Circling The Wagons

By DAVID BROOKS.

E-mail: dabrooks@nytimes.com

June 5, 2004

Over the next few months, I hope to write a fair bit about the dominant feature of our political life: polarization. I hope to figure out how deeply split the nation is, and what exactly it is we are fighting about -- questions that leave me, at present, confused.

Today's topic is what it means to be a partisan, because partisanship is the building block of polarization.

In a perfectly rational world, citizens would figure out which parties best represent their interests and their values, and they would provisionally attach themselves to those parties. If their situations changed or their interests changed, then their party affiliations would change.

But that is not how things work in real life. As Donald Green, Bradley Palmquist and Eric Schickler argue in their book, ''Partisan Hearts and Minds,'' most people either inherit their party affiliations from their parents, or they form an attachment to one party or another early in adulthood. Few people switch parties once they hit middle age. Even major historic events like the world wars and the Watergate scandal do not cause large numbers of people to switch.

Moreover, Green, Palmquist and Schickler continue, people do not choose parties by comparing platforms and then figuring out where the nation's interests lie. Drawing on a vast range of data, these political scientists argue that party attachment is more like attachment to a religious denomination or a social club. People have stereotypes in their heads about what Democrats are like and what Republicans are like, and they gravitate toward the party made up of people like themselves.

Once they have formed an affiliation, people bend their philosophies and their perceptions of reality so they become more and more aligned with members of their political tribe.

Paul Goren of Arizona State University has used survey data to track the same voters over time. Under the classic model, you'd expect to find that people who valued equal opportunity would become Democrats and that people who valued limited government would become Republicans.

In fact, you're more likely to find that people become Democrats first, then place increasing value on equal opportunity, or they become Republicans first, then place increasing value on limited government. Party affiliation often shapes values, not the other way around.

Party affiliation even shapes people's perceptions of reality. In 1960, Angus Campbell and others published a classic text, ''The American Voter,'' in which they argued that partisanship serves as a filter. A partisan filters out facts that are inconsistent with the party's approved worldview and exaggerates facts that confirm it.

That observation has been criticized by some political scientists, who see voters as reasonably rational. But many political scientists are coming back to Campbell's conclusion: people's perceptions are blatantly biased by partisanship.

For example, the Princeton political scientist Larry Bartels has pointed to survey data collected after the Reagan and Clinton presidencies. In 1988, voters were asked if they thought the nation's inflation rate had fallen during the Reagan presidency.

In fact, it did. The inflation rate fell from 13.5 percent to 4.1 percent. But only 8 percent of strong Democrats said the rate had fallen. Fifty percent of partisan Democrats believed that inflation had risen under Reagan. Strong Republicans had a much sunnier and more accurate impression of economic trends. Forty-seven percent said inflation had declined.

Then, at the end of the Clinton presidency, voters were asked similar questions about how the country had fared in the previous eight years. This time, it was Republicans who were inaccurate and negative. Democrats were much more positive. Bartels concludes that partisan loyalties have a pervasive influence on how people see the world. They reinforce and exaggerate differences of opinion between Republicans and Democrats.

The overall impression one gets from these political scientists is that politics is a tribal business. Americans congregate into rival political communities, then embrace one-sided attitudes and perceptions. That suggests that political polarization is the result of deep and self-reinforcing psychological and social forces.

This theory doesn't explain how the country moves through cycles of greater and lesser polarization. Still, I have to say, depressingly, this picture of tribal and subrational partisanship does accord with the reality we see around us every day.

URL: http://www.nytimes.com