For the equations below, variables significant at p < .10 are underlined, variables significant at p < .05 are in bold, variables significant at p < .001 or better are in italics, and insignificant variables are in regular typeface.

WILLINGNESS TO SERVE ON JURY (from 1992 ANES): The ranges reflect the probit results reported in Uslaner (1998a) and the reanalyses I conducted based upon a simultaneous equation estimation with trust as endogenous. Other variables in the model are: discussing politics, trust in government, high school and college education, being divorced, the number of hours worked each week, being self-employed, talking to others about election campaigns, being married, and the number of hours one's spouse works each week.

IMPORTANCE OF CLASSICS (from 1993 gss, estimated by ordered probit): Other variables included education, ideology, a dummy variable for the South, relative financial status, party identification, gender, subjective social class, fundamentalism, size of community, frequency of attendance at religious services, age, and dummy variables for being black, Catholic, or Jewish.

ETHNIC REPRESENTATION (1994 Gss): In a two-stage least squares estimation with trust endogenous, the other predictors of ethnic representation are expectations that the national economy would improve, fundamentalism, age, and whether the government pays sufficient attention to blacks.

MORAL STANDARDS (1981 World Values Study in the United States): The measures of moral standards are all ten-point scales ranging from least to most acceptable. I estimated a seemingly unrelated system of equations for these standards of moral behavior together with other measures (joyriding, lying, cheating on taxes, and avoiding fares on public transportation).