PROJECT OVERVIEW

Income Replacement During Retirement
James P. Smith

Income security during retirement is widely regarded as a primary social achievement of the 20th century. There is concern, however, as to whether all the income security gains of the past century will be sustained. The ratio of retirees to workers is expected to increase, suggesting that social security benefits may fall. Meanwhile, the prevalence of private pensions, at least among younger workers, has ceased to expand.

Smith sheds light on such concerns by describing patterns of income replacement during retirement over the last 25 years. He shows how income replacement rates have varied with household income and marital status and how these relationships are changing. To do so, he develops a new methodology for computing income replacement. Smith's approach to these estimates reduces potential bias in several ways:

• The use of median (or quartile) instead of mean data. The prevailing use of mean incomes tends to overstate the level of income replacement for most households.

• The use of after-tax instead of pretax data. The prevailing use of pretax incomes seriously understates income replacement.

• Tracking income only within marital-status groups. Cross-status tracking of individual income confounds income changes due to alterations in marital status with changes due to retirement per se.

Smith also computes income replacement rates for groups with different income levels. For example, most households in the lower quarter of the income distribution currently enjoy full income replacement. This no doubt goes a long way towards explaining why many of these households save very little and retire at relatively young ages.