April 1, 2006
Jonathan A. Schwabish
Studies of earnings trends and earnings distributions are often hampered by public-use data sets that either do not adequately survey high earners or topcode high earnings. This paper uses the Social Security Administration’s Continuous Work History Sample (CWHS) and the March Current Population Survey (CPS) to analyze trends in earnings inequality during the 1990s and early 2000s, and to assess where significant differences may exist in the two data sets. The CWHS is used to track the changes in the share of earnings received by those at the very top of the distribution—in particular, those in the top 1 percent, 0.5 percent and 0.1 percent. The analysis shows that the earnings distribution in the CWHS is more evenly distributed at the top and includes more observations at the bottom than the CPS. The analysis also suggests that the share of earnings at the top of the distribution rose quickly between 1987 and 2000 but since that time has fallen precipitously. Similarly, inequality trended upward during the first part of the period before falling markedly during the stock market boom of the late 1990s.