There is a considerable debate about the direction and causes of change in U.S. student performance over the last 25 years. This study (1) estimates the net effect of changing family characteristics and demographics on aptitude scores and (2) compares the expected changes to actual changes to estimate the effects of factors unrelated to family. The conclusions undercut the conventional wisdom about failing schools, deteriorating families, and ineffective public investments and policies. The study estimates that changing family characteristics would boost scores by about 7 percentile points. These gains come primarily from higher parental education and smaller family size, which translates into more resources per child. For non-Hispanic white students, the actual gains in scores were approximately the same as expected from family changes. However, black and Hispanic students made far larger gains than non-Hispanic white students, and only about one-third of the gains could be explained by changing family characteristics. These large unexplained gains for minority students may be evidence that additional public investment in schools and social programs and equal educational opportunity policies have had marked benefits. The authors caution that the results should not be interpreted to mean that conditions have improved for every student, family or school, only that averaging across all 14-18-year-old students over the last 20 years indicates a positive change.