On February 24, 2006, the Urban Institute held a conference examining social security reform in six OECD countries. It was a time in which the social security debate in the United States seemed moribund. President George W. Bush had made some dramatic proposals in the previous year, but they were opposed vigorously by interest groups representing the elderly and by the Democratic party, and they gained little support elsewhere.
Yet, we know with certainty that the nation will have to return to this debate eventually. Our aging population combined with soaring health costs will create intolerable budget pressures within the next two decades. Although it is our health programs, Medicare and Medicaid, that will cause most of the problems, Social Security reform should be easier conceptually, if not politically. Consequently, exploring the varying approaches to social security reform in other democracies seemed interesting.
The good news is that reform is possible in a democracy, despite powerful opposing forces. The bad news is that in almost all the countries examined, reform had to wait until severe economic or budget problems, or both, arose. Such severe budget problems will not likely emerge in the United States for a decade or more, perhaps not until the 2020s.
Many countries have undertaken social security reform. In a variety of former Communist countries and in certain emerging economies, reform was necessary either because no system existed or because the system paid very low benefits. Replacing a defunct system with a more generous one is clearly much easier politically than cutting back on a functioning, generous system. For that reason, only developed countries with systems that functioned reasonably well prior to reform were chosen as subjects for the conference. Not all developed countries with interesting reforms could be accommodated, but those that were selected illustrate a wide variety of approaches. Conference papers were commissioned for Canada, Sweden, Japan, Germany, the United Kingdom, and Italy.
All countries represented have a more severely aging population than the United States. In some, such as Canada, the difference is slight, but in countries such as Japan and Italy, the population is aging extremely rapidly. In countries other than Canada and the UK, prereform public pensions were much more generous than in the United States, often providing much higher replacement rates at lower retirement ages. It is probably somewhat easier politically to cut back a highly generous system than one providing fairly low replacement rates.
The UK represented the most unusual case at the conference. Since its public pensions are price indexed and have eroded relative to the working population's standard of living, it is the only country now discussing an increase in its generosity. A commission has recommended that the UK adopt wage indexing, to be paid for by increasing the retirement age. Other countries' experiences suggest that the UK would make a huge mistake to do the former if it does not do the latter.
The UK has also taught the rest of the world valuable lessons in how not to implement individual accounts. The UK developed a system similar to that proposed by President Bush: individuals could opt out of the public system and obtain a payroll tax rebate to be invested privately. Here, the UK made two big mistakes. First, the information provided by the government to help people choose between private and public pensions underestimated changes in expected life, thus making the private approach appear more advantageous than it actually was. Second, sellers of mutual funds and of other private investments were not adequately regulated; overenthusiastic and outright dishonest salesmen put people into ill-advised investments. The Bush proposal avoided the latter problem by designing private accounts to resemble the civil service thrift plan, which offers a limited number of highly diversified, relatively safe investments. However, like the UK's approach, the Bush proposal may have been somewhat misleading. Cuts in traditional benefits were to repay deposits into the private accounts. The repayment carried a real interest rate of 3 percent. People may not have understood that the rate was probably higher than that which could be earned on a risk-free investment in inflation-indexed bonds or in a typical life-cycle fund.
The Canadian reform also had some unusual features. Canada decided to partially fund its pension system by increasing payroll taxes significantly—the rate went from 5.6 to 9.9 percent—and by depositing the resulting surplus into a pension trust fund. This fund could be invested in private assets to obtain higher returns. Because the pension system is a joint federal-provincial responsibility, the trust fund is entirely separate from the budgets of the federal government or the individual provinces. This makes it a good deal harder to use the trust fund indirectly to finance spending outside the pension system. Canada also created an independent board to be sure that investment decisions were not biased by political influences. Pension plan benefits were reduced somewhat by increasing the number of years in the benefit formula over which income is averaged, but the bulk of the reduction in Canada's long-run deficit came from the tax increase and the higher rate of return on their growing trust fund.
It is hard to imagine a similar reform in the United States. Americans would probably not tolerate such a significant increase in the payroll tax, and in many states, the so-called independent boards that invest state pension money have not been able to avoid responding to political pressures.
Most other countries represented in the conference started with payroll tax rates far higher than Canada's, and they faced the prospect of ever-increasing rates if the growth of benefits were not curbed. In Germany, Japan, and Italy, a series of reforms has decreased replacement rates and increased retirement ages. In Japan, reform has been aided by the fact that the system must be reviewed every five years and the legislature must respond when financial problems are reported.
Probably, the most interesting reform occurred in Sweden. They converted their defined benefit pay-as-you go system into something called a notional defined contribution system that is also pay-as-you-go. Tax payments are accumulate in a paper (or notional) account, and an interest rate equal to the rate of wage growth is paid on the account balance. No real investment underlies these accounts. Upon retirement, the accumulated amount is converted into an annuity appropriate to a person's age at retirement and the life expectancy of his or her cohort. The Swedes avoided divisive debates about the appropriate retirement age and allowed people to retire at any time after age 61, but, of course, the later the retirement age, the higher the pension. The annuity assumes a real rate of return of 1.6 percent per year. If wages increase by a greater amount, annuities will increase and if by a smaller amount, annuities will decrease.
Note all of the automatic adjustments in the Swedish system. The system adjusts to changes in the rate of economic growth, because the interest rate paid on the notional defined contribution accounts is linked to wage growth, as is the generosity of the annuity. Changes in expected life are automatically considered when calculating the annuity. As if this were not enough protection against economic and demographic surprises, there is a further automatic adjustment: if the present value of the system's assets falls below the present value of its liabilities, the interest rate paid on accumulated balances and the benefits received by pensioners are automatically adjusted downward.
A funded individual account system was added to the pay-as-you-go system and will help offset some of the benefit reductions, compared to the benefits that were earlier promised. Investors are offered a complex choice of more than 700 mutual funds, but there is also a default investment that many choose.
Variants on the Swedish system have now been adopted in Italy and in some former Communist countries. That the system automatically gravitates toward sustainability is one reason for its appeal.
Canada, Japan, Germany, and Italy all have automatic mechanisms, similar to Sweden's, that reduce the growth of benefits when certain triggers are pulled. The triggers depend on economic or demographic variables, or on an actuarial review. Such mechanisms may be politically acceptable, because their effects are harder to predict than those of more explicit benefit cuts. Admittedly, most of the reforms considered in this conference are extremely complex and their implications may not have been clearly understood by most voters. This may be distressing to advocates of transparency in government, but it was certainly convenient politically.
During the social security debate in the United States, widespread agreement held that current retirees and those near retirement should not be affected by benefit cuts. (Near retirement was defined as being 55 and older.) Essentially, we could not reduce the rate of benefit growth and the associated benefits of compounding savings for 10 years. We might question this value judgment, because it is older generations who have benefited most from the current system. Also, the judgment created a major problem for the Bush proposals. The diversion of payroll taxes into individual accounts would have initially been paid for by borrowing and the borrowing would have lasted a long time—until cuts in traditional benefits gradually closed the deficit. The resulting accumulated debt could have been significantly smaller were benefit growth to be reduced sooner, by affecting the retired and the nearly retired.
The countries represented in the conference treated the already retired very differently. Japan and Italy reduced their promised benefits significantly while Canada and Sweden were fairly lenient, although the already retired could be affected by their automatic balancing mechanisms.
Different countries used different political approaches to reform. In Canada, the provinces' important role in the pension system meant that the federal ruling party had to cooperate with the different parties ruling the provinces. There was no choice but to be multipartisan. In Sweden, a parliamentary committee of seven parties worked on the reform and in the end, five of the parties agreed to move forward. In the United States, it is not likely that progress could be made without bipartisan cooperation. Our recent debate was hampered when President Bush's opponents did not put forward proposals of their own, so it was hard to start a bargaining process that could have led to reform.
In Italy, the legislature delegated the responsibility for designing reforms to the Ministry of Finance. Only a wild dreamer could imagine such an approach in the United States.
Although other countries must be congratulated for undertaking painful and therefore courageous reforms, to say that all have permanently solved their problems would be an exaggeration. In all cases, the automatic balancing mechanisms are relatively new and have not yet gone into effect. If any of the triggers are pulled, the political reaction will be interesting. In some countries, the reforms will take effect slowly and perseverance will be necessary. That is particularly true in Italy, with its 10-year phase-in period. (Unfortunately, while benefit costs remain high, Italy will require a payroll tax rate of over 30 percent-a severe drag on employment.) Most countries also face unresolved problems related to rising health costs, a problem that is much more severe in the United States than the problem posed by our social security system.
Yet, an American cannot help but be envious of the progress made abroad. We can only hope that we can make similar progress, but without the budget and economic traumas that spurred reform in other countries. I think that the lessons for American legislators are clear. First, progress here, without the cooperation of the major political parties that characterized the Swedish and Canadian reforms, is impossible. Second, once the reform debate begins, we can consider options that imply a radical change in the structure of the Social Security program. Third, America would be wise to buttress reforms with automatic mechanisms that ensure the program will be sustainable, even if the demographic and economic assumptions underlying reform turn out to be less favorable than expected. All in all, the Swedish reform has some attractive features, especially the many provisions that automatically ensure sustainability, and should be considered when the reform debate resumes in the United States.
