Development Brief Number 44
November 1994
Redistribution across income levels in pay-as-you-go
Pay-as-you-go pension schemes don't redistribute much from the rich to the poor---and some redistribute the other way
Pay-as-you-go pension schemes can redistribute lifetime income from the rich to the poor. To achieve this redistributive objective, the United States and the Philippines provide higher wage replacement rates for individuals with lower lifetime wages. The Netherlands uses a flat benefit and Switzerland a combination of employment-related flat and progressive earnings-related benefits. Many developing countries use explicit minimum or maximum pension levels for the same purpose. None of these mechanisms seems to work as expected.[1]
Why the poor don't benefit more than the rich
Once certain income-specific characteristics of workers are taken into account, public schemes redistribute little lifetime income to the poor. For one thing, the well-off live longer than the poor and so collect pensions over a longer period. Many poor people contribute but die before they become old enough to collect benefits.
People who live longer than average require larger lifetime incomes to maintain their standard of living, and pensions are designed to insure against this eventuality. However, the longevity differences between rich and poor are predictable, not random, so the differences in benefits received because of these longevity differences are expected and constitute redistribution rather than insurance. Does society wish to transfer income to the rich because they live longer? We do not know whether well-informed citizens and policymakers would consider this equitable---yet that is exactly what many societies do.
Besides drawing a pension over fewer retirement years, lower-income workers usually enter the labor market earlier because they leave school earlier. Yet they often fail to accrue much extra pension credit during these extra contributing years. Germany and Hungary grant pension credit to university students, who thus earn pension benefits during those years without contributing for them. Pension benefit formulas that give more weight to earnings in the years just before retirement also penalize low-income workers, as compared with more educated, higher-income workers whose earnings rise with age. Such formulas are quite common in developing countries.
Financing mechanisms introduce another regressive element to most public pension plans. In OECD countries, where coverage is nearly universal, systems are financed through payroll taxes rather than general income taxes, and taxable earnings are usually capped. In developing countries, where coverage is sparse, the regressive effect comes from using general revenues to subsidize pension plans for the better-off. Guatemala pays a third of the costs of its public pension plan out of general revenues collected from the entire population even though its pension program covers less than a third of the population. Tax deductibility of contributions, a common feature of public pension plans, also favors higher-income workers over low-income workers.
Patterns of redistribution in industrial countries
Considering all these forces of regressive distribution, it is not surprising that empirical studies for the Netherlands, Sweden, the United Kingdom, and the United States show little if any redistribution from the lifetime rich to the lifetime poor. The public pension scheme in the Netherlands was found to have redistributed income primarily from unmarried to married individuals, regardless of income. A study of the Swedish system detected no intragenerational redistribution and suggested that the scheme might be regressive if mortality differences were taken into account. Income-related differences in mortality rates and earnings profiles in the United Kingdom offset progressive benefit features. In Italy, intragenerational redistribution has regional implications as well. One study found a pattern of transfers from the poor South to the richer North, the result of steeper age-earnings profiles, longer lifetimes, and a higher old age dependency ratio in the North.
Several studies have looked at the intragenerational effects of the U.S. social security system and found little or no difference in the rate of return to low- and high-income workers, particularly after adjusting for mortality rate differences. Studies have also shown a variety of other nontransparent redistributions: from single workers to married couples, from dual to single wage-earner families, from women in the work force to other women.
Regressive pension schemes in developing countries
Indirect evidence for developing countries suggests that intragenerational redistribution in public pension schemes is actually regressive. Typically, only formal sector workers in urban areas are covered by the plans. The very poor are usually excluded. Income-related differences in mortality rates, age of entry, and age-earnings profiles are even more pronounced than in OECD countries.
Many developing countries have multiple plans that provide more generous terms to privileged groups of workers with more political clout. In Brazil, higher-income workers have an easier time documenting their covered employment, so they are eligible for generous length-of-service-related benefits. Civil servants in Egypt and Mexico receive better inflation protection than private sector workers, and in Colombia they contribute at a lower rate than private sector workers, yet get the same wage replacement rate.
In Brazil, Hungary, Turkey, and many other countries, white-collar workers are more likely than blue-collar workers to retire at age 40 or 50, increasing their rate of return relative to those who continue working. In Colombia, though all current retirees receive net transfers that constitute approximately the same proportion of benefits for all income groups, the absolute value of the transfer is eight times as large for a high-income worker as it is for a minimum wage worker (see the figure).
Misuse of pension fund reserves contributes further to regressive intragenerational transfers. In Ecuador and the Philippines, pension fund reserves are lent to high-income workers at negative real interest rates of as much as 40% a year. In Trinidad and Tobago, well-off workers borrowed pension reserves for mortgages at below-market rates.
Not a single study for any country has presented strong evidence that the public pension scheme has substantially redistributed income from the lifetime rich to the lifetime poor once mortality differences are taken into account. In fact, in some countries, the redistribution goes from the poor to the rich.
http://worldbank.org/html/dec/Publications/Briefs/DB44.html