The argument that competition increases polarization assumes that liberals want to read liberal newspapers and conservatives conservative ones. Natural as that assumption is, it conflicts with one of the points on which left and right agree - that people consume news and opinion in order to become well informed about public issues. Were this true, liberals would read conservative newspapers, and conservatives liberal newspapers, just as scientists test their hypotheses by confronting them with data that may refute them.

But that is not how ordinary people (or, for that matter, scientists) approach political and social issues. The issues are too numerous, uncertain and complex, and the benefit to an individual of becoming well informed about them too slight, to invite sustained, disinterested attention. Moreover, people don't like being in a state of doubt, so they look for information that will support rather than undermine their existing beliefs. They're also uncomfortable seeing their beliefs challenged on issues that are bound up with their economic welfare, physical safety or religious and moral views.

[However], for that sliver of a sliver that invites challenges to its biases by reading The New York Times and The Wall Street Journal, that watches CNN and Fox, that reads Brent Bozell and Eric Alterman and everything in between, the increased polarization of the media provides a richer fare than ever before.