Law & Society Blog

Values are sets of normative principles and resulting cognitive beliefs that guide the acquisition of human knowledge and provide criteria to judge human actions and emotions. The empirical study of values provides, among other things, fascinating insights into which beliefs tend to occur in clusters. For example, respondents who answer the question “Is God very important in your life?” in the affirmative, are highly likely to also be anti-abortion, anti-gay, anti-divorce, deferential to national political authority, suspicious of foreign influences, and to favor obedience and religious faith over independence and self-determination in the education of their children. Compared to those who subscribe to secular-rational values, traditionalists also favor strong families, tend to have more children, seldom or never discuss politics in public, and believe in absolute moral standards. Similarly characteristic clusters of principles and beliefs emerge along the axis of survival values versus values of self-expression. Survival values focus on securing physical security and basic material needs; values of self-expression favor trust in others, political activism, and gender equality. For example, if someone rejects foreigners, gays and lesbians, or people with AIDS as neighbors, he or she is also likely to believe that men make better political leaders, that a woman needs a child to be fulfilled, that personal happiness is not an overriding goal in life, and that technological progress, on balance, does more good than evil. (For a more detailed account, see the statistical tables of the world values survey, as reported in Baker, America’s Crises of Values. Reality and Perception (Princeton Univ. Press 2004). The key findings are also reported in Ingelhart & Baker, Modernization, Cultural, Change, and the Persistence of Traditional Cultural Values, 65 Am. Soc. Rev. 19-51 (2000). Dan Kahan and Ronald Braman have undertaken a similar study, previously discussed in this blog, that arranges clusters of value-orientation along the dimensions of individualist/collectivist and hierarchical/egalitarian. While by no means a perfect match, it seems that the traditional/secular-rational dimension of the world values survey captures some of Kahan’s and Braman’s hierarchy/egalitarian axis. Whether and the what extent the survival/self-expression dimension corresponds to the collectivist/individualist axis, is a tougher question.)

One of the many possible applications of such research in the legal field would be (i) the creation of a predictive model for judicial behavior; (ii) empirical research into whether a court is attuned to or out of touch with prevailing values; and (iii) a cross-country comparison of judicially expressed versus commonly held values. Both (ii) and (iii) could help generate an index of judicial activism.

As to the creation of a predictive model of judicial behavior, suppose that values v1, v2, v3, and v4 are significantly positively correlated with each other, and significantly negatively correlated with v5 and v6. If judge A wrote a dissent in 1990, strongly endorsing v2, a concurring opinion endorsing v3 in 1995, and another dissent in 2000, disagreeing with a majority opinion endorsing v5, then it is reasonable to predict that A would vote to endorse v1 and v4 (with x% and y% probability, respectively) and to disagree with v6 (with z% probability). Given that the world values survey, briefly discussed here, provides correlations for responses to over 20 significant questions respectively along the traditional/secular-rational and the survival/self-expression axes, one could probably create a meaningfully detailed, predictive model. The beauty is that one could actually test the accuracy of the model on the basis of past decisions. In other words, feed ten early Scalia (Thomas, Breyer, Kennedy, Ginsburg, etc.) decisions into the model and see if it predicts his or her recent record. Some judges will be more predictable than others, but that’s useful information too, because the correlations established by the survey would permit us to quantify the degree of predictability by judge by issue. I’d be very interested in learning if anyone is already working on something along those lines. If so, please email me.

License

This work is published under a Creative Commons Attribution-Noncommercial 2.5 License.