You can go on the Internet and choose where you want to live and what you want to hear.

Some years ago, a historian examined the voting records of Members of Congress during the 1800’s to see if he could determine why they voted as they did. Was the dominant factor geographical — Westerners voting as a block against Easterners or Southerners — or political — Members voting with their parties — or professional — farmers voting against bankers and lawyers. He was unable to find any sort of pattern that would explain what went on in the Members’ minds as they cast their votes.

Then he had another idea. In the 1800’s, Congressional sessions were only a few months, somewhat like the State Legislature is now. None of the Members lived in Washington. Instead, they took rooms in the various hotels and boarding houses that were available. Could it be that conversations at the dinner table and afterward in those boarding houses determined how the Members would vote?

Yes. By checking their voting records against their places of lodging, he found his point of coordination. Regardless of where they came from, which party they belonged to or what their professions were, the Members’ votes were influenced most by what they heard from their fellow boarders.

I submit that the influence of boarding house conversations still impacts our politics today except that the boarding houses are now electronic. You can go on the Internet and choose where you want to live and what you want to hear. To get a list of possibilities, go to You’ll find Paul Krugman’s latest attack on Republicans in general and the tea party in particular right next to Rush Limbaugh’s most recent diatribe against all things and people he considers liberal. George Will and Charles Krauthammer opine alongside Chris Matthews and Rachel Maddow. Look at the menu and take up residence in any electronic boarding house you want.

The differences between them are significant. In those on the right, you’ll be told that “Republicans won the shutdown battle.” — Erick Erickson, In those on the left, you will hear that “Obamacare is working.” — E. J. Dione, The Washington Post. They offer not just a polarization of opinion but of perception of the facts.

If you stay with a single worldview, you never need consider anything that contradicts your own preconception of how things ought to be. That can be very comforting but leaves you vulnerable to some jolting surprises. It’s called “the Pauline Kael effect,” named after the film critic of The New Yorker magazine who was absolutely dumbstruck when Richard Nixon carried every state except Massachusetts in 1972. “It’s not possible!” she cried. “I don’t know a single person who voted for him!”

Those suffering from the Pauline Kael effect refuse to believe any polls whose outcome they don’t like. True, polls can be skewed by bad wording of the questions or improper drawing of the samples. To correct for that, RealClearPolitics lumps all of them together and publishes an average, which diminishes the opportunity for spin. Current RealClear polling information contains bad news for both sides:

The shutdown really did hurt Republicans. They are going to lose in Virginia, where their candidate is a Tea Party favorite. (However, they will win in New Jersey, where Governor Chris Christie very publicly distanced himself from Ted Cruz.)

The rollout of Obamacare really did hurt President Obama. His disapproval numbers are back where they were before the shutdown gave him a boost.

What does this mean for the future? No one knows; political memories can be very short. However, another bit of polling makes it “real clear” that public opinion is disgusted enough with the current polarization that it is beginning to turn a little mean. When that occurs, anything can happen and everyone’s personal opinion can be in danger, particularly those who “don’t know a single person” who disagrees with them.

Robert Bennett, former U.S. senator from Utah, is a part-time teacher, researcher and lecturer at the University of Utah's Hinckley Institute of Politics.