Recent research suggests that the "typical" workers used to evaluate the distributional impact of Social Security reform are not so typical after all. If reformers want to better understand how their proposals will affect typical workers, they must start with an accurate reflection of their true earnings.