Key points are not available for this paper at this time.
Motivated by Benin and Johnsonss (1984) study of 2 samples of adult Nebraska siblings the authors model reciprocal influence between male siblings to determine whether there is reciprocal influence between siblings or a predominant flow of influence from older to younger siblings in the determination of educational attainment. Contrary to Benin and Johnson this reanalysis of their data shows that differences in common family variance among types of sibling pairs are not derived from an unusually high level of fraternal resemblance but only from the unusually low similarity between older sisters and younger brothers. While the Nebraska data do not identify a model with reciprocal effects between brothers educational attainments suitable data are available in Olnecks (1976) study of brothers from Kalamazoo Michigan. These permit the estimation of reciprocal effects of brothers educational attainments. In modeling the Kalamazoo data the authors use a mixture of multiple proxy reports and self-reports to take account of random and correlated response errors in the measured variables. In brief they find evidence of reciprocal influence of brothers levels of educational attainment net of the common effects of family background and the effect of each brothers mental ability on his schooling. The unconstrained effect of older brothers schooling on younger brothers schooling is larger than the reciprocal effect but there is no significant difference between the 2 effects. This leaves open the questions whether reciprocal influence occurs in other types of sibling pairs and whether there is a predominant effect of older on younger siblings. (authors)
Hauser et al. (Sat,) studied this question.