Testimony: Subcommittee on Social Security, Committee on Ways and Means
June 16, 2005
Over the past 20 years more than 30 countries, spread across Latin America, Eastern, Central and Western Europe, Australia and Hong Kong, have added funded privately managed plans to their mandatory social security systems. They did this to prevent public costs from rising, to increase national saving and to improve work incentives. Contributions to the accounts range from 2.5% of wages in Sweden to 12.5% in Chile and they are projected to supply between 30 and 90% of total benefits. In this discussion I consider both income streams-that coming from the personal accounts as well as the traditional defined benefit--as part of the country's social security system, so long as it is financed by taxes or mandatory contributions. I discuss how these countries funded the accounts, financed the transition, protected low earners and kept administrative costs low. The major lessons I draw are:
In comparison with other countries, current and projected benefits from and contributions to in the US social security system are relatively low. We need to decide whether we wish to maintain or to cut the scheduled level of mandatory retirement income-this is our central question. To maintain the scheduled level of social security benefits will cost more money than we are paying now, whether we do it through the personal account part or the traditional part. We will get higher returns and better economic incentives from this extra money if it goes into personal accounts-but this implies funding the accounts mainly through a small add-on. Although most countries in Latin America and Eastern and Central Europe with funded private plans used a carve-out from existing payroll taxes, several OECD countries used a mandatory add-on. Our initial conditions have more in common with these OECD countries.
One reason why countries adopted personal accounts was to increase national saving and economic growth-therefore more goods and services for everyone. This effect depends heavily on how the accounts and the transition are financed. National saving will increase if the accounts are funded by an add-on or by a carve-out whose transition costs are not debt-financed. Chile financed its transition out of a surplus on the rest of its public budget-and extensive analysis shows that this is largely responsible for the increased saving that has fueled Chile's economic growth.
Every country that has personal accounts also has a public safety net, including a minimum pension, to cushion financial and labor market risk.
Accounts can be organized through the retail market, as in most of Latin America and Eastern and Central Europe, or through the institutional market, as in several other countries and in the Thrift Saving Plan in the US. The institutional market has much lower administrative and marketing costs because it benefits from greater economies of scale and bargaining power (but the trade-off is less worker choice and less insulation from political pressures).
1. Funding the accounts through an add-on versus a carve-out
Countries that include funded privately managed plans in their social security systems fall into two different groups: Latin America and Eastern and Central Europe, on the one hand, and several OECD countries, on the other hand. In most Latin American and Eastern and Central European countries personal accounts were created during the 1990's and early 2000's as a remedy for public systems that were already on the verge of insolvency. In these countries payroll taxes were extremely high (often over 25% of wages), tax evasion was also high, workers retired well before age 60 and promised replacement rates were overly generous-for example, 70-80% of the worker's wage. An add-on contribution for the accounts clearly was not an option for this group. The new accounts were funded by a carve-out from the existing payroll tax, which was already too high.
In contrast, several OECD countries, such as Australia, Switzerland, Netherlands and Denmark, started out with relatively modest public benefits-not very different, on average, from benefits in the US, but more redistributive. In addition, in these countries employers have long provided pension plans, on a voluntary or collective bargained basis, which covered about half the labor force. As an alternative to increased public expenditures on pensions and as a way to raise national saving, in the 1980's and early 1990's these countries decided to make funded employer-sponsored plans mandatory for virtually the entire labor force. In effect, these plans became part of their social security systems, through an add-on for employers that didn't already provide them. The add-on has reached 9% in Australia, more in the other countries, and the combined target replacement rate is now 60-65%. The old employer plans were mainly defined benefit but the new ones are mostly defined contribution and many employers are also shifting their old plans to defined contribution (that is, to individual accounts).
I believe this add-on strategy for financing individual accounts would be the best for us too, except that I would organize it through workers, not employers, and I would aim for a much smaller add-on (of about 2%). This would hold our total benefit roughly where it is projected to be now, while allowing the traditional part of the system to become smaller and remain solvent.
2. Transition costs
The OECD countries did not face transition costs, because they did not divert money, they added-on. The Latin American and Eastern-Central European countries did face transition costs. They financed these costs in part by downsizing benefits and by a mix of fiscal stringency and debt finance. The transition has been intensively studied in Chile. Chile cut its obligations by raising retirement age substantially. It accumulated a budget surplus ahead of time to cover its early transition costs. The government is still paying 2-3% of GDP per year for the remnants of the old system, but it does so entirely by generating a surplus in the rest of its budget. In other words, no public debt finance is involved. Financing the transition without increasing the public debt is the major reason why Chile has increased its national saving, which in turn has increased its rate of economic growth.
If we in the US want to use pension reform as a way to increase national saving, either we must use an add-on or we must come up with a plan for transition finance that does not depend heavily on enlarging the public debt. Otherwise, we will be canceling out the increased private saving with increased public dissaving.
3. Minimum pension and other safety nets
All these countries include a safety net and guarantees to protect low earners. Every one of them has a minimum pension of some sort-examples are a minimum pension guarantee in Chile, a flat benefit that goes to every older resident in the Netherlands, or a widespread means- and asset-tested benefit in Australia. This cushions the risk from the accounts and from the labor market. In Switzerland, Mexico and Estonia the safety net rises with years of contributions, to bolster work incentives.
4. Administrative costs-retail versus institutional market
Most of the countries in Latin America and Central-Eastern Europe used the retail market to put workers into funds. But some countries, such as Bolivia, Panama (civil service) and Kosovo, have experimented with the institutional market. Large employer funds in the OECD also use the institutional market.
In the retail market pension fund managers can freely enter the industry, they establish a direct relationship with workers, and administrative costs tend to be relatively high, because of high marketing costs, diseconomies of dealing with many small accounts and price-inelasticity of demand in retail financial markets. Costs start out well over 10% of assets. Even though they fall steeply with asset growth, in countries as advanced as the UK they still exceed 1%, which will reduce final pensions by 20%.
In contrast, in the institutional market records are usually centralized, the money in small accounts is aggregated, and a competitive bidding process is held to choose a limited number of fund managers, among whom workers can choose. Administrative costs are cut by 2/3, because scale economies and bargaining power are larger and marketing costs are smaller. There is a trade-off of course--less choice for workers and greater danger of political pressure, if the bidding is organized by government. Sweden tries to mimic the fees of the institutional market while keeping the greater choice of the retail market, but it does so by imposing price controls. The institutional model is used by the U.S. Thrift Saving Plan and it was recommended by the President's Commission for our individual account system. I think this is appropriate for a system with many small accounts. After 7-8 years, administrative costs in this system would be .3%, which is lower than in other countries and also lower than in most mutual funds in our country.