Achievement and Ascription in Educational Attainment: Genetic and Environmental Influences on Adolescent Schooling

1. Shared Environments of Twins The model estimated with the 6 types of siblings assumes that shared environmental effects are the same for twins and full siblings. Since twins share the same womb and are exact contemporaries, it could be argued that the environments of twins are more similar than that of full siblings. An alternative argument is that the environments of twins, who compete for maternal resources in the womb, might be less similar than the environments of full siblings who do not compete to the same extent. If the environments of twins are more similar one would expect correlations between twins to be larger than correlations between siblings. In fact Table 1 shows the opposite pattern: correlations for full siblings are systematically larger than those for DZ twins (.411>.356 for VIQ, .360>.332 for GPA, .332>.264 for CPL). This pattern is inconsistent with the hypothesis that the environments are more similar for twins than for full siblings, although it is consistent with the alternative hypothesis of greater dissimilarity of the environments of twins because of more intense competition. Another concern that looms large in critiques of twin studies is the possibility that the greater similarity of MZ twins compared to DZ twins is due in part to greater similarity of the environments of MZ twins (perhaps because their greater physical resemblance causes them to be treated more alike by parents and others). Extensive research has failed to support this hypothesis; the capsule result is that the environments of MZ twins are indeed more alike, but to the extent that environmental similarity is associated with phenotypic similarity it is as a consequence rather than a cause, in the sense that MZ twins are treated more alike because they look and behave more alike, not that the more similar environments make them more alike (Neale and Maes, forthcoming; Rowe 1994).