Now that we know California county participation by party and how it compares to registration we can go about applying it to the congressional races. I’ve used the data two ways to determine November results:
In the first method, I first divided the third party vote between the two candidates 50-50. This is an assumption that anyone can quibble over, but Washington has shown us that the 3rd party vote can go to either candidate. Then I assumed that the race will return to the registered voter difference and adjusted voting accordingly. For example, since Sacramento county had a 7.8% Democratic participation advantage but a 9.6% registration advantage, I adjusted the CA-7 race so that it’s 1.8% closer.
CA-7 will go from 52.7%-41.0% Republican to 55.9%-44.1% Lungren. We then move the district 1.8 points closer and get a 55.0%-45.0% Lungren win. This method is dispassionate. It makes no assumptions and uses a simple formula. You can certainly argue a number of reasons that these numbers are wrong. Maybe one candidate didn’t try that hard in the primary. Or he has a cash advantage and is a strong campaigner. Maybe the district will move more to the left than predicted or maybe less. Maybe the district isn’t quite the same as the county.
Still, this provides us with a number that I’ll say is the center of a range. I believe 9 of these 12 races will be within 3 points of what’s here. Lungren will get 52-58% of the vote.
The second method is a more nuanced approach where I predicted each district’s turnout by county and party and then applied it to the race based on adjustments to share of vote in the district. I tried not to do this with bias and you’ll see many of the numbers are similar. What it does do, however, is account for weak Democratic turn-out in the Central Valley.
I want to emphasize these aren't predictions. In fact, the point here is to remove all subjectivity. This is a conversion of the primary vote to the general election vote based on turn-out. This provides objective data. How other factors will play into the races is up to you.
I’ll go from the most Democratic to the least.
CA-47 – I don’t like this result at all. This is in LA County and it’s conceivable I could become involved in this district. I don’t want to be the guy who predicted Gary DeLong would lose big. Certainly with LA and Orange counties have many congressional districts it’s possible I am wrong. I don’t like the idea of saying DeLong can’t win, so I’ll just make it Likely Democratic at this point.
Method A: Lowenthal 55.3%-44.7%
Method B: Lowenthal 55.9%-44.1%
CA-16 – Costa actually did worse in the general election than in the primary in 2010, something that’s rare with Democrats. Yet there was still a high Republican primary turn-out. He doesn’t have much of a C-O-H advantage either. Method B produces a much closer race, but one that Brian Whelan doesn’t come close. Likely Democratic.
Method A: Costa 58.5%-41.5%
Method B: Costa 53.6%-46.4%
CA-3 – This figures to be a tough climb for Kim Vann, although she actually has more C-O-H than incumbent John Garamendi. Likely Democratic.
Method A: Garamendi 53.9%-46.1%
Method B: Garamendi 53.4%-46.6%
CA-52 – I know IJB won’t like this one, but it’s possible that many of the factors I mentioned earlier work in Brian Bilbray’s favor. Maybe Republican turn-out wasn’t as elevated here as it was in the rest of the county. Bilbray figures to run a better campaign than he did in the primary. Bilbray does have more than 9 times C-O-H right now. Lean Democratic
Method A: Peters 52.9%-47.1%
Method B: Peters 53.0%-47.0%
CA-41 – I’m on record that I don’t think much of Mark Takano as a candidate and I’ll stick with that. The numbers are the numbers, however, and they show a Democratic district with elevated Republican primary turn-out. John Tavaglione does have a bigger cash advantage. Toss-up
Method A: Takano 50.9%-49.1%
Method B: Takano 51.0%-49.0%
CA-26 – It’s hard to bet against Republican Tony Strickland, since he has (cue Dr. Evil) $1 million in the bank, more than three times what Julia Brownley has. He is one of the best challengers the Republicans have this cycle. This district is tougher to predict than the ones above, because Linda Parks got so many votes. I think she stole a lot more Democratic votes than Republican ones, based on how Brownley and Strickland did. Toss-up
Method A: Brownley 51.1%-48.9%
Method B: Brownley 50.6%-49.4%
CA-24 – Lois Capps is the only Democrat in a competitive district who has a significant C-O-H advantage right now. Her opponent Abel Maldonado has proven himself to be a strong fundraiser and he has augmented his total with some of his own money. The Republican primary advantage was small, but it should keep this race very close. This is the closest race according to the model, but it’s also one of 4 that are Toss-up, so I won’t predict it’ll be the closest.
Method A: Maldonado 50.2%-49.8%
Method B: Maldonado 50.3%-49.7%
CA-9 – The popular meme is that since Jerry McNerney won in 2010 and his district became slightly more Democratic then he was a shoe-in. He lost the primary by 4 points and the Republican turn-out wasn’t all that remarkable. It’s tough to bet on someone as young as Ricky Gill but he did well in the primary. Another toss-up.
Method A: Gill 50.4%-49.6%
Method B: Gill 50.9%-49.1%
CA-36 – The McNerney meme is also applied to Mary Bono Mack, a candidate who overperformed her district repeatedly. Her 2008 margin over John McCain’s vote total was the largest of any Republican in a competitive district. And she won by 16 points in the primary. My gut says she’ll coast to re-election but Republican turn-out in Riverside county far outpaced Democratic. This wasn’t a situation of high Republican turn-out. In fact, Riverside had one of the lowest Republican turn-outs in the state. Democratic turn-out was pathetic, just as it was in Los Angeles and Orange counties. The GOP has room to grow here, but the Democrats have a ton of room to grow. My gut tells me Likely Republican or even Safe, but the numbers say Lean Republican.
Method A: Bono Mack 52.7%-47.3%
Method B: Bono Mack 52.6%-47.4%
CA-7 – Another district where the narrative doesn’t match up with the results. Bera did very well against Lungren in 2010 and this district is slightly more Obama also. In Sacramento county Republican turn-out was only slightly higher than Democratic turn-out. Bera has room to grow but Lungren has about the same room. The 11.7 point primary deficit should be too much for Bera to overcome. This is a case, however, where the Method B helps the Democrats and indicates that Bera may have a shot. Likely Republican
Method A: Lungren 55.0%-45.0%
Method B: : Lungren 52.8%-47.2%
CA-21 – This is the district Method B was invented for. Republican turn-out was excessive, just as it was in these counties in 2010. Of course in 2010 the Republicans actually did much better in the general than in the primary, the exact opposite of what you’d expect. It appears that Democrats in this district, who are heavily Hispanic, just don’t vote. So it’d be a shock if election day turn-out approaches registration.
The surprise second place finisher, John Hernandez, raised and spent almost nothing on the primary and had a negative C-O-H in May. I understand how you can have debt, but I don’t get how you can get a negative C-O-H. He has yet to report June 30 fundraising, but he’s up against a good candidate. The only way he wins is if Democrats suddenly start turning out here. Likely Republican and that’s only because of registration.
Method A: Valadao 51.2%-48.8%
Method B: : Valadao 54.7%-45.3%
CA-10 Did anyone on earth expect that Jose Hernandez was going to win here? Yes, that was an astronaut joke. Hernandez lost by 21 points in the primary, with Chad Condit and his meager budget taking 15% of the vote. Jeff Denham, on the other hand, is an experienced politician who has run close races before and is a prolific fundraiser. I think it’s Safe Republican, but if I’m not giving Alan Lowenthal safe I won’t here. Likely Republican
Method A: Denham 55.5%-44.5%
Method B: : Denham 55.5%-44.5%
No comments:
Post a Comment