivity, this will generate additional resources for the treasury that can be used to finance the transition.

"Over the long run, transition costs become transition gains."

Making the Carve-out Small. The transition financing gap will be reduced if some part of the new system remains pay-as-you-go, so contributions continue flowing into the pay-as-you-go pillar. This has been accomplished in several ways:

1. Keeping a sizable public pillar and instituting a smaller private pillar. My research shows that countries with a large implicit pension debt, such as Sweden, Hungary and Uruguay, tended to keep a large public pillar, because they felt that otherwise they could not cover their transition costs. In contrast, such countries as Chile, Bolivia and Kazakhstan, which started with a relatively small pension debt, resorted to a small public pillar – a minimum pension guarantee, received only by lifetime low earners.¹⁷ The continued inflow of funds to a large pay-as-you-go pillar reduces the transition financing gap in the short run. But if benefits are too generous (actuarially unsound), the reform will not be sustainable in the long run.
2. Excluding some workers. Exclusions from the new system may include such people as the military, the police or older workers (as in Chile) or may make the second pillar mandatory only for certain groups such as high earners (as in Uruguay).
3. Making the switch voluntary for current workers. Most reforming countries have followed this approach, making the multipillar system mandatory for new workers but allowing current workers to stay in the old system if they wish. Usually workers over age 45 choose to stay in the old system while most younger workers switch. The former group continues contributing to the pay-as-you-go pillar, thereby reducing the financing gap, while the latter group partially withdraws with the expectation that the individual accounts will build up by the time they retire. One advantage of a voluntary switch is that it mitigates opposition to reform from groups most anxious to stay in the old system and permits a lower value to be placed on past service credits for those who switch. By choosing the minimum terms that are needed to convince the desired number of workers to switch, a government can substantially downsize its recognized debt and transition costs (as in Hungary). Obviously, the higher the expected rate of return on the individual accounts, the lower the compensation needed to induce workers to switch. In effect, the transition can be partially self-financed by building a strong second pillar.

Which of these methods would work in the United States? Since our nation has a relatively small implicit pension debt and financing gap (compared with other industrial countries), we could finance a largely funded privatized system if we chose to do so. That is, we could carve out half or more of the contribution (as in several recent proposals by Syl Scheiber, Martin Feldstein and Lawrence Kotlikoff) for the new funded pillar. In the Feldstein and Kotlikoff proposals, the individual account by itself is large enough to maintain currently scheduled replacement rates. In the Scheiber proposals, a 5 percent contribution to an individual account is accompanied by a smaller flat benefit in the pay-as-you-go pillar. Together, the two parts maintain currently scheduled replacement rates and progressivity. However, most proposals visualize keeping a substantial pay-as-you-go pillar, diverting only 2 or 3 percentage points, or 20 percent to 30 percent of the total contribution, to the funded pillar. In following this course we would gain the advantage of keeping the transition financing gap small.¹⁸

The Commission's proposal reduced this gap still further by making the switch voluntary. A voluntary switch is difficult to implement under the complex benefit formula in the U.S. system. Careful measures must be taken, or else high earners might be the largest group to opt out of the pay-as-you-go pillar, thereby withdrawing part of their contributions that were used to subsidize the benefits of low earners. Low earners would be left in the pay-as-you-go system, exacerbating its financing problem. The switching terms must be carefully devised to maintain fiscal balance and overall system progressivity. The Commission recommended that high earners should be more limited than low earners in the amounts of payroll tax they can reallocate, partially mitigating this problem. Also, each dollar reallocated by high earners would result in a larger percentage cut in defined benefits than for low earners. Workers over the age of 55 would continue making their full contribution to the old system and their future benefits would not be changed.

"Future Social Security or budget surpluses can be harnessed to cover part of the transition cost."

Downsizing Benefit Obligations. Before or in the course of making the transition, most countries have reformed their old systems by downsizing benefits, raising the retirement age, raising penalties for early retirement, tightening eligibility for disability benefits, and changing the indexation of initial benefits to price rather than wage indexation after retirement. Chile, Argentina, Uruguay, Hungary and Poland followed this strategy, which may be virtually indispensable to a good pension reform – especially in countries that start out with bloated benefits. (Of course, benefit cuts would be necessary in these countries whether or not they undertook structural reform.) These measures cut the transition cost problem, but only to a limited extent because they are generally phased in very gradually. Some, but not all, of these benefit cuts are typically made up through the growth of the funded pillar. That is, typically in Latin America and Eastern Europe, total benefits from the combination of both pillars in the new system were somewhat reduced, but benefits stemming from the pay-as-you-go pillar alone were cut much further. [See the sidebar for a discussion of U.S. Social Security Benefits under Reform.]

Using Existing Assets to Pay Off the Pension Debt and Cover Transition Costs. In some reforming countries, such as Peru and Poland, where public enterprises are being privatized, part of the proceeds have been used to pay off the pension debt – applying long-term assets against long-term liabilities. China is trying to sell off its shares of state-owned enterprises to generate resources that will finance its transition to a funded individual account system. This is not a potential revenue source in the United States. More relevant for our purposes is the use of treasury surpluses or temporary surpluses in the existing Social Security system.

While the Latin systems generally did not have a surplus in their old Social Security systems, the United States does, in the Social Security trust fund. The trust fund is not large, it will not last long and it consists exclusively of special non-tradeable government bonds. But until the trust fund is exhausted, redeeming these bonds is one of several methods that can be used to generate cash to pay off old obligations.

The Latin countries (aside from Chile) also did not have surpluses in their general treasuries. Chile is said to have built up a surplus in its public treasury in preparation for financing its pension reform. In contrast, the U.S. budget surplus has disappeared; but it may reappear in the future, still in time to help. If a surplus exists that otherwise would be used to increase government spending or cut taxes, then its use to allow the buildup of retirement accounts enhances national saving and labor productivity.

Issuing General Treasury Debt to Cover the Remaining Cash Gap in the Short Run. Because of the fungibility of money, we do not know to what extent the transition in other countries has been financed from benefit cuts plus asset reallocations, but 30 percent might be a good estimate in many cases. The remainder is usually financed by government debt, with the intent to gradually repay this out of taxes and system savings that continue after the old obligations have passed. The relevant proportion for the U.S. could only be determined by crunching the numbers for many alternatives.

"Government borrowing to find the transition debt is merely an exchange of implicit public debt for explicit public debt."

Government borrowing has increased in the early years of the reform in almost every Latin American and Eastern European country. Indeed, some use of temporary debt finance is almost inevitable so that a heavy double burden is not imposed on the transition generation of workers. Temporary borrowing with gradual repayment allows policy makers to determine how the burden of the debt should be distributed among cohorts. Since young and future workers will benefit most from the reform, by receiving larger pensions or paying lower contributions than they would have otherwise, it is appropriate that they should also pay part of the cost. This is accomplished by borrowing to cover part of the transition cost and repaying the loan later on.

It is crucial to realize that total public debt is not increased by this financing arrangement. The pension debt exists right now, in every pay-as-you-go system. Rather, borrowing temporarily as part of a transition to a funded system is simply an exchange of hidden implicit debt for more observable explicit debt. In fact, in the course of the reform the total pension obligation of the government has been reduced in almost every country. This also holds for the plans recommended by the Commission in the United States. The fact that the remaining debt becomes more explicit increases the likelihood that pressures will be brought to bear to pay it off. Using special issue transition bonds with scheduled retirement dates could enhance such pressures. In countries that have already reformed, some of the new bonds were sold to the pension funds in the new second pillar. Government debt and bank deposits have been the largest initial investments of practically all new pension funds in Latin America and Eastern Europe, although they are gradually moving toward much more diversified portfolios.

Paying Off the Debt in the Long Run. However, if one object of the reform is to increase national saving, then the painful fact is that someone's consumption must be cut, at least temporarily, relative to what it would have been otherwise; and pure debt finance will not accomplish that. Eventually, taxes must be raised or benefits and other government expenditures cut enough to pay off the debt. The slower the payoff, the lower the required tax rate per year and the more politically acceptable may be the reforms, but this also delays the timing of increased national saving for productive investment. Indeed, if the transition is largely debt-financed and if the debt is not paid off in a timely way, the benefits of "pension reform" are substantially mitigated.

"If part of the U.S. Social Security system were converted to a private system, as recommended by the Commission, the transition cost could be financed with a small temporary increase in the payroll tax rate or base."

Each financing method has its downside. Cuts in expected defined benefits hurt workers as they retire, although this is largely compensated by their purchase of private annuities from the personal accounts. Reducing other government expenditures may be difficult to achieve, and some government services that enhance labor skills and capital, hence productivity and economic growth, might be lost. If taxes are increased during the transitional period, we must decide which taxes. The choice is between increasing payroll tax rates, increasing the payroll tax base or turning to general revenues. A broader tax base would allow a lower payroll tax rate to achieve any given target revenue. It would be less distortionary to labor markets and more growth-enhancing. It would be more progressive, which many people would interpret as more equitable. But it also might be the most difficult to implement politically. Argentina has imposed a new tax explicitly for this purpose, but this action is rare.

It has been estimated that if half the current pay-as-you-go system were converted to a funded direct contribution system, the financing gap could be paid off by a payroll tax rate of about 1.5 percent or a consumption tax of 1 percent for about 70 years in the U.S.¹⁹ A smaller carve-out, as proposed by the Commission, could be paid off with a smaller tax, in fewer years. If the new pension system increases economic growth, the growth premium would allow the transition costs to be paid off faster, while still leaving a consumable surplus for workers and retirees in the long run.

[page]

Twenty countries in Latin America, Europe and the Asia-Pacific region have structurally reformed their Social Security systems over the past 20 years to make the systems more sustainable, equitable and growth-enhancing. At least another 10 countries are considering such reforms. They have tried to achieve these beneficial effects by increased prefunding, with the investment managers chosen on a competitive basis by workers, unions and/or employers. They have tried to insulate the system from political pressures that might emphasize short-run gains over long-run stability. The funded private arrangement has been accompanied by a publicly managed social safety net, for risk diversification and mitigation as well as redistribution. While the reformed systems in most of these countries are still too new to evaluate, Chile's 20-year-old pension reform appears to have made a major contribution to that nation's high economic growth rate.

What does all of this imply for the forthcoming Social Security debate in the United States? Despite the very real problems in our system, it is healthier than those in many other countries that have not yet reformed their systems. Our benefit rate is modest, our retirement age realistic and rising, and actuarial penalties discourage early retirement. Yet we will have to do something to maintain the solvency of the system. We can cut benefits and raise contribution rates, or we can change the basic structure. Since we have to change the system in some way, we should give a lot of thought as to what is the best way. We should choose a fix that lasts, instead of one that will prove insufficient after 10 years. And we should think about how the various proposals will affect the aggregate economy. For example:

• Keeping the payroll tax low can help preserve incentives for employees to work and employers to hire labor;
• Penalties for early retirement and rewards for continued employment can increase the supply of older workers;
• The shift toward individual savings accounts can generate a stock of investable resources; and
• Private management of these funds can help ensure that investment decisions are based on economic rather than political considerations.

"The longer we wait, the more difficult and expensive it will be to enact reforms."

Most analysts on all sides of the political spectrum now believe that some prefunding is desirable – to make the financial balance of Social Security less sensitive to demographic change, to reduce the pension debt we bequeath to our children and grandchildren, and to build national savings for the long term. But once we agree to prefund, the question immediately arises: Who should manage the funds? And here there is greater disagreement. Some argue that a single centrally managed fund backing the current defined benefit formula would be cheaper and better than a shift to individual accounts. However, the experience of many countries suggests that it is difficult to insulate publicly managed funds from political manipulation and this manipulation leads to low rates of return for the economy and the pension funds. Moreover, a Social Security surplus that is under the government's control may increase government's deficit spending rather than saving. This is the basic rationale for private competitive management. Despite much controversy about the administrative costs and fees in such a scheme, analyses of retirement savings plans in the U.S. and abroad suggest that it is possible to carefully design the system to keep these low.

As discussed above, most of the financing for these accounts can come from a carve-out of the existing contribution rate – so long as policy makers are willing to cut benefits from the pay-as-you-go pillar that would remain, below the level that is currently scheduled. Based on the experience of other countries, this cut would be gradual and would not affect current pensioners. The reduced defined benefit would be largely or fully recouped through annuities purchased by the assets in the personal accounts of contributing workers. The fact that the individual accounts are likely to earn a rate of return higher than the return in a pay-as-you-go scheme helps to restore these benefits without the large increase in contribution rate that would otherwise be necessary. Moreover, if the higher rate of return corresponds to increased saving, labor supply and productivity, this enhances real output and enables the real income of both retirees and workers to increase.

This paper has surveyed some of the key implementation issues, including how to contain administrative costs and risk, how to achieve an equitable distribution of benefits and how to cover transition costs. These issues are difficult, but they are not unsolvable. Other countries have solved them. The report of the Commission to Strengthen Social Security demonstrates that we can solve them, too. The longer we wait, the fewer our degrees of freedom and the more benefits will have to be cut or contribution rates raised. The sooner we reform, the larger the accounts baby boomers will build up before they retire, hence the less painful the transition will be. Policy makers will be doing the country a service, as well as acting on behalf of our children and grandchildren, if they reform Social Security now.

NOTE: Nothing written here should be construed as necessarily reflecting the views of the National Center for Policy Analysis or as an attempt to aid or hinder the passage of any bill before Congress.

[page]

1. The first half of this paper draws heavily on an earlier article, Estelle James, "Reforming Social Security in the U.S.: An International Perspective," Business Economics, January 2001.
2. For examples of the extensive literature on the impact of Social Security on retirement age and savings, respectively, see Jonathan Gruber and David Wise, Social Security and Retirement Around the World (Chicago: University of Chicago Press, 1998); and Lawrence Kotlikoff, Jagadeesh Gokhale and John Sabelhaus, "Understanding the Postwar Decline in United States Saving: A Cohort Analysis," in The Brookings Papers on Economic Activity (Washington, D.C.: Brookings Institution, 1996). Also see a summary of this literature in World Bank, Averting the Old Age Crisis: Policies to Protect the Old and Promote Growth (Washington, D.C.: World Bank and Oxford University Press, 1994).
3. Alan L. Gustman and Thomas L. Steinmeier, "How Effective Is Redistribution under the Social Security Benefit Formula?" National Bureau of Economic Research, NBER Working Paper No. 7597, 2000; and Julia Lynn Coronado, Don Fullerton and Thomas Glass. "The Progressivity of Social Security," National Bureau of Economic Research, NBER Working Paper No. 7520, 2000.
4. For a summary of the large literature on intra- and intergenerational redistributions under PAYG in other countries, see Averting the Old Age Crisis. For applicability to the U.S. and distributional effects of proposed reforms, see Lawrence Kotlikoff, "Generational Policy," in Alan Auerbach and Martin Feldstein, eds., The Handbook of Public Economics (New York: North-Holland, forthcoming); Alan Gustman and Thomas Steinmeier, "How Effective Is Redistribution under the Social Security Benefit Formula?" National Bureau of Economic Research, NBER Working Paper No. 7597, 2000; Julia Lynn Coronado, Don Fullerton and Thomas Glass, "Distributional Impacts of Proposed Changes to the Social Security System," in J. Poterba, ed., Tax Policy and the Economy, Vol. 13, 1999, pp. 149-86; Coronado, Fullerton and Glass, "The Progressivity of Social Security," National Bureau of Economic Research, NBER Working Paper No. 7520, 2000; and various papers in Martin Feldstein, ed., The Distributional Effects of Social Security Reform (Chicago: University of Chicago Press, 2000).
5. Many analysts argue that over the past century, in the U.S. and other countries, the equity premium (the difference between the rate of return on stocks and bonds) has exceeded the amount needed to adjust for risk, and some expect this premium to continue for the next century. See Elroy Dimson, Paul Marsh and Mike Staunton, Triumph of the Optimists (Princeton, N.J.: Princeton University Press, 2002).
6. Alberto Musalem and Mario Catalan, "Contractual Savings and Capital Markets Development: An Analysis of OECD Countries' Evidence," World Bank Working Paper, 1999.
7. For further details on the Latin American and Eastern European reforms, see Luis Cerda and Gloria Grandolini, "Mexico: The 1997 Pension Reform. Policy," World Bank, Research Working Paper No. 1933, 1997; A. Chlon, M. Gora and M. Rutkowski, "Shaping Pension Reform in Poland: Security through Diversity," World Bank, Social Protection Discussion Paper No. 9923, 1999; Robert Palacios and Roberto Rocha, "The Hungarian Pension System in Transition," World Bank, Social Protection Discussion Paper, 1997; Monika Queisser, The Second-Generation Pension Reforms in Latin America (Paris: OECD Development Centre Studies, 1998); Klaus Schmidt-Hebbel, "Does Pension Reform Really Spur Productivity, Saving and Growth?" Central Bank of Chile Working Papers 33, Santiago, Chile, 1998; Schmidt-Hebbel, "Chile's Pension Revolution Coming of Age," paper prepared for the DIA Project: "Vergleich alternativer Alterversicherungssysteme in Chile, Deutschland, Groossbritannien, den Niederlanden und den USA," 1999; Schmidt-Hebbel, "Latin America's Pension Revolution: A Review of Approaches and Experience," World Bank ABCDE Conference, Washington, D.C., 1999; Salvador Valdes-Prieto, "Pension Reform in Latin America: Transition and Implementation Problems," Annals of Public and Cooperative Economics, Vol. 69, No. 4, 1998; and Rodrigo Acuña and Augusto Iglesias, "Chile's Pension Reform after 20 Years," World Bank, draft of working paper, 200l. For further details on the OECD reforms, see Hazel Bateman and John Piggott, "Mandatory Occupational Pensions in Australia," Stefan Hepp, "The Swiss Multipillar System," and Paul Johnson, "Pension Reform in the United Kingdom," all from Annals of Public and Cooperative Economics, Vol. 69, No. 4, 1998; and Martin Feldstein, ed., Privatizing Social Security (Chicago: University of Chicago Press, 1998).
8. For an econometric analysis of this issue see Michael Marlow and Mark W. Crain, "The Causal Relationship Between Social Security and the Federal Budget," in Carolyn Weaver, ed., Social Security's Looming Surpluses; Prospects and Implications (Washington D.C.: American Enterprise Institute Press, 1990).
9. Valdes-Prieto, "Pension Reform in Latin America" and Schmidt-Hebbel, "Chile's Pension Revolution Coming of Age;" also see Musalem and Catalan, "Contractual Savings and Capital Markets Development."
10. On the existence of scale economies, see Estelle James and Robert Palacios, "Costs of Administering Public and Private Pension Plans," Finance and Development, Vol. 32, No. 2, pp. 12-15, June 1995; Estelle James, Gary Ferrier, James Smalhout and Dimitri Vittas, "Mutual Funds and Institutional Investments: What Is the Most Efficient Way to Set Up Individual Accounts in a Social Security System?" in John Shoven, ed., Administrative Costs and Social Security Privatization (Chicago: University of Chicago Press, 2000); Estelle James et al., "Administrative Costs and the Organization of Individual Account Systems: A Comparative Perspective," in R. Holzmann and J. Stiglitz, eds., New Ideas About Old Age Security (Washington, D.C.: World Bank, 2001); and Olivia Mitchell, "Administrative Costs in Public and Private Retirement Systems," in Martin Feldstein, ed., Privatizing Social Security.
11. For more details, see James et al., "Mutual Funds and Institutional Investments"; James et al., "Administrative Costs and the Organization of Individual Account Systems;" Annika Sunden, "The Swedish Notional Account System," Annals of Public and Cooperative Economics, Vol. 69, No. 4, 1998; and Hermann Von Gersdorff, "The Bolivian Pension Reform: Innovative Solutions to Common Problems," World Bank Working Paper, 1997.
12. Max Alier and Dimitri Vittas, "Personal Pension Plans and Stockmarket Volatility," in R. Holzmann and J. Stiglitz, eds., New Ideas About Old Age Security (Washington, D.C.: World Bank, 2001). Also see the discussion in Liqun Liu, Andrew J. Rettenmaier and Zijun Wang, "Social Security and Market Risk," National Center for Policy Analysis, NCPA Policy Report No. 244, July 2001.
13. This rule requires pension trustees to act as "prudent men," choosing only those investments that seem to offer reasonable income and safety relative to expected gains.
14. Acuna and Iglesias, "Chile's Pension Reform after 20 Years."
15. See Coronado et al., "Distributional Impacts of Proposed Changes to the Social Security System" and "The Progressivity of Social Security"; Feldstein, The Distributional Effects of Social Security Reform; and Gustman and Steinmeier, "How Effective Is Redistribution under the Social Security Benefit Formula?"
16. See Estelle James, Alejandra Cox-Edwards and Rebeca Wong, "The Gender Impact of Pension Reform: A Cross-Cutting Analysis," World Bank working paper, 2002.
17. See Estelle James and Sarah Brooks, "The Political Economy of Structural Pension Reform," in R. Holzmann and J. Stiglitz, New Ideas About Old Age Security (Washington, D.C.: World Bank, 2001).
18. See Edward Gramlich, "Different Approaches for Dealing with Social Security," Journal of Economic Perspectives, Vol. 10, No. 3, 1996, pp. 55-66; Martin Feldstein and Andrew Samwick, "The Transition Path in Privatizing Social Security," National Bureau of Economic Research, NBER Working Paper No. 5761, 1996; Martin Feldstein and Andrew Samwick, "Allocating Payroll Tax Revenue to Personal Retirement Accounts to Maintain Social Security Benefits and the Payroll Tax Rate," National Bureau of Economic Research, NBER Working Paper No. 7767, 2000; and Lawrence Kotlikoff, Kent Smetters and Jan Walliser, "Privatizing Social Security in the U.S.: Comparing the Options," Review of Economic Dynamics, Vol. 2, No. 3, 1999.
19. Gramlich, "Different Approaches for Dealing with Social Security."

[page]

Rodrigo Acuña and Augusto Iglesias, "Chile's Pension Reform after 20 Years," World Bank, draft of working paper, 2001.

Max Alier and Dimitri Vittas, "Personal Pension Plans and Stockmarket Volatility," in R. Holzmann and J. Stiglitz, eds., New Ideas About Old Age Security (Washington D.C.: World Bank, 2001).

Hazel Bateman and John Piggott, "Mandatory Occupational Pensions in Australia," Annals of Public and Cooperative Economics, Vol. 69, No. 4, 1998.

Luis Cerda and Gloria Grandolini, "Mexico: The 1997 Pension Reform Policy," World Bank, Research Working Paper No. 1933, 1997.

A. Chlon, M. Gora and M. Rutkowski, "Shaping Pension Reform in Poland: Security through Diversity," World Bank, Social Protection Discussion Paper No. 9923, 1999.

Julia Lynn Coronado, Don Fullerton and Thomas Glass, "Distributional Impacts of Proposed Changes to the Social Security System," in J. Poterba, ed., Tax Policy and the Economy, Vol. 13, 1999, pp. 149-86.

Julia Lynn Coronado, Don Fullerton and Thomas Glass, "The Progressivity of Social Security," National Bureau of Economic Research, NBER Working Paper No. 7520, 2000.

Elroy Dimson, Paul Marsh and Mike Staunton, Triumph of the Optimists (Princeton, N.J.: Princeton University Press, 2002).

Martin Feldstein and Andrew Samwick, "Allocating Payroll Tax Revenue to Personal Retirement Accounts to Maintain Social Security Benefits and the Payroll Tax Rate," National Bureau of Economic Research, NBER Working Paper No. 7767, 2000.

Martin Feldstein and Andrew Samwick, "The Transition Path in Privatizing Social Security," National Bureau of Economic Research, NBER Working Paper No. 5761, 1996.

Martin Feldstein, ed., Privatizing Social Security (Chicago: University of Chicago Press, 1998).

Martin Feldstein, ed., The Distributional Effects of Social Security Reform (Chicago: University of Chicago Press, 2000).

Edward Gramlich, "Different Approaches for Dealing with Social Security," Journal of Economic Perspectives, Vol. 10, No. 3, 1996, pp. 55-66.

Jonathan Gruber and David Wise, Social Security and Retirement around the World (Chicago: University of Chicago Press, 1998).

Alan L. Gustman and Thomas L. Steinmeier, "How Effective Is Redistribution under the Social Security Benefit Formula?" National Bureau of Economic Research, NBER Working Paper No. 7597, 2000.

Stefan Hepp, "The Swiss Multipillar System," Annals of Public and Cooperative Economics, Vol. 69, No. 4, 1998.

Augusto Iglesias and Robert Palacios, "Managing Public Pension Funds," Social Protection Working Paper, World Bank, 2000.

Estelle James and Robert Palacios, "Costs of Administering Public and Private Pension Plans," Finance and Development, Vol. 32, No. 2, pp. 12-15, June 1995.

Estelle James, "The Political Economy of Social Security Reform: A Cross-Country Review," Annals of Public and Cooperative Economics, Vol. 69, No. 4, 1998.

Estelle James, "New Models for Old Age Security: Experiments, Evidence and Unanswered Questions," World Bank Research Observer, Vol. 13, No. 2, 1998.

Estelle James, Gary Ferrier, James Smalhout and Dimitri Vittas, "Mutual Funds and Institutional Investments: What Is the Most Efficient Way to Set Up Individual Accounts in a Social Security System?" in John Shoven, ed., Administrative Costs and Social Security Privatization (Chicago: University of Chicago Press, 2000).

Estelle James, James Smalhout and Dimitri Vittas, "Administrative Costs and the Organization of Individual Account Systems: A Comparative Perspective," in R. Holzmann and J. Stiglitz, eds., New Ideas About Old Age Security (Washington, D.C.: World Bank, 2001).

Estelle James and Sarah Brooks, "The Political Economy of Structural Pension Reform," in R. Holzmann and J. Stiglitz, New Ideas About Old Age Security (Washington, D.C.: World Bank, 2001).

Estelle James, Alejandra Cox-Edwards and Rebeca Wong, "The Gender Impact of Pension Reform: A Cross-Cutting Analysis," World Bank Working Paper, 2002.

Paul Johnson, "Pension Reform in the United Kingdom," Annals of Public and Cooperative Economics, Vol. 69, No. 4, 1998.

Lawrence Kotlikoff, Kent Smetters and Jan Walliser, "Privatizing Social Security in the U.S.: Comparing the Options," Review of Economic Dynamics, Vol. 2, No. 3, 1999.

Lawrence Kotlikoff, Jagadeesh Gokhale and John Sabelhaus, "Understanding the Postwar Decline in United States Saving: A Cohort Analysis," in The Brookings Papers on Economic Activity (Washington, D.C.: Brookings Institution, 1996).

Lawrence Kotlikoff, "Generational Policy," in Alan Auerbach and Martin Feldstein, eds., The Handbook of Public Economics (New York: North-Holland, forthcoming).

Michael Marlow and Mark W. Crain, "The Causal Relationship Between Social Security and the Federal Budget" in Carolyn Weaver, ed., Social Security's Looming Surpluses; Prospects and Implications (Washington D.C.: American Enterprise Institute Press, 1990).

Olivia Mitchell, "Administrative Costs in Public and Private Retirement Systems," in Martin Feldstein, ed., Privatizing Social Security (Chicago: University of Chicago Press, 1998).

Alberto Musalem and Mario Catalan, "Contractual Savings and Capital Markets Development: An Analysis of OECD Countries' Evidence," World Bank Working Paper, 1999.

Robert Palacios and Roberto Rocha, "The Hungarian Pension System in Transition," World Bank, Social Protection Discussion Paper, 1997.

Monika Queisser, The Second-Generation Pension Reforms in Latin America (Paris: OECD Development Centre Studies, 1998).

Klaus Schmidt-Hebbel, "Does Pension Reform Really Spur Productivity, Saving and Growth?" Central Bank of Chile Working Papers 33, Santiago, Chile, 1998.

Klaus Schmidt-Hebbel, "Chile's Pension Revolution Coming of Age," paper prepared for the DIA Project: "Vergleich alternativer Alterversicherungssysteme in Chile, Deutschland, Groossbritannien, den Niederlanden und den USA," 1999.

Klaus Schmidt-Hebbel, "Latin America's Pension Revolution: A Review of Approaches and Experience," World Bank ABCDE Conference, Washington, D.C., 1999.

John Shoven, ed., Administrative Costs and Social Security Privatization (Chicago: University of Chicago Press and National Bureau of Economic Research, 2000).

Annika Sunden, "The Swedish Notional Account System," Annals of Public and Cooperative Economics, Vol. 69, No. 4, 1998.

Hermann Von Gersdorff, "The Bolivian Pension Reform: Innovative Solutions to Common Problems," World Bank Working Paper, 1997.

Salvador Valdes-Prieto, "Pension Reform in Latin America: Transition and Implementation Problems," Annals of Public and Cooperative Economics, Vol. 69, No. 4, 1998.

World Bank, Averting the Old Age Crisis: Policies to Protect the Old and Promote Growth (Washington, D.C.: World Bank and Oxford University Press, 1994).

[page]

Estelle James is principal author of Averting the Old Age Crisis: Policies to Protect the Old and Promote Growth (Washington, D.C.: World Bank and Oxford University Press, 1994) and is currently a consultant to the World Bank and other organizations. She was previously Lead Economist in the Research Department of the World Bank and Director of its Flagship Course on Social Security Reform. She also served as a member of the President's Commission to Strengthen Social Security in the United States.