Over a year before this November’s election, FairVote made its 2014 US House projections as part the biennial Monopoly Politics report. FairVote projects the outcome of 368 of the 435 House races:  205 to Republican Party candidates and 163 to Democratic Party candidates (Table 1). These projections are based on a cautious approach, carefully designed to be 100% accurate, that takes into account only the 2010 and 2012 election results in each district and the presence of an incumbent. Using the same methodology, FairVote will project outcomes for the November 2016 election on November 6, 2014 — just two days after this year’s election.  We expect to see an even greater number of extremely safe seats for each party, despite the nation being relatively evenly split between the two political parties.

Table 1: Monopoly Politics Projections for 2014 US House Elections

Most US House races are endemically noncompetitive and so FairVote does not need to wait for polling data, the results from the challenging party’s primary or campaign spending data to make its projections.  The combination of politically polarized voting (in which ever fewer voters split their tickets) and the geographical “sorting” of voters (so that Democratic voters tend to be concentrated in urban areas and Republican voters in rural areas) means there are few congressional districts in which real two-party competition exists. The vagaries of election season—individual campaign strategies, fundraising and spending, scandals, gaffes and policy proposals—are relevant only in the handful of districts in which underlying partisanship is evenly balanced or, in a declining number of districts, where an incumbent represents a district that leans toward the other party. Indeed, a full year after FairVote’s initial projections, we have updated projections only in a few districts where an incumbent is no longer running and the partisan landscape is, potentially, competitive.

FairVote’s Monopoly Politics projections show Republican candidates assured of winning two seats than they did not win in 2012 (NC-7, UT-4). One Democratic candidate is projected to win a seat (CA-31) won by the Republican candidate in 2012 (Table 2).

Table 2: Net Impact of Monopoly Politics Projections for 2014 US House Elections

FairVote used 2010 and 2012 election results to assess the underlying partisanship of each district (discussed here) and each incumbent’s personal vote (which FairVote terms “performance over average candidate”, an allusion to the incumbent’s value to their party along the lines of the measure “value over replacement player” in baseball).  Using these data, FairVote categorized districts as safe, likely, lean or tossup and projected winners in the vast majority of safe and likely seats (Table 3). FairVote uses the term “projection” (rather than “prediction”) because we assume American voters are evenly split between the Republican and Democratic parties—which, on average, they are. FairVote further assumes that, where one party’s slate of candidates receives a greater than average proportion of the vote nationally, the “swing” toward that party is uniform across all districts in the nation.  As part of the Monopoly Politics report, FairVote explores the impact of different two-party votes in more depth in this 2013 analysis of our projections. An interactive spreadsheet allows readers to apply FairVote’s model to simulate election projections for different two-party votes here.  

Table 3: Monopoly Politics House District Classification and Projection of Winners

In 2012, using nearly the same methodology, FairVote projected winners in 333 (76.5%) races.  We projected these 333 seats with 100% accuracy, even though we did not use any polling data, spending data, or any other information aside from recent election results and whether an incumbent was running in the race.  FairVote expects to be as accurate in these midterm elections, with only calamitous or completely unforeseeable developments, like the arrest of a projected winner or a court decision removing the candidate’s party from the ballot, threatening the accuracy of FairVote’s projections. To reiterate: we can accurately call the results in 85% of congressional districts without needing to refer to the issues, the characteristics of challenging candidates, scandals, campaigning, and fundraising and spending specific to each 2014 race!

Now that we are nearing the end of election season, projections about which party will win each congressional district have come thick and fast. Three prominent election forecasts are the Cook Report, Larry J Sabato’s Crystal Ball (which Politico uses) and Real Clear Politics (RCP). Like FairVote, on October 24, 2014, all three were predicting the Republican Party to gain seats and easily keep control of the House (Table 4). Yet, they differed a little from the Monopoly Politics projections because these forecasts do concern themselves with the minutia of each district’s 2014 horse race. While it is difficult to pin down their exact methodology, it is clear that each uses election results and the presence of incumbents plus a mixture of polling data (particularly important for RCP), characterizations of the relative strengths and weaknesses of the candidates, the mood of the district or state and, perhaps, data on campaign fundraising and spending.

Table 4: Forecasts for 2014 US House Elections (as at 10/24/14)**

For the most part, the forecasts are similar (Table 4), demonstrating once again the lack of competition in most US House districts. Reflecting their use of additional sources of data, the other three election forecasts are updated frequently and, as of October 24, identify more safe seats and are left with fewer tossup seats than FairVote’s projections.

If we home in on the few truly competitive races in 2014, in 11 of FairVote’s 25 tossup seats, some of the other election projections agreed that the outcome is unpredictable (Table 5). Indeed, for 7 of the 25 tossup races (AZ-2, CA-7, CA-52, IL-12, IA-3, MN-8 and NH-1), the election projections were in unanimous agreement that the results will be close. These are, unequivocally, the races to watch.

Table 5: Individual “Tossup” Districts Identified in the Monopoly Politics Projections and the Three Election Forecasts

FairVote’s projections tend to differ with the three forecasts only in those few instances where the particular details of the horse race in a district are so salient as to have overcome the underlying predispositions of the electorate. An example of such a district may be Arkansas’ 2^nd district.  FairVote projects a Republican victory in AR-2, an open seat. This prediction is based on the underlying partisanship of AR-2, which—in the absence of an incumbent—should see a typical Republican candidate receive almost 58% of the vote. (Remember, these predictions were made well before candidates nominated for, campaigned in and eventually won, their primary elections.)

All three of the election forecasts classified AR-2 as a tossup, based on the performance, strengths and weaknesses of the candidate eventually nominated by the Republican Party. The Republican candidate, French Hill, has never held legislative office and is a party insider, having served as an advisor to President George W. Bush and Governor Mike Huckabee. Meanwhile, the Democrat nominee was a long-time Mayor of North Little Rock, Chief Legal Counsel for the Arkansas Secretary of State and a legislator in the Arkansas House of Representatives. Polls show statistically equal levels of support between the two candidates. In this instance, an otherwise safe seat may have been made competitive—although FairVote’s expectation is that the district’s partisan balance will, in the end, lead the district to break toward the Republican candidate.

Similarly, in Nebraska’s 2^nd district, the Republican incumbent, Lee Terry, may have made one too many gaffes—in Sabato’s estimation—to win in 2014, including unashamedly accepting his pay check during 2013’s federal government shutdown.

Instances, like that in AR-2 and NE-2, in which a safe US House district is turned competitive are few and far between—they are the outliers at each election on which we naturally become transfixed. Interestingly, we can use the differing methodologies of FairVote’s projections and the Cook, Sabato and RCP forecasts to (crudely) quantify in how many instances an individual candidate and/or campaign might affect 2014 US House election results.

Only 11 districts are simultaneously identified as tossups in the three election forecasts and not identified as tossups in the Monopoly Politics report. Table 6 presents a summary of why FairVote did not consider each of those 11 districts to be tossups, based on the natural levels of competition in those districts. (It should be noted, than in 9 of these seats FairVote did not project a winner. It is only two of these races—AR-2 and WV-2—that threaten the 100% accuracy of FairVote’s cautious projections.) But, if we think about it, the fact that these districts are not naturally competitive but have been identified by Cook, Sabato and/or RCP as tossup seats means that races in these 11 districts have been made competitive by the particular candidates running in the races.

Table 6: Districts Identified as Tossups in the Forecasts (10/24/2014)**

Conversely, 14 districts were identified by the Monopoly Politics report as potentially competitive races but do not feature in any of the other three election projections. Table 7 presents a list of these districts, with an explanation as to why they ought to be considered naturally competitive districts. We could think about these 14 districts as naturally competitive races that have been made safe by the particular candidates and campaigns run in 2014.

Table 7:Districts Identified as Tossups in the Monopoly Politics Report

If we take the 11 districts made competitive, the 14 districts made safe, and add the 11 seats that both the Monopoly Politics report and some of the other forecasts identify as tossups, we have 36 districts (8.3% of all districts). These are the 36 districts in which the campaigns run in 2014 in those seats could really affect the election outcome. In the remaining 399 districts, individual candidates and individual voters have little power to affect the outcome of the election — even in a nation that is so closely divided.

And so, this analysis shows, there are isolated instances in which the specific candidates presented by each party do affect which party wins that district. Yet, in the vast majority of House races, there is no real competition after primary election day. This endemic lack of competitiveness in congressional elections need not be so. FairVote has a plan to ensure every voter casts their ballot in a meaningfully contested congressional race and has a real chance to elect a preferred candidate. Our fair representation plan would require only a simple statutory change. For more, visit out Monopoly Politics page and read up on our Fair Voting Solution.

* “Other” refers to districts that would otherwise be likely or safe for a particular party, but in which the incumbent was in his or her first term (for example, Chris Collins in NY-27) or in which the incumbent performed so badly in the 2012 election compared to the underlying partisanship of his or her district that we cannot be confident of their victory in 2014 (for example, Dan Benishek in MI-1).

 ** Some forecasts have subsequently revised their projections, but not so much as to significantly alter the analysis presented here.