Michael Barone writes that in the post New Deal era political scientists were frustrated that parties were ideologically scrambled.  How inconvenient that there were progressive Republicans and conservative Democrats.

The prayers of the political scientists in the 1940s, 1950s, and into the 1960s was that our party system would evolve into one with one clearly liberal party and one clearly conservative party.

According to Barone, they reasoned thusly:

How could voters make rational choices when the parties presented no coherent set of policies? Wouldn’t it be more democratic to have one liberal and one conservative party?

And now their prayers have been answered, the parties are nothing if not ideologically sorted, but today's political scientists:

…lament gridlock and polarization and partisanship and reminisce about those (mostly mythical) days when politicians of both parties got along.

Accepting that there was indeed a kind of consensus among political scientists (i.e, liberal political scientists) the question is how directly did they influence our present polarization? Barone himself seems to attribute our division mostly to geographical sorting, or  "political geography." 

Still, this is an important caveat from our intellectual history. Today it is progressive thinkers who most deplore polarization. Conservative cogitators seem much less concerned.