Many political observers have lamented the increasing partisanship and dysfunction of Congress. Polarization and the vitriolic politics that it generates have been especially pronounced in the House of Representatives, where the common ground between the parties has vanished as the contingent of independent-minded centrists has disappeared at a remarkable rate in recent decades - from 192 such Representatives in 1972, to 129 in 1992, to just 12 in 2012.
These centrists have not necessarily shared a common ideology; rather, they share a willingness to take some positions that are out of step with their party and, as a result, have the ability to forge the compromises necessary for effective policymaking in our constitutional system of checks and balances. The absence of such moderates willing to work across the aisle is perhaps the most powerful example of how unrepresentative our winner-take-all system of elections has become, as an upcoming FairVote analysis on the disparities between the ideological makeup of the U.S. electorate and the U.S. House will explain.
The charts above depict the evolution of the ideological distribution of Representatives in the U.S. House based on their individual "DW-NOMINATE" scores, dramatically illustrating the degree of polarization that has occurred as both major parties have shifted further and further away from the center over the last forty years.
The DW-NOMINATE metric has become the gold standard in the world of political science for evaluating the ideology of legislative bodies, parties, and individual politicians. Developed by Keith Poole and Howard Rosenthal, DW-NOMINATE maps the ideological positions of legislators by identifying coalitions - determining which legislators vote together and how consistently. It assumes that the strictest ideologues in each party will be those least likely to vote across party lines, and places every legislator on a scale from -1 (the most liberal score possible) to 1 (the conservative extreme).
The chart for the House in the 92nd Congress shows two clear poles, representing the mainstream of the two major parties, but also shows a significant contingent of moderates bridging the gap between them. There was, and had been for decades, ideological overlap between members of the two parties, with many Democrats voting more conservatively than many Republicans. The data for the 102nd Congress shows that a shift has occurred, with the ideological centers of the two parties moving further apart, and a cleft opening between them as the number of moderates declined. By the end of the 112th Congress in January 2013, this cleft had become a wide gulf, with significant ideological distance between the most conservative Democrats and most liberal Republicans.
While increasingly partisan primary voters have been a primary driver of this change, the significant number of moderate voters that remain have also become progressively more underrepresented. The representation of these moderate voices in Congress is critical for the success of a government that relies on compromise and bridge-building for its effective operation.
The table below presents the ideological distribution of the U.S. House of Representatives in the 92nd, 102nd, and 112th Congresses, based on the adjusted DW-NOMINATE scores of individual members.
The decline in the number of moderate representatives over this period is remarkable. In the U.S. House during the 92nd Congress, 198 members (45.1%) could be classified as moderates, while just 52 (11.8%) were strong liberals or strong conservatives. By 2012, the number of moderates in Congress had fallen from 198 to just 12 (2.7%), and the number of strong liberals and conservatives had skyrocketed, increasing by nearly five times to 240 (54.3%).
Poole and Rosenthal describe any legislator whose score falls between .25 and -.25 as a moderate, with the center point, zero, effectively representing the all-time ideological midpoint of congressional politics. While this is useful for describing Members of Congress in a historical ideological context, it is less helpful for identifying the representatives in the true center of any given Congress. As American politics has evolved over the last four decades, the ideological midpoints of recent Congresses have shifted away from the midpoint as seen from a broader historical perspective. As a result, some legislators who would be described as moderates under Poole and Rosenthal's definition will likely be seen as firmly on the left by many modern observers, especially when comparing them to a Republican party whose representatives have tracked further and further to the right.
For the purposes of this analysis I have attempted to address this problem by amending the scores slightly, with a figure that represents the public perception of the ideological midpoint of congressional politics in a given year. For each congress, the ideological center of the House can be understood as the midpoint between the average Democratic representative's score and the average Republican representative's score. Since public perception of the ideological midpoint will lag behind changes in congress, this midpoint value is averaged with the midpoint values of Houses in the preceding 20 years. This 20 year midpoint value is then added to the scores for each representative. The adjustment values applied to each Congress are listed in the first column of the above table.
This adjustment does not dramatically alter the classification of representatives, but does bring them more in line with an understanding of the location of the political center that is more appropriate for the time period in question.
Where Have They Gone?
Moderate representatives come most often from districts with relatively even levels of partisanship, and are especially likely to come from districts in which their party has a slight electoral disadvantage. These "crossover representatives," defined as members that represent a district in which the opposite party received more than 50% of the two-party vote for president relative to the national average, can often be counted on as bridge-builders in Congress. They are more likely to work across party lines and engage in "crossover voting," as their reelection is dependent on gaining support from voters on both sides of the ideological divide. Of the twelve moderates in the House of Representatives in the 112th Congress, seven were crossover representatives.
The disappearance of moderates in the House of Representatives is in large part a function of the declining number of these crossover representatives. An upcoming FairVote analysis of crossover voting, part of the Monopoly Politics 2014 report, finds that the number of crossover representatives in the House has fallen from 116 in 1992, to 44 in 2012, to just 26 today, largely because of the decline in the number of competitive congressional districts.
The decreasing number of competitive districts also contributes to the vanishing of moderate representatives by allowing increasingly partisan primary voters to become the de facto electorate in many districts. This phenomenon is often cited as a source of the rightward shift of Republicans in the U.S. House.
Though primary voters have played a role in pushing the parties away from one another, they do not represent the American electorate as a whole. FairVote analysis of Pew polling data indicates that the 2.7% of representatives with moderate ideology represent an estimated 19% of Americans with similar views.
The decline of centrist politics in the United States is self-perpetuating. As the number of moderates in Congress decreases and parties vote in lockstep, voters increasingly identify more with parties than with individual candidates, and are even less likely to elect new moderates. The inevitable result is a gridlocked Congress operating from two partisan poles rather than a pragmatic center, and fewer and fewer competitive congressional races in which the voters can make a real difference.
Restoring the Voice of Moderates
The best way out of this vicious cycle is structural reform of U.S. House elections, as our winner-take all electoral system creates few incentives for interparty cooperation.
One potential reform is the modification of California's top two primary system into a top four primary. All candidates would appear on the primary ballot, and the top four candidates would advance to the November election, regardless of party affiliation. In November, ranked choice voting would avoid concerns about vote-splitting and encourage candidates to reach out to supporters of other contenders and other viewpoints in order to secure a majority.
While such a top four primary can be introduced in states right now, Congress would need to pass a statute to enable the best solution: adoption of the choice voting form of fair representation voting. This reform calls for existing districts to be combined into three or five seat "super-districts" in which voters would use fair voting methods to elect their representatives. Each super-district would be likely to elect at least one moderate candidate to Congress in every election cycle. FairVote projects that the adoption of fair voting for all U.S. House races would result in the election of over 60 moderates in the current climate, restoring the voice of centrists in American politics.