In Chapter 3, I hypothesize that the probability of the Supreme Court achieving high behavior conformity when issuing a ruling will increase when one of the following conditions has been met:

1. (1) The Court issues a ruling in a vertical issue area, or

2. (2) The Court issues a popular ruling in a lateral issue area.

However, the probability of the Court achieving high behavior conformity will decrease when:

1. (3) The Court issues an unpopular ruling in a lateral issue area.

This hypothesis can be formally stated as:

where μ is the average behavior conformity, α is a constant, lateral is an indicator variable for lateral issues, lateral × popularity is an interaction term between an indicator variable for lateral issues and a trichotomous indicator variable indicating the popularity of the Court's ruling, β1 and β2 are coefficients for the effects of these two variables, and ε is the unobserved disturbance. I code the trichotomous indicator variable as 1 if less than 30 percent of respondents in national public opinion surveys indicate opposition to the Court's ruling; as 0 if more than 70 percent of respondents indicate opposition, and as 0.5 otherwise. For those Supreme Court cases in which no national public opinion poll was conducted, I assume the Court faces little public opposition and code the trichotomous indicator variable as 1.