Date of this Version
The continued scalability of a set of attitude items used in national sample surveys separated by 19 years is a remarkable and useful result. In a discipline lacking valid measures of socially important attitudes, such an outcome merits close study and wide dissemination. But exhibiting the stability of Stouffer's tolerance scale is not the prime end of the Williams et al. paper; they pursue an accounting of the "origins of tolerance." Nor is it the prime interest of the larger research from which their paper arises. Our concern in the larger study is with the political tolerance of the American people, their willingness to practice the democratic restraint shown by Lipset and Raab to be central to democracy in America. We are studying the conditions under which democratic restraint prospers or declines, and the changes over time in these attitudes and propensities. In my judgment, the paper by Williams et al. establishes the reliability of Stouffer's tolerance scale in 1954 and 1973, but it seriously distorts rather than illuminates understanding of political tolerance.