>It used to be such studies focused on twins raised apart, but
>current studies instead look at twins raised together. One fairly
>common observation seems to be forgotten, however: identical twins
>raised together are likely to form tighter bonds with one another
>than fraternal twins. this may be simply because outsiders cannot
>tell them apart, or who knows what, but what it certainly does mean
>is that identical twins will be more likely to share opinions on key
>issues, especially ones with deep emotional content.

Lewontin points out that parents may treat identical twins
differently from fraternal twins.

>Another factor to consider is that conservatives against abortion
>tend to have more children than liberals. they then raise them as
>conservatives, and among these large families are likely to be more

This doesn't seem like a plausible way of explaining the results. If
it is true that conservatives have more children, then they will have
more non-identical ("fraternal") twins as well as more identical
twins, and the study purports to find greater agreement in political
views between the latter than between the former.