Politics across Generations: Family Transmission Reexamined (original) (raw)

2009, The Journal of Politics

Issues of measurement have also come to light that help account for variations in dyadic agreement. See Dalton (1980 analysis of the original Jennings and Niemi data that uses LISREL techniques to "correct" for measurement unreliability. A consequence of doing so is to increase the apparent level of parent-child agreement. See Appendix for further discussion. 2 2 All respondents were interviewed face-to-face in 1965, as were the great majority in 1973 and 1982 as well, when an abbreviated mail-back questionnaire was used for the more remotely located individuals. In 1997 approximately one-half interviews were face to face and the other half by telephone; computer-assisted technology was used for each mode. 3 Panel attrition and the absence of an initial parent interview account for the difference between the 935 members of th wave high school senior panel and the 636 parent-child pairs. The retention rate from the original 1556 pairs is 41%. 3 to decline, to level off, or even to increase as the offspring enter life stages resembling those occupied by their parents and political landscape shifts. A second central question about the dynamics of parental influence involves the presence of variations across political orie and subgroup variations among parent-child dyads. As noted above, certain kinds of parental orientations are more succe implanted by late adolescence than are others. Are these the ones that survive most readily over time? What properties are associated with these traits? We also noted that certain aspects of the parent-child nexus, especially parental characteristic enhanced the likelihood of political reproduction. Although a variety of such mediating variables have theoretical credenti can affect parent-child similarity on particular measures, explicitly political characteristics are of particular interest. After subject at hand is political socialization. In terms of social learning theory, transmission success should vary according to the strength of cue giving and reinforcem part of the socializer. Our analysis will employ two measures to evaluate this expectation, to be explained in more detail b of these mediating variables is traditional: the level of politicization within the family. The second capitalizes on the longi design of the study and is an indicator of parent salience and cue giving with respect to specific political orientations. The third topic we address is that of replication. One potentially troubling aspect of the Jennings and Niemi reports is that based on pairs formed from high school seniors of 1965, a cohort coming of age during such dramatic happenings as civil disturbances, the Vietnam war, political assassinations, and Watergate. That being so, it has been suggested that our findin cohort-centric, that preceding and succeeding cohorts would show different patterns of relationships, presumably including faithful political reproduction of their parents (Sears 1990; Sears and Funk 1999). Although some partial replications do not support this proposition (Allerbeck, Jennings, and Rosenmayr 1979), a more tho would include a replication of the measures with a subsequent cohort of parent-child pairs, one where the offspring were so under quite different historical circumstances. To make the test even stronger, these pairs should have the same family line original pairs. Which is to say, a third biological generation should be added to the two already available. One of our maj then, is to ascertain to what extent patterns of parent-child correspondence transcend eras. Study Design To address these topics we draw on a portion of the longitudinal parent-child political socialization project carried out by t University of Michigan's Survey Research Center and Center for Political Studies. Constituting the original core of the pro interviews with a national sample of 1669 high school seniors from the graduating class of 1965. Subsequent surveys cond 1973, 1982, and 1997 resulted in a four wave panel of 935 individuals, which represents an overall, unadjusted retention ra 56%. 2 During the first three waves efforts were also made to interview at least one parent, thereby enabling the constructio parent-child pairs as units of analysis. Altogether there are 636 pairs that have survived over the course of the study. 3 We to such dyads as being composed of Generations 1 (the parents) and 2 (their offspring).