Virginia’s June 10th primary elections were supposed to a quiet affair. They mostly were, with most major party nominations already having been decided by conventions and privately-administered “firehouse primaries” and Democrat Don Beyer surging to a comfortable primary win in the heavily Democratic 8th congressional district being vacated by Jim Moran. But of course there was another contest: the historic upset of House majority leader Eric Cantor in the Richmond suburbs.
Cantor’s shocking defeat doesn’t change the fact in the general election ten House races are effectively settled, including a projected comfortable win for Cantor’s vanquisher, Dave Brat. Democrats have a shot at picking up just one seat to narrow the current 8-3 Republican advantage.
I am currently an intern at FairVote, an organization that advocates for electoral reform, and I thought I would take some time to share our projections and proposed reforms for states given that there has been relatively little attention to FairVote’s analysis of congressional elections and explanation of how Democrats are disadvantaged by the electoral system. All race projections are from our Monopoly Politics 2014 report, which projected 333 races last cycle, each one correct. For this election, our updated report projects 368 winners, although most of the remaining 67 races have a clear lean as well. Given that we rely on relative presidential outcomes in districts and relative performances of House candidates, we’ll be releasing our 2016 House election projections on November 6.
To be clear, two days after the 2014 elections and two years before the 2016 elections, we will be able to project almost 400 house races, only changing our predictions if a seat opens up. This is not a testament to our forecasting abilities, but rather a demonstration of how flawed our democracy truly is.
Virginia – The Landscape
Virginia’s congressional delegation has grown increasingly Republican over the last two decades. What was once a 7-4 split in favor of the Democrats (1992) has become the 8-3 Republican majority we see today. Given that only one seat is truly in play this cycle, this majority will not be going anywhere in the near future.
Virginia does have two other relatively balanced districts in terms of partisanship, but both have solid incumbents who are expected to win by more than 10 points. One to keep an eye on is VA-2, where Republican Scott Rigell represents a 51% Republican district, making him the most at-risk incumbent in the state. It’s no coincidence that Rigell was the only Member of Congress to vote against tying defunding Obamacare to the government shutdown. That said, given that he won by 8% in the Democratic year of 2012, we project he will retain his seat
Methodology
FairVote’s projections are grounded in two factors:
1. The partisanship of a district. This is determined by the performance of major party candidates in a presidential race relative to their performance nationally. FairVote developed this partisanship measurement in July 1997 in its first Monopoly Politics report; later, it was adapted by Charlie Cook with his partisan voting index, but FairVote still finds that using a single presidential election provides a very reliable guide to the underlying partisan balance in a district or a state.
2. Performance Over Average Candidate (POAC). This is determined by how a candidate fares relative to what we would expect for an average candidate of the same party and incumbency status (see FairVote’s discussion of the “incumbency bump”) given the district’s partisanship and the national partisan balance.
The incumbent with the strongest POAC score is retiring Republican Frank Wolf in VA-10, who had a +5.5 POAC that boosted his winning margin by 11%. The weakest incumbent was Gerry Connolly in VA-11 with a -8.5% POAC, although he still won easily. You can see our full Virginia report here.
Redistricting and Election Competitiveness
Republicans narrowly gained effective control of the State Senate in 2011, giving them control of the redistricting process. The state legislature passed its new congressional redistricting map on January 21, 2012. Not a single Senate Democrat voted in favor of the map. Democrats rejected the map in part because it did not create a second majority-minority district. The map was subjected to multiple lawsuits, but courts upheld the map’s validity in every case, and on March 28, 2012 it was pre-cleared by the Department of Justice. In a state that is split 50/50 Republicans/Democrats, Republicans have five districts that are least 55% Republicans and three more Republican-leaning districts with Republican incumbents.
Virginia’s swing state status in the presidential race contributed to relatively high voter turnout in 2012, with a voter turnout rate of 64.4% and a corresponding relatively high percentage of eligible voters who voted for winning candidates. Only one House race was won by less than 10% in 2012.
2014 Projections
We project districts 1, 6, 7, and 9 as safe for Republicans and districts 3 and 8 as safe for Democrats (despite District 8 being an open seat). Districts 2 (Rigell), 4 (Forbes), and 5 (Hurt) are projected as Republican victories as well in more competitive districts. Rigell and Forbes have districts that are 49% and 47% Democratic, respectively, but their superior POAC combined with their slight partisan advantage means that they would need an unlikely personal controversy or an equally unlikely Democratic wave to remove them from office. Barring any major surprises, these 10 districts should all remain in the hands of their current party.
The one district that is a toss-up is district 10. This race will determine who replaces long-serving incumbent Frank Wolf. Virginia Delegate Barbara Comstock (R) has already won the Republican primary (which was a party-run “firehouse primary”), and Fairfax County Supervisor John Foust (D) has won his party’s nomination by default. The district covers a portion of Northern Virginia, including Manassas, Winchester, and part of McLean.
Other projection outfits tend to have similar projections for Virginia this cycle. However, we have not projected a winner for VA-10 while both Cook Political Report and RollCall both have called it “lean Republican”. This trend is present across our projections: there are races that we choose not to project that other models do. The reasoning for this conservatism is that other models include candidate analysis, whereas we are only concerned with structural factors – including how Virginia-10 is only a 52.5% Republican district and hardly out of range for Democrats in an open seat contest. In essence, we can project more than 80% of races with anticipated 100% accuracy without knowing anything about an incumbent besides his/her party and the number of years he/she has been in office. The model, then, serves as a reminder that our current system is not working for Virginia, or America. Only structural reforms can bring about the necessary change.
What Virginia Needs
Virginia obviously could have had a better redistricting map for Democrats if they had drawn it. They probably would have had a fairer map if an independent commission had drawn it as well.
But the root of the problem for Democrats goes much deeper. As people continue to sort themselves into likeminded communities, urban areas are becoming more liberal and less urban areas becoming more conservative. This process means that the normal solution (independent redistricting commissions) is unlikely to be the silver bullet that its proponents want it to be.
Independent redistricting can make districts more competitive, but only at the expense of geographical compactness. In other words, since most Democrats are concentrated in cities, creating even districts would require snake-like districts that cut into portions of urban areas and slither out into the country side (sound familiar?). Since the entire point of representative democracy is to, well, represent a group of people, this solution only seems like a slight improvement.
The best answer to problems with Virginia districts –and the country as a whole -- is to create larger “super districts” that each elect 3 or 5 Representatives. FairVote has developed a national plan to do that, as it shows with this interactive flash map analyzing districts as they are and as they could be. We recommend that candidates be elected by ranked choice voting (RCV), where voters rank candidates. In multi-seat districts, the percentage of the vote declines with the number of seats – it is just over 25% of the vote in a three-seat district and about 17% seat districts. Done nationally, it would completely remove the incredible partisan bias we now see in U. S. house races, where Democrats are unlikely to retake the House without surpassing 55% of the vote, as we explain in this analysis.
In contrast to the current plan, our Fair Voting plan projects an even playing field: likely wins for five Democrats and five Republicans, with one toss-up seat. Those projections depend on voter participation, though, so all voters in every district would be urged to participate. We’d likely see greater diversity within the parties, as well as the ability for independents and minor party candidates to hold the major parties accountable.
Adopting a plan that actively creates competition (and allows both parties to earn their share of seats) is the best way forward. All it requires is a statutory change from the US Congress. While that might not seem feasible now, it could very well happen in the future – especially if we spend more time contrasting the problematic plans we have now compared with making every voter count in every election.