Documentation and methodology
The figure is adapted from Acs and Zimmerman (2008b), Figure 5, “Characteristics Associated with Leaving the Bottom Quintile.” Coefficients are based on a linear probability regression that includes these characteristics as well as dummy variables for age, the presence of children, and the presence of other adults in the household. Own and spouse work hours are measured in thousand-hour units. Acs and Zimmerman do not differentiate between spouses and permanent cohabiters, and interact the spouse hours variable with a dummy variable for the spouse’s presence. Only characteristics with statistically significant coefficients in at least one time period are shown. In the 1984–1994 time period, the coefficients for white, more than high school, disability, and spouse present are statistically significant at the 99 percent confidence level; high school education is statistically significant at the 95 percent confidence level; and male, homeowner, own hours, and spouse’s hours are statistically significant at the 90 percent confidence level. In the 1994–2004 time period, the coefficients for more than high school and own hours are statistically significant at the 99 percent confidence level.