As measured by the most recent income data available from the Current Population Survey (CPS), between 1996 and 2002 median household income increased 4.7 percent more than inflation. That statistic compares a "snapshot" of households in 1996 with another "snapshot" in 2002. It is not a picture of what happened to the same households over that time period. Medians, like those from the CPS, can conceal an enormous amount of movement in the income of individual households. This report uses the most recent longitudinal data available from the Survey of Income and Program Participation (SIPP), to examine movements in the incomes of the same households within the income distribution during the first part of this period — 1996 to 1999 (see Text Box: Household Income).¹

Income quintiles were constructed for each year — 1996 through 1999 — (see Text Box: Constructing Income Quintiles) using data collected in the SIPP. Using longitudinal data it is possible to identify and discuss factors that result in the rise and fall in the incomes for the same U.S. households (see Text Box: What Makes SIPP a Longitudinal Survey?).

• Of  U.S. households, 66 percent in the top quintile and 62 percent in the bottom quintile were in the same quintile 3 years later (see Figure 1).
• Between 41 percent and 47 percent of households in the middle three quintiles were in the same quintile 3 years later (see Figure 1).
• Of U.S. households, 13 million (13 percent) experienced changes in their annual income between 1996 and 1999 that resulted in their moving either up or down two or more quintiles in the income distribution (see Table 3).
• Of these 13 million households, the 34 percent (4.3 million) in the bottom and the second quintiles experienced the largest percentage gain in annual household income between 1996 and 1999, and the 11 percent (1.5 million) that started in the middle quintile experienced the largest absolute gain in annual household income (see Table 3).
• Of these 13 million households, 22 percent (2.9 million) of those that started in the top quintile experienced a decline of two or more quintiles between 1996 and 1999 (see Table 3).
• Householders with lower levels of education were more likely to remain in or move into a lower quintile compared with householders with higher levels of education (see Figure 2).
• Householders who were wid-owed were more likely to remain in or move into a lower quintile compared with householders who were not widows or widowers (see Figure 3).
• Householders under age 45 were more likely to move into a higher quintile than older householders, while householders age 55 and over were more likely to move into a lower quintile than younger householders (see Figure 4).

_______________
¹ The population represented is the civilian noninstitutionalized population of the United States. See the “Source of the Data” section for more details.

² The estimates in this report are based on responses from a sample of the population. As with all surveys, estimates may vary from the actual values because of sampling variation or other factors. All statements made in this report have undergone statistical testing and pass Census Bureau standards for statistical accuracy.