Later this month, FairVote will release its complete report on Monopoly Politics 2014, including an in-depth analysis on the decline of the number of congressional representatives from districts that favor the other party and its apparent effect on voting behavior. This analysis applies the same analytical techniques used in the report to a recent votes on a bill on abortion law, and compares it to partisan voting patterns on an analogous bill from 1993.

On June 18, the House approved a bill that would restrict abortions after the first 22 weeks of a pregnancy in a 228-196 vote, almost entirely along partisan lines. Only six Republicans voted against the bill - with two of those opposition votes because they thought it was not restrictive enough (due to exceptions), while just six Democrats voted in favor of the bill.

Many have commented on the increased polarization in Washington, with intra-party homogeneity being especially pronounced in the House, where Members are often more concerned with appealing to a narrow base of supporters than moderates on either side of the aisle. As this vote illustrates, Republicans and Democrats are looking more and more like parliamentary parties, a troubling development in a system that often requires inter-branch and cross-party compromise. We find this most recent vote to be revealing both for demonstrating the degree to which Congress has changed in the past 20 years and what it says about the reasons for that change.

Below, is a breakdown of the votes on this bill, called "The Pain-Capable Unborn Child Protection Act", focusing on the partisanship of the districts represented by the defectors. The numbers demonstrate that representing one of the relatively few remaining "competitive districts" (partisanship between 47-53%) or "crossover districts" - meaning one that favors the other party, - makes a Member of Congress significantly more likely to vote with the other party.  Indeed, of the 10 "defector" votes due to taking a position more like the majority of the other party, only two (both Democrats) are among the 8 Members of Congress representing a non-competitive district leaning toward the other party.

• 6 Republicans defected from the majority of their party to oppose the 2013 Abortion Bill:
  • 3 of the Republican defectors represent competitive districts, making up 10.8% of the total Republicans from competitive districts in the 113^th Congress
  • 1 of the Republican defectors represents a crossover district, making up 10% of the total Republicans from crossover districts in the 113^th Congress
  • *2 of the 6 Republican defectors didn't vote for the bill because it was not conservative enough, meaning that all four of the defectors with a position more sympathetic to pro-choice policies represent either a competitive or crossover district
• 6 Democrats defected from the majority of their party to oppose the 2013 Abortion bill:
  • 4 of the 6 Democratic defectors (67%) represent crossover districts, making up 25% of the total Democrats from crossover districts in the 113^th Congress

20 Years Ago: Abortion Restrictions in 1993

Members of Congress have not always voted along such strict partisan lines, however - even on divisive issues like abortion. For instance, the 1993 House vote on the Hyde Amendment (banning of federal funds to be used on abortions), experienced much more bipartisan support. In total, 99 Democrats crossed over to vote in favor of the amendment, while 16 Republicans opposed the measure. Members representing both competitive and crossover districts were much more likely to defect from the majority of their party.

A similar analysis of the Hyde Amendment vote is provided below. Again, Members were more likely to defect when they represent competitive districts, and even more likely to vote with the majority of the other party when they represented crossover districts. The major difference in voting behavior can be explained by how many more of such districts existed then than there are now - roughly five times as many crossover districts, in fact. 

• 16 Republicans defected from the majority of the party to oppose the Hyde Amendment:
  • 6 of the 16 Republican defectors represented competitive districts, making up 13% of the total Republicans from competitive districts in the 103^rd Congress
  • 4 of the 16 Republican defectors represented crossover districts, making up 14.8% of the total Republicans from crossover districts in the 103^rd Congress
  • Of the 10 votes in 'non-competitive districts', 2 (20%) were in crossover districts - meaning they had to be especially mindful of voters of the opposite party when voting on bills
• 99 Democrats defected from the majority of the party to support the Hyde Amendment:
  • 47 of the 99 Democratic defectors represented competitive districts, making up 54.7% of the total Democrats from competitive districts in the 103^rd Congress
  • 56 of the 99 Democratic defectors represented crossover districts, making up 61.5% of the total Democrats from crossover districts in the 103^rd Congress
  • Of the 53 no votes from Democrats in 'non-competitive districts', 26 (49%) were in crossover districts - meaning they had to be especially mindful of voters of the opposite party when voting on bills

Although the trends in crossover voting in both abortion votes are similar, polarization has increased substantially over the past twenty years. This increase in polarization can be blamed primarily on the decline of the number of districts that support different parties in the presidential and congressional race. Out of the 113 districts represented by a crossover Member in 1993, only 26 remain in 2013. There also has been a sharp decline in competitive districts - form 132 to 52 - due partly to redistricting, but likely more so due to demographic trends.

A few developments have been especially influential the decline of crossover districts. In the past twenty years, Southerners have moved away from voting for Democrats in congressional elections, often opting conservative Republicans instead to replace them. Currently, only 28% of southern districts (from the states of Arizona, Alabama, Florida, Georgia, Kentucky, Louisiana, Mississippi, North Carolina, Oklahoma, South Carolina, Tennessee, Texas, Virginia, and West Virginia) are represented by a Democrat, down from 67% in 1991. Likewise, moderate, urban Republicans have had difficulty winning elections in the Northeast in that same time span. As recently as 1994, Republicans won 8 of the 23 seats (35%) in the New England states - all of which are now held by Democrats.

Many of the most liberal and conservative Members of Congress currently represent uncompetitive districts, that contrary to speculation, are not necessarily the result of partisan gerrymanders. While it is true that redistricting typically makes some of the most vulnerable incumbents safer, a great majority of the safest seats have been created through a demographic trend outlined in Bill Bishop's 2009 book, The Big Sort'. He contends that voters of similar political inclinations are geographically sorting themselves into the same communities, creating extremely homogenous congressional districts, and thusly decreasing the competition in House races.

Therefore, in today's political climate, moderating one's views to appeal to voters in the opposite party actually hurts reelection chances of incumbents. In order to break through this stalemate, what is needed is a higher proportion of candidates who win office authentically representing voters who prefer them to act as genuine bridge builders among the various factions. However, the winner-take-all rules employed in US elections inevitably produce uncompetitive districts that distort national preferences in Congress.  Only doing away with single-member-districts that force politicians to adhere to extreme factions will remedy the situation. 

FairVote's fair representation plan, which combines the current single seat districts into three to five seat 'super districts', would help bring back crossover voting in the House by better representing the political spectrum currently present in the American electorate. Fair voting would pave the way for genuine 'bridge builders' who run bipartisan campaigns and are willing to work across the aisle in Congress.