Maybe we just really like to fight. Photographer: Nicholas Kamm/AFP/Getty Images
Maybe we just really like to fight. Photographer: Nicholas Kamm/AFP/Getty Images

The most important problem facing our national government is political polarization, because it makes governing virtually impossible -- unless one party can dominate the House, Senate and White House.

What’s driving the problem? One view, favored by political scientist Alan Abramowitz of Emory University, is that Americans are growing more polarized and their increasingly extreme differences are reflected in Congress. Another view, associated with Morris Fiorina of Stanford University, is that polarization in Congress stems more from a disconnect between politicians and the general public. (By the way, both camps reject the view propounded on cable television: that gerrymandering is the main cause.)

I have long put my chips down on the Abramowitz side. And two new analyses suggest I've made a good bet.

The first, from the Pew Research Center, shows that the share of Democrats with very unfavorable opinions of Republicans, and the share of Republicans with very unfavorable opinions of Democrats, has risen to about 40 percent, from about 16 percent 20 years ago. And today, more than a quarter of Democrats and more than a third of Republicans believe that the other party is a “threat to the nation’s well-being.”

Yet the public is still significantly more centrist than Congress is, the Pew research shows. And that raises a core puzzle that has long favored the Fiorina view: If political polarization is driven by the public, how can Congress be so much more polarized than the people?

The answer may lie in new evidence that voter preferences have been misread and in fact vary substantially within even “moderate” congressional districts. These findings are from a study by political scientists Nolan McCarty of Princeton University, Jonathan Rodden of Stanford, Boris Shor of the University of Chicago, Chris Tausanovitch of the University of California at Los Angeles and Chris Warshaw of the Massachusetts Institute of Technology.

The old assumption that populations in moderate congressional districts tend to be centrist is crucial to Fiorina's suggestion that politicians are disconnected from the people. Extremist members of Congress from such districts are, after all, the poster children for the view that politicians exert the polarizing force. Similarly, when an incumbent loses a general election in such a district, the new representative (from the other party) typically votes in a manner much different than that of the previous representative, which is interpreted as evidence that politicians go to extremes on their own.

McCarty and co-authors begin their analysis by acknowledging how differently Democrats and Republicans from moderate districts represent their constituents. They then show, however, that those districts aren’t really so moderate: If you judge by looking at either the mean or the median voter, a population consisting of 100 extreme Republicans, 100 extreme Democrats and one person in the middle would be characterized as “moderate.” As the example shows, “districts that are moderate on average often do not contain large densities of moderates.”

Furthermore, the more internally polarized a district is, the greater the divergence in voting patterns by elected officials, the political scientists found. That may be, they suggest, because the elected officials are unsure about the precise number of people in each camp and also the number who will actually vote. If, for example, a Republican member of Congress assumes there are 102 extreme Republicans rather than 100, and the Democrat believes there are 102 extreme Democrats, then they would represent the district in much different ways.

McCarty and team conclude that “it is quite plausible that the rise of polarization in the U.S. Congress has also been driven by increasing within-district polarization associated with demographic and residential sorting in recent decades.” And that leads back to my core concern: To the extent that we the people are driving polarization, it is much harder to fix.

More time and attention needs to be spent finding ways to continue governing effectively despite polarization. (I have proposed some; not surprisingly, the reaction to those ideas has generally been quite negative.) Government by inertia is no effective government at all.

To contact the writer of this article: Peter Orszag at orszagbloomberg@gmail.com.

To contact the editor responsible for this article: Mary Duenwald at mduenwald@bloomberg.net.