Earnings inequality is best measured on full-time workers aged 25-55 to avoid compositional noise
Evidence
We restrict the scope of this measure to the core workforce, deļ¬ned as full-time and fullyear workers ages 25ā55. This restriction avoids the interferences of parttime, school attendance, and retirement trends. We then adjust the earnings for top coding and inļ¬ation into 2001 U.S. dollars and exclude workers with annual earnings equal to or lower than $100. ā Ken-Hou Lin & Donald Tomaskovic-Devey, Financialization and U.S. Incomeā¦, 81