Using a system like the Cook Partisan Voting Index (PVI), I derive my own PVI which is a little more detailed than the one used by Cook. Instead, I look at the population trends within that congressional district down to growth of traditional blocs that tend to vote Democratic or Republican. For example, average household income and ethnicity as well as union households are used as modifiers against presidential voting behavior. Additionally, I also factor in voter registration figures as well as how candidates from both parties perform within that district. There are, for example, several districts that vote Democratic in presidential elections, but who send Republican congressmen to the House and vice versa for the Republican presidential vote districts sending Democrats to the House. In the former case, when looking at their voting records and how they score on an ideological grid, Republican House members from Democratic presidential voting districts tend to be described as either "rank and file" Republican at best (from the conservative viewpoint) to moderate Republicans (often erroneously labeled "RINO" by conservatives). Meanwhile for Democratic House members from districts that vote Republican in presidential elections, they invariably are described at best as moderate (from the liberal viewpoint) to conservative at worst.
These figures are then compared against a trend over time as to the partisanship of the congressional district. Naturally, redistricting really complicates the picture which is why it is so very important that the GOP build from the local level up and control the redistricting process at the state level every ten years. In the last round of redistricting, Republicans adopted a stance to strengthen incumbent Republican House members. Unfortunately, this had some unintended consequences as it also strengthened Democratic districts. In effect, they jettisoned any hope of capturing these districts in the future. For example, in Pennsylvania, where the GOP controlled the process, approximately 38% of the state's total population lives in Philadelphia and the immediate surrounding counties- Chester, Delaware, Montgomery, Berks, Bucks and Lehigh. Much of the population growth in the surrounding counties is due to the decline in population in Philadelphia itself. These are not conservative voters fleeing Philadelphia for the suburbs, but liberal, Democratic voters infiltrating traditionally Republican area. And, in fact, although they have sent Republicans to Congress in the recent past, the trend has increasingly been towards Democratic votes for president. Hence, one would expect the congressmen representing these districts in these counties, if Republican, to be moderate and, in fact, they are. [Personal note: in a previous entry, someone stated I put too much emphasis on Philadelphia in Pennsylvania politics. One cannot rationally "overestimate" or ignore 38% of a state's population.]
Thus, without further ado, where will seats be gained and lost in 2014? Please be advised that is only a preliminary review and can drastically change between now and November 2014. Thus far, twelve incumbents have announced their retirements from the House- seven Republicans and five Democrats. All five Democrats are running for another elected office and all represent fairly reliable Democratic districts. Five of the seven Republicans are running for another office with a sixth retiring into academia (University of Alabama) and the other one (Michelle Bachmann) is a wild card as to her future. Like the Democrats, all hail from fairly reliable Republican districts. Perhaps the only area for worry- and not much at that- is GA-10 being vacated by Phil Gingrey.
With the open seats as of this point basically safe for the incumbent party, the next place to look at is those districts that trend one way but are represented by someone from the opposition party. Here, it would appear is a problem for the GOP as there are 22 such districts (Republicans representing a Democratic district) as opposed to only seven Democrats representing Republican districts. However, looking at the electoral performance of the incumbent, only two Republicans won in 2012 with less than a five percentage point difference against the Democratic opponent- Mike Coffman in Colorado's 6th District and Mike Grimm in New York's 11th District. However, there are three Democrats who won by less than 5% in 2012- Patrick Murphy (FL-18), Steve McIntyre (NC-7) and Jim Matheson (UT-4). Assuming these districts flip in 2014, that actually would represent a 1 seat gain for the GOP.
The final category is those districts where Republicans represent Republican districts, but either the Republican trend is not strong and/or the incumbent failed to win in 2012 by more than a 5 percentage point difference. Although there are 22 "weak" Republican districts represented by Republicans, only seven failed that 5% margin of victory threshold. They are Steve Southerland (FL-2), Rodney Davis (IL-13), Kerry Bentivolio (MI-11), John Kline (MN-2), Tom Reed (NY-23), Chris Collins (NY-27) and Robert Pettinger (NC-9).
For the Democrats, there are 11 weak Democratic districts. However, of the eleven incumbents, eight failed to reach the 5% margin of victory threshold in 2012. They are Ann Kirkpatrick (AZ-1), Ron Barber (AZ-2), Krysten Sinema (AZ-9), Carol Shea-Porter (NH-1), Ann Kuster (NH-2), Tim Bishop (NY-1), Sean Maloney (NY-18) and Pete Gallego (TX-23). Again, if we assume these seats to flip in a worst case scenario, the Republican Party would pick up another net one seat.
There are three districts that are rated absolutely even, all represented by Republicans. They are Paul Ryan (WI-1), Jamie Herrera-Beutler (WA-2) and Mike Rogers (MI-8). However and again, all won in 2012 by more than a 5 percentage point difference and would appear headed for reelection. This would all translate into a net gain of two seats for the Republican Party as things stand now.
Naturally, there are some districts where watching is warranted. For the Republicans, they are Scott Tipton in Colorado, Daniel Webster in Florida, Justin Amash in Michigan, and James Renacci in Ohio. For the Democrats, the list includes Joe Garcia of Florida, Steve Horsford in Nevada, and Bill Owens in New York. Even assuming these seats flip, it would still be a net gain of one seat for the GOP in the House.
Looking at the three major election prognosticators/handicappers- Sabato, Rothenberg, and Cook- and where they agree on the races of interest, it boils down to 20 contests where there is the greatest chance of a seat flipping parties. One may have it "likely" while another may have it "leaning," but I looked at only the ones where there was some common ground and then assigned a score to where they placed the race to derive an overall chance of the seat changing party hands. In no instance do all three put the district as a toss-up, although the closest they came to consensus was the Arizona 2nd District race (two had it as a toss up. The seats with greatest chance of turnover are the Arizona 2nd (held by a Democrat) and the Colorado 6th (held by a Republican). As mentioned earlier, they have 20 races in common where there is a good chance to some degree of turnover. However, their analysis pretty much mirrors that of this writer- 10 seats are Republican held and 10 seats are Democratic held. The Republican seats, besides the Colorado 6th (Mike Coffman) are: California 21st (David Valadeo), California 31st (Gary Miller), Florida 2nd (Steve Southerland), Illinois 13th (Rodney Davis), Indiana 2nd (Jackie Walorsky), Michigan 1st (Dan Benischek), New York 19th (Chris Gibson), New York 11th (Mike Grimm), and Ohio 14th (David Joyce). My analysis would include the California races, Colorado 6th, Illinois 13th and Ohio 14th.
For the Democrats, the races the professionals rate as having a high likelihood of party turnover are: Arizona 1st (Ann Kirkpatrick) Arizona 2nd Ron Barber), California 36th (Raul Ruiz), California 52nd (Scott Peters) Florida 18th (Patrick Murphy), Illinois 10th (Bill Foster), New Hampshire 1st (Carol Shea-Porter), North Carolina 7th (Steve McIntyre), Texas 23rd (Pete Gallego), and Utah 4th (Jim Matheson). My analysis would include both Arizona races (however, I would also add the Arizona 9th race- Krysten Sinema, a Democrat), the Florida 18th, New Hampshire 1st, North Carolina 7th, and Texas and Utah races. Naturally, many may quibble with these races and have their own favorite to throw in there, or believe that one of these incumbents will survive for whatever reason. And those reasons and explanations may be quite valid. For example, Joe Garcia in the Florida 26th, a Democrat, is being mentioned as vulnerable (in fact, the liberal websites are starting to worry about that race). The more important point is not the specifics of the races, but the trend picked up along the way. That is, everything points to two things: (1) Republicans will keep control of the House come 2015 and (2) whoever picks up a net number of seats, it will not be a very large number of seats. I would estimate at the most a net seat pick up of five either way, but more likely a minimal net two seat Republican pick up in seats.
One very important consideration is that the 114th Congress that is seated in 2015 will be the one that legislates through the all-important 2016 Presidential election. If the GOP keeps the House as expected and if they take the Senate (a real possibility), then they can set the agenda to thwart the liberal one of Obama. In effect, although he is doing a great job of making HIMSELF a lame duck this early in the cycle, he would either be a two-year lame duck, or he can moderate and actually talk to and work with Republicans. It is my hope he would opt for the latter strategy, but being a realist, I doubt he has it in him based on his performance of 1+ term thus far.