ARE BAD ESTIMATES GOOD ENOUGH FOR THE COURTS

The controversy distills to the old ecological fallacy question: Can t he individual-level relationship between race and vote be reliably inf erred from aggregate-level election results? The SSQ critics of Grofma n's ER (ecological regression) method rely largely on sloganeering rat her than on careful statistical exegesis, and there is little to add t o Grofman's response. Nevertheless, there are serious problems with Gr ofman's method. The Supreme Court made a supreme mistake in endorsing that method.