Comment: Public and Private Retirement Systems in oecd Countries

Robert C. Pozen

Council on Foreign Relations

November 15-16, 1996

The papers presented here on the global pension crisis in oecd countries focus primarily on publicly funded retirement systems. But the objective is to provide retirement security to citizens from all sources. Therefore, it is important to examine both publicly and privately funded retirement systems in these countries.

Publicly funded systems all involve mandatory contributions, and usually offer a flat benefit together with an earnings-related benefit. These two components can be seen, for example, in France and Japan. Privately funded systems are typically voluntary and based on employment, although some oecd countries now offer personally funded pensions. Personal pensions include peps in Britain, and iras in the United States.

Let me discuss public and private retirement systems separately.

Public Retirement Systems

Funding Structure. All public retirement systems are defined-benefit in that the government is promising retirement payments based on years of service and, to some extent, on earning. However, these government benefit promises are not generally funded by a distinct trust fund. Rather, they are funded on a pay-as-you-go basis. That is, contributions from today's workers are paying for benefits to today's retirees. This is a vulnerable funding structure, given the changing demographics in all oecd countries.

Funding Trends. At present, most public systems are taking in more money than they are paying out. But this is a temporary phenomenon, as the demographics are turning against the public systems. The ratio of workers to retirees will be declining sharply in most countries because the Baby Boomers will retire and they will live longer. This will happen most dramatically in Japan, as Haruhiko Kuroda explained, and will also occur in Europe and the United States As a result, these public retirement systems will experience substantial shortfalls between contributions and benefits during the first half of the 21st century. Such long-range shortfalls have significant implications for overall government budgets.

Pension Accounting. The accounting for these impending shortfalls in public retirement systems is quite weak throughout oecd countries. The temporary surpluses currently enjoyed by such systems are often included to help balance budgets, but the accrued long-term liabilities are often ignored. In the United States, for example, all proposals to balance the Federal budget are built on the back of the temporary surpluses of Social Security, which will begin to run deficits around 2012. Similarly, the accounting for public retirement deficits is obscured in the budgets of many European Union countries that are struggling to meet the debt-to-gnp ratios for monetary union. Unless there is an honest and widely disseminated accounting for public retirement systems, it will be difficult to generate citizen support for needed reforms.

Traditional Approaches. Many countries have moved recently to constrain the growth of benefit obligations by moving back the retirement age, reducing the size of certain benefits, and/or putting more emphasis on means testing. These trends in Europe are well summarized in Norbert Walter's paper today. However, it is politically very difficult to enact broad-based increases in participant contributions. For many citizens, these contributions are the largest tax they pay; for over 70 percent of U.S. citizens, their Social Security contributions are larger than their income taxes. It is obviously unpopular for politicians to offer only reductions in benefits, increases in retirement ages, and higher requirements for contributions. They may want to offer some "sweetener" by allowing citizens to control part of their contributions in a manner that may produce higher returns.

Privatization Initiatives. Privatization is a misnomer because it connotes total transfer of public systems to private systems. In fact, the realistic alternatives in all oecd countries involve partial privatization--a portion of participant contribution goes to an employer-based scheme or personal pension. The best illustration is serps in Britain--the earnings component of the national system--which may be contracted out to an occupational scheme or to a personal pension. Such contracting out has considerable appeal because it retains the public safety net as a floor, while allowing participants the opportunity to seek higher investment returns on a portion of their contributions through accounts under their control.

Specific U.S. Responses. Questions have been raised at this conference about whether the remaining public safety net(after contracting out a modest portion (2 percent) of Social Security contributions to a private security account (psa) (would be sufficient for low- and moderate-income workers. While any psa involves some risk, the net investment return from a psa is very likely to exceed the 1 percent to 2 percent return projected for Social Security during the 21st century. Moreover, the investment risk in the psa can be constrained by reasonable diversification requirements. A good model is the Section 404(c) regulations under erisa which, by analogy, would require psa participants to be offered at least three diversified pools with different risk/return characteristics, one of which must be a money market--type account. This may very well be a more attractive model than allowing psa investments in any individual security.

Questions have also been raised about whether the administrative charges to a psa would eat up any potential higher returns. First, the total annual expenses of fixed income funds in the United States average slightly less than l percent, and somewhat above 1 percent for equity mutual funds. The annual administrative cost for Social Security is just below 1 percent. Second, the key figure is the net investment return after administrative expenses. This net result has been on average over the last decade in the 6 percent range for fixed income funds and above 10 percent for equity funds; both are significantly higher than the 1 percent to 2 percent return projected for Social Security.

Private Retirement Systems

Funding Structure. Most private plans in oecd countries are employment-based and structured as defined-benefit plans. Japan has only defined-benefit plans and such plans are dominant in continental Europe. The Anglo-Saxon countries, by contrast, have been gradually shifting from defined-benefit to defined-contribution plans. Moreover, the Anglo-Saxon countries have started to offer individual- as well as employer-based defined-contribution plans. These trends are in evidence in Canada and Australia as well as the United States and Britain.

Funding Trends. It is more than a coincidence that higher funding levels for pensions are found in Anglo-Saxon countries than in the rest of the oecd countries. The Anglo-Saxon countries are the ones that have moved toward defined-contribution plans and these, by definition, are fully funded when the employer makes the appropriate contribution. The serious funding issues relate to defined-benefit plans--whether employers have made sufficient contributions plus investment returns to support the retirement promise. The funding of defined-benefit plans in Germany, France, and Italy is quite modest, as compared to the higher levels in the United States. Even so, some defined-benefit plans in the United States have a substantial shortfall between assets and liabilities. While this shortfall is guaranteed in part by a U.S. government corporation with the acronym pbgc, the pbgc is itself underfunded.

Pension Accounting. Like the public side, the company side of defined-benefit plans is poorly accounted for in many oecd countries. This is true of many continental European and Asian countries. In Japan, for example, the accounting standards do not require publicly traded companies to disclose their pension assets relative to their pension liabilities. Over 20 Japanese firms, because they are listed on the New York Stock Exchange, do disclose their pension assets/liabilities in accordance with fasb #87. For these Japanese firms, there was a 31 percent deterioration in pension-based underfunding in fiscal 1995, as compared with fiscal 1994. All oecd countries should move to an uniform accounting standard for pension funding, along the general lines of fasb #87. Such accounting would allow a more realistic assessment of a company's financial position by investors and public officials as well as by company directors.

Defined-Benefit Strategies. A number of countries are permitting defined-benefit plans to take a more flexible approach to investing assets in an effort to boost returns. This is generally leading European pension funds to invest more in stocks, although Germany and France continue their historic emphasis on bonds. In Japan, the government has substantially liberalized investment restrictions on pension funds after the traditional insurance and trust companies reduced their annual returns on pension assets to only 3 percent. Companies are also trying to reduce the size of their defined-benefit promise going forward, though they face constraints from collective bargaining. In addition, some companies have shifted partially from defined-benefit to defined-contribution structures.

Defined-Contribution Trends. As mentioned above, the trend is clearly from defined-benefit to defined-contribution. In the United States, there are already more participants in defined-contribution than in defined-benefit plans. While defined-benefit plans still have more assets than defined-contribution plans in the United States, the defined-contribution asset growth rate over the last five years has been 12 percent as compared with a 6 percent rate for defined-benefit asset growth. The U.K. has also moved substantially to defined-contribution plans, both employer and personal. This trend is developing more slowly in continental Europe, and is under serious discussion in Japan, which has traditionally offered only defined-benefit plans. But this movement from defined-benefit to defined-contribution plans involves a shift from credit risk of employers to market risk of investments. Employers and financial intermediaries need to educate participants in defined-contribution plans about the risks and returns of available investment alternatives.

Specific U.S. Response. One of the topics discussed at this conference is the relative merit of defined-benefit versus defined-contribution approaches if a portion of Social Security is privatized. The defined-benefit approach would have the Treasury investing a portion of Social Security monies in the stock market. As Laura Tyson suggested, such an approach would raise a number of questions, such as who would choose the stocks and how would proxies be voted? In addition, the Treasury would come under tremendous pressure to make socially useful investments, for example in housing and infrastructure, even if the financial returns were lower than those in the stock market.

Many advocates of the defined-benefit approach believe that the Treasury would invest through an index fund such as the s&p 500. This type of mechanized investing, however, could undermine the efficiency of the capital allocation process. The stock of the 498th company would receive a large influx of buying on a regular basis, while the stock of the 502nd company would be shut out of these Social Security flows. The capital allocation process would be better served by a more decentralized system in which capital would flow to the companies with the best prospects, regardless of whether they were included in a particular index. A defined-contribution approach, like the psa discussed above, would be more consistent with such a decentralized process for capital allocation than index investing by the Treasury.

Conclusions

The public retirement systems throughout the oecd will confront difficult financing problems during the 21st century. Many reforms are being considered within these systems, as well as a gradual shift from defined-benefit to defined-contribution structures.

However, time is of the essence. The cardinal rule is that any change in retirement systems must be adopted years before it takes effect. It is extremely difficult to reduce or delay any benefit for those in, or close to, benefit status. Since the baby boomers will reach age 65 in the year 2011, governments should focus now on the financing issues raised by the future retirement of this huge generation.