In a subscriber-only New Yorker piece reviewing what political science books have to say about the current political divide, Jill Lepore says that gerrymandering is much less responsible for the stark polarization in US politics than many people think:

This explanation [gerrymandering as cause of polarization] is wrong. First, polarization has taken place in both chambers at about the same time and rate and, since redistricting does not affect the Senate, it cannot wholly explain what’s happened in the House. Second, much polarization in the House has taken place in districts that have not been redrawn by legislators. Third, much of the polarizing in gerrymandered districts preceded their redrawing, The best calculation is that redistricting accounts for no more than ten or twenty per cent of the polarization in the House. Gerrymandering is bad for all kinds of reasons, but polarization isn’t one of them.