Christopher Hare is a Ph.D. candidate in Political Science at the University of Georgia. Keith T. Poole is the Philip H. Alston Jr. Distinguished Professor in the Department of Political Science at the University of Georgia. Howard Rosenthal is Professor of Politics at New York University and Roger Williams Straus Professor of Social Sciences, Emeritus, at Princeton. They, along with Dave Armstrong, Ryan Bakker, and Royce Carroll, are co-authors of the forthcoming book Analyzing Spatial Models of Choice and Judgment with R.
The most recent data show that polarization in Congress reached a new record high in 2013. Absent heightened electoral pressures or some form of partisan realignment, the trajectory of congressional polarization is unlikely to reverse course anytime soon. Members of Congress are remarkably stable in their ideological positions, and so polarization is likely quite “sticky.”
Using DW-NOMINATE scores, the graph below shows the ideological distance between the parties in both chambers between 1879 and 2013 (DW-NOMINATE scores measure legislators’ liberal-conservative positions using their roll call voting records). After a period of depolarization that ran through much of the mid-20th century, the parties started to become more ideologically distant beginning in the 1970s. This is true in both chambers, although polarization has progressed at a greater rate in the House. Congress is now more polarized than at any time since the end of Reconstruction.
Another way to track congressional polarization is to show the distribution of Democratic and Republican legislators on the liberal-conservative dimension across multiple years. The graphs below show the ideological makeup of the Democratic and Republican parties in the House and Senate in two Congresses: the 97th Congress (1981-1983) and the current, 113th Congress. House Democrats are shown in light blue, while Senate Democrats are shown in dark blue. Republicans are shown in corresponding shades of red. The party means are marked using the dotted blue/red lines in each plot. The positions of several noteworthy members of Congress are also shown.
These graphs illustrate several important points about congressional polarization: First, the dramatic shift to the right by the Republican Party. Second, the disappearance of ideological moderates in both parties. While centrist legislators like Sen. Sam Nunn (D-Ga.) were once common, Nunn would be part of a rare species if he re-entered Congress today. Third, members of Congress who were once solid liberals or conservatives relative to the rest of their parties have become relative moderates. We have demonstrated this point with Sen. John McCain (R-Ariz.). McCain would have been to the right of the mean Republican in the House and Senate in 1981, but is today considered a moderate given the rightward drift of his party. Finally, there has been a rise of ultra-conservative Republicans in Congress. In the 97th Congress, 21 of 193 House Republicans and 11 of 54 Senate Republicans had DW-NOMINATE scores greater than 0.5. There are now 124 of 233 Republicans in the House and 16 of 45 in the Senate greater than 0.5. Not surprisingly, this produces a much less manageable caucus.
Indeed, it is noteworthy that Speaker John Boehner (R-Ohio) broke the “Hastert Rule” and pushed forward several major pieces of legislation despite opposition from a majority of his own party in the 113th Congress (including the vote to end the government shutdown and raise the debt ceiling last October and again this week). These internal Republican splits reflect the increase in polarization rather than the moderation of polarization; namely, just how far the Republican Caucus stretches between the center-right to the far-right.
DW-NOMINATE scores also seem to do a good job of tapping into the rise of what Sean Theriault terms “partisan warfare” or “political warfare” in Congress. To use Theriault’s example, Sens. John Cornyn (R-Texas) and Ted Cruz (R-Texas) are both very conservative legislators who vote the same way on about 90 percent of roll call votes. But it seems reasonable to characterize Sen. Cornyn as a less bombastic and more compromising figure than Sen. Cruz. Even though these differences are primarily stylistic, they do sufficiently manifest themselves in roll call voting behavior that DW-NOMINATE is able to distinguish between the two, estimating a score of about 0.5 for Sen. Cornyn and 1.0 for Sen. Cruz. There are other important non-ideological dimensions–for example, a “talent” dimension–that differentiate legislators and are not explained by DW-NOMINATE scores. But polarization in Congress seems to be primarily based on ideological differences expressed in roll call voting as well as other choice behavior like campaign contribution activity.
Mathematically, partisan polarization in Congress cannot continue to expand indefinitely. But there are several directions that polarization could take in the coming years. Probably most unlikely is a “hot” decline in polarization caused by a breakup of the party system (as occurred before the Civil War) sparked by some calamity like a major economic crisis. More likely would be a “cooling off” period in which one or both parties respond to electoral pressures by gradually shifting back to the ideological center, most likely via replacement (i.e., nominating more moderate candidates). Of course, polarization has been the norm in Congress throughout most of American history–it was the depolarized era in the mid-20th century that was the aberration–and perhaps it is more realistic to expect that congressional polarization may essentially stabilize at or near current levels for the foreseeable future. Indeed, it could even worsen if Democrats begin to mirror the Republicans’ jump away from the center with the rise of unabashed progressive politicians like New York City Mayor Bill de Blasio and the possibility of greater popular support of European-style social democratic programs.
This is the latest post in our ongoing series on political polarization. The previous posts are listed below. -Dan Hopkins