What We Can Learn from the British Experience with Personal Accounts
Britain has used personal accounts in its public pension system longer than any other industrial country. Personal accounts fund individual workers' retirement benefits through savings, whereas pay-as-you-go systems, including the U.S. Social Security system, fund elderly benefits from current workers' taxes. The U.K. experience provides a good model of how not to do things in America. It shows that personal accounts can be good or bad, depending on how they are designed.
The British experience also shows the power of personal accounts. Though critics have pointed out flaws in the system, no British government has proposed getting rid of personal accounts. That is because both Labor and Conservative governments want to avoid the higher tax rates necessary to provide the same projected benefits from a public pay-as-you-go plan.
Evolution of British Social Security. Until 1978 the United Kingdom had only a simple "flat" (uniform) social security benefit - about 24 percent of the average wage - and practically every old person received it. The benefit was simple and modest, but affordable. It was supplemented by generous, voluntary employer-sponsored defined-benefit pensions that covered about half the labor force.
In 1978 the government added an earnings-related component - the State Earnings-Related Pension Scheme (SERPS) - to the flat public benefit. Employers who offered equivalent benefits could opt out and get a tax rebate. Many did. Thus, by the early 1980s, Britain had a funded, privately-managed benefit as part of its tax-financed old age security system. Beginning in 1988, workers were allowed to opt out of their employer's plan or SERPS by putting the tax rebate (also called a carve-out) into a personal account. Initially, there were few regulations on the accounts and little oversight.
Problem: Workers Making Wrong Choices. Since 1988, U.K. workers have had lots of choices. They can stay in SERPS, stay in their employer's plan or opt into their own accounts. Personal accounts are usually managed by insurance companies, who market their retail services directly to workers. About three-fourths of all workers have taken up one of the private options. However, in general, low-income workers would have done better by staying in SERPS, which has a progressive benefit formula, and workers with long-tenured jobs would have done well to stay in their employer's plan, since they lose their accumulated benefits when they opt out. But the commissioned salesmen who sold personal accounts didn't clarify these distinctions, and sold workers on the advantages of opting out of SERPS or their employer plan. The result is that many workers made wrong decisions, switched, and lost billions of pounds in foregone benefits. The lesson for the United States is: Unlimited investment choices in a mandatory old age security system is a mistake because many workers are inexperienced investors.
Problem: High Account Fees. Fees for personal accounts have been largely unregulated, and are often high. There are fees for making deposits, fees for withdrawals and fees for managing the funds in between. Any workerwho made a mistake in their first choice of asset manager then made a second mistake when they switched and had to pay both an exit fee to one and an entrance fee elsewhere. As a result, in many cases, U.K. workers' pensions will be 25 percent to 30 percent lower than they would have been in the absence of fees.
The lesson for the United States: When investments are voluntary and individual, and are sold through the retail market, the net returns are likely to be disappointing. A mandatory personal account system working through the institutional (wholesale) market will secure much lower fees by aggregating small contributions and cutting marketing expenses - and is likely to produce better results. Workers' choices can be limited to a small number of stock and bond portfolios indexed to broadly diversified market benchmarks. Fees for personal accounts modeled on the Thrift Savings Plan for U.S. federal employees are projected to run about 0.3 percent of assets, which would reduce final pensions by about 6 percent. This is very different from the U.K. model.
Problem: Gaming the System. U.K. workers can also opt back into SERPS, and their system provides incentives to do so. You get a bigger bang for your buck by putting it into your personal account when you are 20, since it will compound interest for 45 to 50 years. But when you are 50 you get a bigger bang by putting it into the government's pay-as-you-go system, because at that point you get the same credit for a year of contributions regardless of your age. So it is not surprising that some workers jump back into the state system in their 50s. In fact, the British insurance industry is now advising older workers to do precisely that. But it is costly for the public treasury.
If the United States reforms Social Security by allowing workers to voluntarily switch to personal accounts, rules should limit the option to switch back. Excepting Britain, almost all countries with personal account systems prohibit switching funds back to the pay-as-you-go system.
Problem: Declining Relative Value of the Basic Benefit. At the same time that the British government was adding privately-managed options, it looked at long-term projections, realized that its population was aging, and decided to keep down costs for its public flat benefit. The flat benefit had been wage-indexed, and increased every year at the same rate as pay, remaining at about 24 percent of the average wage. Workers with low contributions got little above this basic safety-net benefit.
In 1981, the government switched to price-indexation. Wages generally go up 1 to 2 percent per year faster than prices, due to productivity gains, so the switch to price indexation saved the government lots of money. The U.K. system is now solvent into the foreseeable future because it pays practically no escalating benefits. Although wages have gone up more than 40 percent, the basic U.K. pension has the same purchasing power that it had 25 years ago. So the basic benefit is now barely 15 percent of the average wage - way below the poverty line or the average standard of living. The benefit is so low that it is below the income threshold for means-tested (welfare) benefits, and more than one-third of British pensioners receive means-tested benefits.
Since the income threshold for means-tested benefits is linked to wages, it will continue to rise, and more and more retirees will become eligible for social assistance. This is inefficient because it discourages voluntary saving, incurs high transactions costs for means-testing, and misses some people who qualify.
In contrast, in the United States the initial Social Security benefit is indexed to wages and thereafter rises with prices. Switching to price indexing of initial social security benefit is one of the proposals under discussion. But under price-indexing the average defined Social Security benefit would remain constant. That benefit, averaged over all recipients, is now 32 percent of the wage of the average full-time worker, but would fall to 16 percent in 75 years, and lower thereafter. Income from personal accounts would fill some of the gap, but not all. If we switch to price-indexing initial benefits today, we will eventually find ourselves in the same unhappy situation as the British.
Conclusion. The British system was designed with too much choice and too little safety net. In designing a personal account system for U.S. Social Security, we should take steps to avoid those twin dangers.
Estelle James, Ph.D., is a consulting economist on Social Security issues to the National Center for Policy Analysis.