The article by Kemper and Murtaugh1 in this issue of the Journal employs a familiar algorithm to suggest the dire consequences of the aging of a society. Generally, the age- and sex-specific rates of elderly people today are multiplied by projections of their numbers in the future. Such a model predicts hordes of disabled older persons needing nursing homes. Indeed, using similar logic, many people have warned of an impending shortage of nursing home beds and have suggested that we start building now. This approach to forecasting evokes two responses. The first is to challenge the statistics, but if anything,.