In 1981, Chile’s military dictatorship sought to reduce government spending and labor costs by privatizing the oldest social insurance program in the Americas. The Chilean pension system was, by all accounts, in need of reform. The system’s deficit had risen to 25% of Chile’s gross domestic product, yet 93% of retirees received only the minimum pension benefit. Today, some people argue that American Social Security is also in need of reform. They urge America to follow Chile’s example by allowing workers to redirect all of their Social Security contributions to personal pension accounts. A closer look at the Chilean system’s performance over the past seventeen years, however, should be cause for caution.

How Chile Privatized Pensions

General Augusto Pinochet’s regime created a system of private funds--called Administradores de Fondos de Pensiones, or AFPs--to manage and administer workers’ individual retirement accounts and survivors’ and disability benefits. Every worker participating in this defined-contribution system designates an AFP to receive a mandatory payroll deduction of 10% of salary (up to $22,000), plus an additional 2.5% to 3.7% for death and disability insurance and administrative fees. (Employees may voluntarily contribute up to an additional $2,000 a month to their retirement accounts, although only the mandatory contribution is tax-deductible.) When workers who have contributed to an AFP for twenty years retire (at age sixty-five for men, sixty for women), they can use the accumulated funds to buy an annuity or draw down their account according to an actuarially determined schedule, as with an individual retirement account (IRA) in the United States.

Retirement plans in the United States and many other countries have traditionally depended on contributions from both employers and workers. But the Chilean system puts the burden on employees alone. In order to offset the absence of employer contributions to AFP accounts (as well as to health care and other social insurance programs), the dictatorship ordered a salary hike of 18%, so that workers took home more in wages even after contributing to the new system. Qualified workers were also issued interest-bearing "recognition bonds," yielding a 4% real return, corresponding to the accrued value of their contributions to the old social security system. There is a backup provision for retired workers with twenty years of previous contributions to AFPs. If their personal accounts become exhausted or cannot provide a specified minimum benefit, the government guarantees a minimum pension that is periodically adjusted for inflation; this minimum pension amounted to $119 a month in September 1997. The government also requires AFPs to pay an average annual return equal to at least 50% of the average return of all AFP accounts or two percentage points below it, whichever figure is higher.

Within thirteen months of imposition of the new plan, more than a million employees, 36% of the Chilean workforce, had signed up with an AFP. By late 1995, the AFPs’ assets totaled $25 billion, equal to 40% of Chile’s GDP, and by the year 2010, they are projected to grow to 110% of GDP. Today, almost 99% of the workforce has, at one time or another, affiliated with an AFP. But, by several measures, Chile’s workers may not be getting their money’s worth.

What Went Wrong

1. Volatility. For more than a decade, the returns on AFP accounts seemed spectacular. The selling off of state enterprises and, from 1985 to 1991, high interest rates contributed to an average annual real return over fifteen years of 16.6%, peaking at 35% from 1989 to 1991. Almost half of the investments were in government bonds that were indexed to inflation, which was high during that period. But subsequently Chile’s economy cooled, and so have returns on personal pension accounts. In 1994, more than half of the AFPs incurred losses. In 1995, average returns fell to -2.5%, and over the past three years they have averaged only 1.8%. Since 1995, the average dollar amount of pensions paid has also dropped.

2. High Expenses and Fees. Total AFP expenses range from 15% to 20% of annual contributions--an average of $62 per enrollee in 1995. (U.S. Social Security expenses, by way of contrast, amount to less than 2% of contributions.) Approximately one third of these expenses represent sales costs, which from 1988 to 1995 more than doubled as a percentage of total expenses as AFPs competed for enrollees, not by reducing charges but by mounting ever more extravagant marketing campaigns. Swayed by free toaster ovens and other prizes or the promise of bigger returns, 25% of enrollees switch AFPs each year. These expenses consume a higher percentage of low earners’ contributions than high earners’ contributions, and they reduce the rate of return for every Chilean. In 1995, fees and commissions amounted to 23.6 percent of contributions, or 2.4 percent of average wages. According to World Bank economist Hemant Shah, commissions reduced individuals' average rates of return between 1982 and 1995 from 12.7% to 7.4%, and between 1991 and 1995 from 12.9% to a mere 2.1%. One actuary has calculated that for a new enrollee the 3.5% gross yield in 1996 actually amounted to a return of -6.8%. That same year the profit margins for AFPs--five of which controlled 80% of the market, constituting an implicit cartel--averaged more than 22%.

3. Evasion and underreporting. According to Chilean economist Jaime Ruiz-Tagle, workers contributing to AFPs earned an average of $1,000 in February 1995 but declared an average taxable income of only $460. Only 58% of workers contributed anything at all. Evasion through underreporting is particularly widespread among low earners, who figure that the guaranteed minimum pension will exceed what their retirement funds can yield. Employers, too, often underreport payroll in order to evade other taxes and charges. In February 1996, there were 150,000 unresolved suits against employers for insufficient or nonexistent deposits of worker contributions.

4. Inadequate coverage. A United Nations Development Program report estimates that 40% of AFP contributors will require additional assistance. The less one earns and the longer one lives, the more likely it is that an AFP account will not suffice. Since women in Chile, as in the United States, earn less on the average, leave the workforce more frequently to bear and raise children, and outlive men, they are particularly at risk. U.S. women, in contrast, benefit from the redistributive nature of Social Security, which provides more generous benefits, as a proportion of income, to low earners. The only safety net for the poor is a minimal pension that provides barely enough to pay for a loaf of bread and a cup of coffee each day. And even that austere program is limited to 300,000 Chileans, excluding thousands of the most destitute citizens. Moreover, most of the self-employed, who constitute more than 28% of the Chilean workforce, are especially vulnerable because participation in an AFP plan is not mandatory for self-employed workers; as of 1996, only 10% of them had voluntarily enrolled.

5. High transition and supplementary costs. Add up the pensions under the new system and those still being paid under the old one, the "recognition bonds," the minimum pension, and other guarantees, and the private pension system is at least three times as costly to run as the system it replaced. Government spending on pensions currently amounts to 6% of Chilean GDP.

A FOOTNOTE:

The junta protected one class of Chileans from privatization. The military continues to this day to receive pensions under the old governmental system.