Sunday, October 31, 2010

Final House Predictions

Am I better at this than the experts? At least I'll commit to each race. No toss-ups. You're always right (and wrong) if you pick it a toss-up.

Democratic to Republican (59 seats)
Arizona 1
Arizona 5
Arkansas 1
Arkansas 2
California 11
California 20
Colorado 3
Colorado 4
Florida 2
Florida 22
Florida 24
Florida 8
Georgia 2
Georgia 8
Illinois 11
Illinois 14
Illinois 17
Indiana 8
Indiana 9
Kansas 3
Louisiana 3
Maine 1
Maryland 1
Michigan 1
Mississippi 1
Nevada 3
New Hampshire 1
New Jersey 3
New Mexico 2
New York 19
New York 20
New York 29
North Carolina 2
North Carolina 8
North Dakota
Ohio 1
Ohio 15
Ohio 16
Ohio 18
Oregon 5
Pennsylvania 10
Pennsylvania 11
Pennsylvania 3
Pennsylvania 7
Pennsylvania 8
South Carolina 5
South Dakota
Tennessee 4
Tennessee 6
Tennessee 8
Texas 17
Texas 23
Virginia 11
Virginia 2
Virginia 5
Washington 3
West Virginia 1
Wisconsin 7
Wisconsin 8

Republican to Democratic (4 seats)
Hawaii 1
Illinois 10
Louisiana 2

Net: 55 seats
Republicans: 234
Democrats: 201

The following is a list of 20 seats that'll flip is it's a big tsunami. I'm not hedging my bet. The picks are the picks. But if 5 of my next 10 flip I think I deserve credit. If it's #5, #11, and #19 I don't.

1. Mississippi 4
2. New Hampshire 2
3. Colorado 7
4. Arizona 8
5. Michigan 7
6. Connecticut 4
7. Connecticut 5
8. Alabama 2
9. Massachusetts 10
10. New York 23
11. Washington 9
12. Michigan 9
13. Ohio 6
14. Illinois 8
15. Virginia 09
16. Mississippi 2
17. Minnesota 1
18. Minnesota 8
19. New York 24
20. North Carolina 7

That'd be a 79 seat gain. It's possible Republicans could pick up seats not on this 79. All the forecasters have more seats than I do that are "Lean Democratic" or better. Real Clear Politics lists 135 Democratic seats potentially in play. I'll stick with my list.

Wednesday, October 27, 2010

Chaffetz Gets It

I want to give an enormous shout out to Jason Chaffetz. I had the pleasure of meeting Jason before he was elected and again after he made it to Washington. I find him earnest and principled. There is a sentiment from the base that "you don't work with Democrats. You run them over." Of course this is the attitude the Democrats had with Republicans. And we see how well that worked for them.

Democrats blame Republicans for obstructionism, when there were few pieces of legislation the Democrats were actually prevented from passing. The most glaring one, climate change, wasn't even obstructed by the Republicans. The Democrats never introduced the bill in the senate. When the Democrats wanted to pass a bill it wasn't that hard to pick off a Scott Brown, Olympia Snowe, George Voinovich, Susan Collins, or Kit Bond. The Democrats wrote the legislation in the back room and then were surprised when it was their own party that obstructed the bill.

No matter how many Republicans get elected Tuesday there will be over 200 Democrats in the House and Senate. All of them will have been elected too. We won't win on Tuesday. We'll win more than we lose. If you want to win independents you have to work with the other party. Republicans need to lead, but that doesn't mean exclude the other guys. The Democrats might have a few meaningful things to add to a bill.

The 90's were a good time because we had a smart effective Republican congress that worked with a Democratic President. America kept the Republicans in the majority until they forgot that's what they were supposed to do. So let's introduce bills in committee, have debate, and hearings. Let's listen to both sides, give a little and get a little. I know we can make Republicans bills that can get Democratic votes and Obama will sign. Sure, we should push Obama and make him veto some legislation, but only after we've got some wins to show America what Republicans can do.

How The House Breaks

Nate Silver suggests it's possible that races could break two points for Democrats and they might not lose the House. A week ago I looked at the results of the expert predictions, showing when the pollsters predict a toss-up the races split and don't go one way or the other.

Since Monday independent pollsters have released 17 House races polled within 4 points. These races could go either way. In 2006 and 2008 Republicans won 31 races by 5 points or less. Democrats won 34. It's certainly possible that where that center is could move a point or two in one direction or another, but it's likely that the two sides will split the toss-ups.

A simple way to look at the forecasts would be to say Republicans win all the races that the forecasters have leaning GOP, lose all the races that are leaning Democratic and split the toss-ups. If that were to happen, Larry Sabato, Stu Rothenberg, Charlie Cook, and CQ Politics would all have 41-43 Republican gains. Real Clear Politics would have 62. Of course you should expect the Republicans will win a few Lean Democratic seats and the Democrats will win a few Lean Republican. That'd make Republican gains bigger. Right now I see a Republican 233 Democratic 202 split based on my formula. Nate Silver has 231-204. While I really think it'll be 225-227 I'm probably being cautious.

Senate Predictions

I can make most of these now, because races don't change in the last few days.

South Dakota
South Carolina
North Carolina
New Hampshire
North Dakota

West Virginia
New York*
New York

That'd be 51-48 Democrats. The one race I have yet to pick is Colorado, but I'm leaning Republican. Alaska will be one of the two Republicans. I'm leaning Miller, but I reserve the right to change that one. Don't take this prediction as contradicting my post yesterday. I'm picking Boxer by the slimmest of margins and I know my California friends are going to do a big GOTV so I'll be proven wrong.

The Campaign Trail

As I mentioned earlier, the Delaware Valley is a huge battleground in House and Senate races. Starting Friday my camera crew and I will be on the campaign trail with a bunch of those campaigns documenting the final hours before the elections. I'm excited to get it. i hope to be able to report back our experiences each day.

These will be the final scenes in Where's The Party?

I will have final House and Senate predictions even if I can't post much.

Tuesday, October 26, 2010

Good News for Fiorina

PPP, a Democratic polling firm, has come out with a poll where Carly Fiorina is losing by 9 points. Clearly, she's done, right? Wrong. This is a great poll for Fiorina. If you look closely you'll see that 26% of this year's voters have already voted and they favor Boxer by only 48%-46%. That's a significant amount of voters, so the 2 point difference is very meaningful and likely highly indicative.

PPP has a 47% Democratic electorate and a 13 point differential between Democrats and Republicans. Democrats were 40%, 41%, and 42% in 2004, 2006, and 2008. The differential was 7%, 6%, and 12%. Once again PPP's electorate is way more Democratic than even California can produce.

Extrapolating their responses by party, the people who have already voted could be 40% Democratic 33% Republican 27% independent, the 2004 electorate or 41% Democratic 35% Republican 24% independent, the 2006 electorate. Both would produce the same results. If Fiorina gets either the 2004 or 2006 electorate and wins independents by 17 points, it's a toss up. It's certainly possible the electorate won't be like 2004 and 2006 and be more like 2008, but since the voting so far, with 26% already voting, has been 2004/2006.

The polls in the last two weeks have ranged from Fiorina +3 to this one with Boxer +9. I'm telling Carly and California Republicans not to give up. We're really neck and neck.

Things to Look For Besides The Results

Obviously who wins and loses is what's most important, but there are a number of things that could be telling about the future.

In 2008 House Democrats won the 18-29 vote 65%-35%. This was significant, but below the 67%-33% that Obama got. Do the Democrats have this group or was it just Obama in 2008? If Republicans do better, Democrats will argue that this is an off year election and a Republican wave. Because it's an off year election less young people will vote, but what we want to know is what the ones that do vote do. Certainly Republicans will do better than they did in 2008. The Democratic argument is that they own the youth vote. If Republicans do well with 18-29 that'll show that they won't win this demographic under all circumstances.

So it be important to look at how the youth vote indexes against the total percentage. This eliminates that it's a wave election because it compares the percentage of 18-29, 65%, to the overall average of 55.6%. The Democratic index was 117, well below Obama's 126. That may show that Obama really inflated the number.

House Democrats won the Latino vote 70%-30%, just as they did in 2006. The Latino indices were 126 and 128 in those years. In 2004 the Democrats only too 55% of the Latino vote, a 114 index. That was a year Bush brought in a lot of Latinos. This year will show whether the GOP is capable of winning Latinos.

People might think the Black vote would be irrelevant to Republicans, but they've traditionally garnered around 10% of the Black vote. It was 5% in 2008. The is a significant amount of votes and could be a big deal in districts that aren't majority Black, but there is a presence. If Republicans are at 5% again that'll show that Blacks are a lost cause until Obama leaves office or that the GOP has lost the few Blacks it had due to Obama opposition.

I'll be watching the tea party candidates. The argument against them is that they won't win. The argument for them is that they are true conservatives, not RINOs. I may not prefer some of the candidates, character counts with me, but every dog should have his day. I'm tired of hearing that conservatives can't win somewhere. We'll test it and see. If they don't win this year we'll know that the state/district is one where only a more moderate candidate can win. We know about the ones in the Senate (e.g. Miller, Angle), but there are a number in House races. Keith Rothfus in PA-4, Anna Little in NJ-6 are two I can think of.

I'll be watching the fake tea party candidates. Democrats have inserted these people in PA-7, NJ-3, FL-8, FL-12, and the Nevada Senate. If conservatives can be fooled into voting for these people the Democrats have a winning strategy of fraud that they'll expand to hundreds of races in 2012.

Monday, October 25, 2010

The Left This Year

I was talking with a politically aware Democrat today who moved a few months ago. He mentioned that he didn't want to go vote in his old neighborhood and needed to re-register. I don't think you can change your registration this late. I don't think he'll drive a half hour just to vote. I think this election just hasn't been important enough for him to think about it. Could more Democrats be that way?

Sunday, October 24, 2010

House Ratings

I'm not projecting Republicans 229.5 Democrats 205.5. I have 9 Democratic districts as Safe Republican. A lot of people don't do this, but if you're predicting the Republicans will pick up 40-50 you have some locks that'll be clear weeks before. Likewise, I put Delaware firmly into the Democratic column.

Real Clear Politics is aggressive in their predictions. They now have 133 Democratic districts in play, with 108 Lean Democratic or better. They're including any district that there's a hint of a competitive race. The formula only puts them at 236-199 Republican, however. 538 is run by Democrats and they list a 230-205 split.

The formula says that Larry Sabato is predicting a 218.5-216.5 Republican advantage. He's already declared a net of +47 for the GOP (226-209), so expect him to move some to the right. Stu Rothenberg is saying Republicans will be +45 (224) to +55 (234). The formula puts him at 45. Cook says Republicans +52 (231), but they are 5 shy of that. The formula puts them at 226. CQ hasn't given a number, but the formula puts them at Republicans 222-213. When all the experts predict that the floor is flipping the House, you can be pretty sure it'll happen.

Saturday, October 23, 2010

It's a Black Thing

Republicans believe in treating everyone equally. The best person should be hired for the job, regardless of color, even if that means hiring 5 African-Americans or no African-Americans.

As a result, Democrats have won an overwhelming share of the African-American vote for the last 40 years, with Republicans resigned to only getting a small percentage of the votes. There have been a few African-American Republican congressmen, notably J.C. Watts and Gary Franks, in congress. Of course, Colin Powell, Michael Steele, Condaleeza Rice, and Clarence Thomas have had even more prominent governmental roles. The GOP hasn't really bothered to reach out.

This year there are 11 Black Republican candidates for the House of Representatives. African-American majority districts elect Democrats. The Black candidates Republicans have run in the past have just been fodder for the Democrats.This year, Republicans are running 6 Black candidates in majority minority districts.

It's the other 5 that I want to note. Tim Scott is running in a district that's 21% Black. Yet it's so Republican that Scott is considered a lock for the seat. Allen West and Ryan Frazier are in very competitive races in districts with few Blacks. Since Blacks vote heavily Democratic, that might be why they have good shots.

Bill Randall and Charles Lollar are running in more Democratic districts, but not ones that are so heavily Democratic that a Republican couldn't win. Even though both have raised a decent amount of money, neither has registered on the national radar. Randall is running a district that's 27% Black, while Lollar is running in one where Blacks are 30% of the population. Both are running against white opponents. Lollar is running against House Majority Leader Steny Hoyer. He, of course, has great clout, but this is a year when having great clout might not be such an asset. I don't expect either candidate to win, but I'm interested in seeing how well they do with the Black community. Can a Black Republican win a decent share of Black voters this year? Or will they get the same 8-12% most Republicans do?

Regardless, there'll be at least one Black Republican in congress next year. Maybe more. Of course, as Republicans, we don't care about the color of a candidate's skin, only if he's the best man for the job. If we celebrate Black Republicans getting elected aren't we hypocrites?

Friday, October 22, 2010

Don't Believe Everything You Read

Maybe you can't figure this one out because it has me stumped. Stu Rothenberg has MA-9 in the toss-up column after not listing it before. Scott Brown won this district, but he did better in 2, 3, 5, and 6. Rothenberg lists none of them. The MA-9 candidate has no money. There's nothing on line about this race. How could this district be in danger? (Edit: Rothenberg has removed MA-9 and moved MA-10 from Lean Democrat to Toss-up.)

Republicans are releasing all sorts of polls showing obscure candidates in tight races, like the one for Washington-6 today.

Internal polls are always suspect, but seeing one for a race that hasn't been polled independently is really suspect. I can't see how either of these districts will flip. Republicans appear to be flooding the zone with competitive polls in non-competitive districts. This is brilliant. Democrats are running around like chickens with their heads cut off. You have to defend Washington-6, don't you? It shouldn't take that much to save it, right? What if you don't and he loses? Democratic resources will be stretched thin.

At his point, any independent poll where the candidate is down means he'll likely lose. We're so close to the election and there's been so much early voting that making up ground now is extremely difficult. Any internal poll that shows a candidate down is really suspect. If the candidate releases a poll where he's down 4, that means he's down a lot more. Maybe he can make it closer, but you don't get a prize for losing 53-47.

If you see three Senate polls where the candidate is up 6-8, don't buy one poll that has him up 2. One poll is an outlier.

If you can find out how many people qualify on a likely voter screen, take that into account. If a poll lets 90% of respondents through, you have to question it. Mid-terms get 53-60% of registered voters. They get 35-40% of voter age people. Assuming that likely voters are more likely to want to answer questions you should be able to get rid of at least 25-30% of the people polled. Since Democrats are less enthusiastic, they are likely including more Democrats than they should.

Pollster Skew

All year Democrats have claimed that Rasmussen polls were bogus, pretty much because they favored Republicans and Scott Rasmussen was a Republican pollster. They posited that the reason Rasmussen's polls were so accurate was that they changed their polling at the last minute to get it right.

I know that as a Democrat if you read something in a liberal blog that makes it true, but I decided to do research to see if things were true. Rasmussen isn't employed by any Republican campaign. He might (or might not) be a Republican, but I'm pretty sure all pollsters vote for someone. That doesn't make them biased or employed by anyone. Scott Rasmussen may have done some polling for George Bush eight year ago, but a few polls back then doesn't put him n anyone's payroll. Being on the payroll does.

Rasmussen has no Republican bias. Their polls skewed Republican, sometimes heavily, earlier this year. But that was when they were using "likely voter" and everyone else was using "registered voter." You can question the wisdom of doing that, but comparing the two is apples and oranges. We've found that likely voter is 4% more Republican than registered voter this year, so it's difficult to say that Rasmussen was favoring anyone.

So I decided to look at each pollster since they all went to "likely voter" after Labor Day. I only included 7 pollsters, as I wanted enough polls to make a determination. As it is, Reuters/Ipsos and Mason-Dixon are a little light with the polls. I divided it into three periods, from right after Labor Day until now to see if Rasmussen changed their polling. The problem is that everyone isn't polling the same races. Anyone polling more Democratic leading races will come out more Democratic than someone polling Republican leaning races.

So I compared two pollsters polling the same race within a week or less and then determined how they were overall to each other. The problem here is that Rasmussen and CNN/Time could average the same for all their common races, but Rasmussen might be 2 points more Democratic than SurveyUSA in their common races, but CNN/Time might be 4 points more Democratic. So I averaged out.

Since we don't know yet who's accurate we can't determine how Republican and Democratic they actually are, but we can determine them relative to each other. Since Rasmussen was in the middle every period I used them as the base.

Since everyone went to "likely voter" models, PPP has been much more Democratic than all the other pollsters. Sometimes they'll be similar and sometimes they'll be 7 points more Democratic. These are averaged out. Quinnipiac, Reuters, and SurveyUSA have been 2-4 points more Republican than Rasmussen. Surely, all those independent pollster can't be in the bag for the GOP.

Since Democrats know PPP is right, I'll show the comparison with that.

This could be 100% true, as we don't know yet. Democrats can save this image and use it as justification that everyone else has their finger on the scale.

Thursday, October 21, 2010

Generic Ballot

The generic ballot is moving in the Democrats favor! Before they go about celebrating, that's not really good news. Of the 9 pollsters who surveyed likely voters last month and this month, Republicans dropped from an +8.1 to +7.4. That's not much of a gain. This close to the election that's not enough of a shift to make a difference. If you include the three pollsters who switched from registered to likely the average is +7.7. In 1994 Republicans won the House vote by 7 points and got 231 seats. In 2002 they won by 5 points and got 227 seats. In 1946 they won the vote by 9 points and got 246 seats. In 1996 Republicans got 49.8% of the vote and retained the majority. That was the only time a party won less than 50% of the vote and got the majority.

The Senate Climb

What makes this year so difficult for the Republicans is that there are so many states with more Democrats than Republicans. There are 23 seats with more Democrats and only 12 with more Republicans. Some of these states, however, have more Democrats but some of them tend to vote Republican on a national level. That varies, of course. West Virginia and Arkansas vote Republican for President, but don't vote Republican for congress or the state legislature.

Most of the states with a Democratic incumbent are an uphill Republican climb. Winning Washington, Wisconsin, or California requires heavy cross over, which is unusual, and a big win with independents. Indiana and North Dakota have a Democrat in a place where, all things being neutral, a Republican will win. Those are the easy ones. Getting to 51 is going to require a lot of things to go right in many of these states. Even with a wave that might happen in Wisconsin, but you need a pretty big wave for things to go right in so many states.

While so many Democratic seats are up in Democratic states, there are many more Democratic seats up in the House in Republican districts. That's why the House will easier to win.


As in 2008, there will be Republican pick-ups where they win by a large margin. What about the close ones? In sports they say that the good teams win the close ones. Is it the same here?

In 2008, it wasn't. Not surprisingly, since more Republican seats were in play, more were decided by 10% or less. They pretty much broke evenly.

In 2008, four of the five forecasters made forecasts. Real Clear Politics, which is favoring the GOP most this year, did not. When all of them agreed on their forecast, flipping or not, the party they agreed on won 79 of 80 times. I think we can count on those. The good news for Democrats is that right now they only all agree on 15 Democratic seats that'll flip.

They did fairly well on the seats that actually flipped. The Crystal Ball was the only one that made predictions on all of them. Larry Sabato got 80% of them right.

Because the Crystal Ball predicted every race their winning percentage was lower. They got 75% right. If they're 75% right this year, the GOP will get a net 21 of the seats they have flipping now. The Crystal Ball has indicated they'll move more seats to the Republicans. Considering he still has 31 seats in toss-up, this number will grow.

Three forecasters left seats in the toss-up category in 2008. As they expected these seats broke fairly evenly. If all of the seats Cook predicts to flip do, as they did in 2008, and the toss-ups break evenly, the Republicans will be win a net off 44. Based on Rothenberg, it'd be a net 34. CQ would have a net of 33.

Wednesday, October 20, 2010

Rankings and Comparison

Here are my latest rankings for the House and Senate. As I mentioned this morning, I've updated the Senate to reflect that the races where it's too late to close.

Here's my predictions with those of the experts. Based on the House formula all of the experts have enough seats in the Lean/Toss-up categories to give the Republicans a majority. Everyone has at least 75 in Lean Democrat or better. Even if a bunch break Democratic, it should be difficult for the Democrats to win enough seats to retain their majority. Expect Larry Sabato to move more races into the Republican column so that individual seats match his +47 prediction.

Who Gets Undecideds?

Jim Geharty is posing undecideds in the last few polls break for the challenger or third parties. We're still two weeks out and there is plenty of time for Jim to be right.

I wanted to see how undecided in close races broke two weeks out.

The data I have doesn't indicate whether the 8% not committed to the two candidates two weeks before were undecided or third party. What we do see is that 5.1% did go with the major parties and, of that, 3.2% went to the incumbent. So, the incumbent does pick up late votes at a better pace. Gaining 1 point at this late stage doesn't help you win.

In cases where there was a 4%+ swing, the incumbent was the gainer 5 times to the challenger's 4. Six of the biggest swings were in races where the leader was ahead by 8% or more. The leader had those gains in 5 of them. The one time the challenger closed was only by 4%. It's telling that no race that had more than 1 point separating the leader and the challenger was won by the challenger.

So let's put North Carolina, Ohio, Louisiana, Florida, Missouri, and New Hampshire in the Safe GOP column. I think you can put Wisconsin in there too. One poll does have it a 2 point race, but the average is Johnson by 6. You can also throw Connecticut in the Democratic column.

Democrats definitely have 48 seats to 44 for the Republicans. That leaves 8 races that could go either way. Despite recent closer polls, I doubt Kentucky is going to flip. Conway has only been ahead in one poll all year.

The other seven are real toss-ups. All of them have had polls in the last two weeks that have favored each candidate. Most of the other polls are very close. The two recent Pennsylvania polls are a bit perplexing, since Toomey was ahead by 5-10 points in every poll since July. Sestak could win, but such a win would be unusual.

Brilliant Republican Strategy

Barney Frank is desperately raising money to defend his seat. Jim Oberstar and John Dingel are worried. I can't see any of these guys losing. Yet... There's been a poll released. Maybe some whispers on the Internet. There's a chance the congressman in a seemingly safe district could lose. Congressmen who've never worried about re-election panic. They don't know what a competitive race looks like and, well, this is a Republican wave year. Who knows? Maybe there is a 10% chance he'll lose.

So Barney Frank concentrates on his own re-election. He isn't helping other Democrats. He's raising money, contributions that could go elsewhere. The DCCC redirects money to save a chairman's seat. If Democrats spend money in seats where they are 90% likely to win without spending the money, Republicans have a better chance to pick up actual close races.

Tuesday, October 19, 2010

The Black Voter

The Black Obama voter, on the other hand, might not be up for grabs. Obama took 95% of Black voters, leaving 4% for John McCain. George Bush took 11% of Black voters. In the Senate, some Republicans did better. George Voinovich got 32% of the Black vote in his Ohio Senate race in 2004. Richard Burr got 11% of the Black vote the same year. Elizabeth Dole got 1% in 1988. If Dole had gotten 11% of the Black vote, she'd have lost by 4 points instead of 8.

Obama retains a 91% approval rating with African-Americans. There's been virtually no erosion in their support. That loyalty might translate to voting for Democrats on November 2. While White Obama voters who aren't planning on voting might be undesirable more often than not, the Black Obama voter should be counted on to vote Democratic. The more Black voters who show up, the better it'll be for Democrats.

The Obama Voter

Democrats have the idea that if they can just get all the Obama voters to the polls, they'll do fine. After all, in most polls Obama voters say they're voting Democratic over 80% of the time. The perception is that the Obama voter is a wide eyed 20 year old progressive who is set to vote Democratic for life.

That's unlikely.

The Obama voter contained a number of first time voters, some of whom had recently turned 18. The election contained some unusual circumstances. Republicans were disliked more than they'd been in any Presidential election since 1976. A number of voters weren't voting because they were Democrats or even liked the Democrats. They disliked the Republicans. Some voters were sold on Obama's charisma and his message of "hope" and "change." These people weren't dyed in the wool Democrats either. Obama got 53% of the vote, more than any Democratic Presidential nominee got since 1964. It's fair to say that some people were voting for a Democrat for President for the first time. Considering that McCain got 2 million votes less than Bush, some of them were likely Bush voters.

Democratic congressional candidates got 65 million votes, 55.6% of all House votes. The 65 million votes was 12.5 million more than the Democrats had ever gotten. The 55.6% was better than they'd done in a House election since 1982. Many were likely voting for a Democrat for congress for the first time. A mid-term election generally draws about 72-75% of the vote a Presidential election draws. Considering Obama has 36% approval with whites and got 43% of their vote in the election, it's fair to say that some white Obama voters are unhappy with him. They might not be motivated to vote Republican this year, but they sure aren't voting for Democrats. If they were motivated to vote, I think more would Republican than Democrat.

Do Not Believe Everything You Read... Unless You're a Democrat

The Internet is abuzz because PPP has a couple of polls out with better Democratic polling better and even those evil rascally Rasmussen folks are showing something. As any Democrat will tell you, Rasmussen is bought and paid for by the RNC in order to drive the election towards Republicans. Of course, I showed last month that ever since everyone went to the likely voter model, Rasmussen hasn't had a Republican skew compared to the other pollsters.

The story may be that things are moving Democratic. Don't believe it. As I said last week, early voting has already started. Nevada started voting Saturday and there are already a significant number of votes. These votes can't change. Most people have made up their minds. Many of those that haven't are probably not going to vote.

There are a number of races which are too close to call right now. They'll likely remain that way for most of election night. In 2008, Democrats and Republicans both won races decided by 5 points or less. They will again this year. But if 16 polls from 11 different pollsters show that Toomey is ahead by at least 3 and by 7 more than half the time, the race isn't going to suddenly shift this close to election day. That doesn't mean it can't move a few points either way, but senators who are far behind with less than a month to go never win.

On the other hand, don't believe that Republicans are going to take 70 or 80 seats in the House. Politico is touting "99 seats in play." Charlie Cook lists 115 races. Real Clear Politics has 133! Republicans have been putting out a lot of polls that have obscure candidates within 8 points of Democrats who were thought of as safe. They are safe, even if the experts fall for these polls. If a Republican had himself within 6, he's probably 12 points down and not in the race.

Republicans will win 223 to 230 seats in the House and 46 to 49 seats in the Senate. Which ones might change a little bit either way, but things aren't changing one way or another.

Monday, October 18, 2010

What Happens in Vegas...

... Well, it won't stay in Vegas.

I spent the weekend in Las Vegas canvassing neighborhoods for Dr. Joe Heck. I met Dr. Heck and he seems like a good guy. I like where he stands. His district will likely be one Republicans need to get to a majority. People were very receptive to his message and seem lukewarm on Dina Titus. When it comes to the Angle-Reid race, I can tell you that there wasn't a lot of enthusiasm for Sharron Angle. There was an enormous amount of enthusiasm, however, for Harry Reid to lose. "Enormous" understates it a bit. If Heck wins, it'll be because people prefer him over Titus. If Angle wins, it'll be because they want to vote for whoever is running against Harry Reid.

I met Gary Johnson, the former governor of New Mexico. I'd never heard of him until I started noticing enthusiasm for his Presidential candidacy on the web. So I asked him if he was running for President. He told me that due to timing he couldn't tell me yes. Of course that means he is. He didn't tell me not to say anything to anybody, so I'm comfortable writing about it here. (I mean, really. I probably only have about a dozen readers.) So my scoop is that Gary Johnson is definitely running for President.

The most disturbing thing that I heard is that Dina Titus' campaign is calling voters telling them that the Heck campaign unregistered them and that they need to go register again immediately... and vote for Dina Titus in early polling. I don't know if this is true. No one I know heard these calls, but people did call the campaign very irate. As they should be. Of course, no one can unregister you and no one has. Suppressing votes is illegal and lower than I ever imagined Democrats would stoop. Granted, they haven't had a problem with lying or putting up phony candidates. But this? Maybe I'm just naive.

So there you have two scoops no one else has.

Thursday, October 14, 2010

House Prediction Update

One of the nagging questions I've had is how accurate campaign polls are. If I see a poll from a Republican challenger that has him down 8, is he close? Nate Silver has access to a lot of polls. His analysis says that there is a 6 point difference between campaign polls and independent polls. So the Republican above is likely down around 14 points and is only showing this internal so that people will think he has a chance. If they think that, he might actually have a chance.

Knowing this, I was able to re-evaluate the races and dispatch the races that just aren't even in the realm of possibility. I dropped four races as potentially flipping, dropping vulnerable Democrats from 115 to 111. Now I have 83 Lean Democrat or better, compared to 89 a week or so ago. I'm lowering my prediction from Republicans 229-206 to Republicans 227-208. This is actually one more seat than I predicted on October 6. Any Republican will be happy with 226 seats, a 48 seat net.

On the other hand, Real Clear Politics now has 130 vulnerable Democrats, while Charlie Cook is at 115. Republicans aren't going to take 130 seats or even in 115. They don't have to. The narrative here is that the playing field is so vast that the Democrats don't know where they have to defend and will spread their resources then.

Can Things Change?

It's an old maxim that "after all this advertising and all these months everyone knows who they're going to vote for." I don't know if that's true, but it certainly sounds true. Yes, there are some voters who don't even pay attention until they get in the booth, but there probably aren't enough of them to swing an election.

There is a new fly in this ointment. Early voting. Two thirds of the states now allow it and an increasing number of people are doing it. I already voted. Michelle Obama voted. [Wait. Doesn't she live in a big White House in DC now?] Once people have voted the ballot is set. No matter what happens. For some, the influencing period already ended.

That makes it less likely things can change.

Wednesday, October 13, 2010

What Democrats are (Probably) Learning

George Soros announced he's sitting this election out. Bill O'Reilly doesn't believe him, but I do. Say you're a Republican donor who donated a few million in 2008. It probably didn't do any good and you wish you had that money back for this election. Even if you have a gazillion like Soros there are better places to donate your money that it'll actually have an effect.

Republicans didn't show up at the polls in 2006 and 2008. A number of Republican congressmen and senators lost, many of whom were perceived as RINOs and part of the problem. This year, they hope that replacing them with new Republicans will restore the party's principles. Progressive Democrats have been frustrated by the Blue Dogs. The lesson here is that if they sit the election out, the Blue Dogs will lose and they can elect progressives in 2012. I'm guessing they see that and won't be voting for those who've strayed from party orthodoxy.

Tuesday, October 12, 2010

Safe Districts?

John Willoughby was on Dennis Prager a few minutes ago. Remember this is my sleeper race, one absolutely no one is talking about.

When I was at the Tea Party Convention in February, an activist told me that his Democratic congressman was going down. At the time Charlie Cook had 87 seats on his list. This one wasn't on it. Cook now has 115 Democratic seats on the list. That district still isn't among them. But neither is Hawaii-2. When Steny Hoyer and Barney Frank are running advertising, is any district safe?

Monday, October 11, 2010

Generic Ballot

Two weeks ago people were making a big deal out of a few polls like Gallup and Newsweek that had the Democrats doing better. Had the Republicans peaked? Not quite. Both of these forecasters were using registered voters. In this cycle those will look more Democratic than likely voters.

I've divided September into an early and late. So it has two points on here. The likely voter polls have averaged 7.3 in early September, 6.7 in late September, and now 7.2 in October. From January through August, only 2-3 pollsters were using likely voter, so September is a good place to start looking. I used the lower number for Gallup because Republicans +18 looks too high. Interestingly, Rasmussen is now the most favorable pollster for Democrats. For the last year, Democrats have yelled that Rasmussen is too Republican leaning.

The green line is "registered voters." It's true that "registered voters" were trending Republican until mid-September. Late September polls were only Republican +1. October is shown as even, but that's only because no pollster has polled the generic ballot using only registered voters. Thus, Likely and All voters are the same here.

House and Senate Race Ratings

Prediction: Republicans 229 Democrats 206

Prediction: Republicans 49 Democrats 51

How You'll Know Who is Going to Control the House

As election night returns come in, we’ll want to know whether Republicans are on pace to take a majority in the House or if the Democrats will retain control. We’ll be able to judge that based on how many net seats Republicans have picked up. There are two caveats to the following timeline:

1. Exit polls might not be available for many races. So we’ll rely on the SoS tally in each state. Some states are faster than others. So a state might not be done before later states come out with their returns.

2. Some races will be so close that the results won’t be known until days later. Those seats count just the same as ones won on election night.

7 PM Polls Close: South Carolina, Georgia, Indiana, Vermont, Kentucky, Virginia
Democratic Seats possible: 13
Democratic Seats that are very vulnerable: 7
Republican Seats possible: 0
Republican Seats that are very vulnerable: 0
Republican net pick-up count: +4

Republicans have strong pick-up possibilities in Virginia and Indiana. If they’re not flipping, Republicans could be in trouble. Seats in Georgia and Kentucky are less vulnerable. If they’re flipping, it’ll be a long night for Democrats.

7:30 PM Polls Close: North Carolina, Ohio, West Virginia
Democratic Seats possible: 13
Democratic Seats that are very vulnerable: 7
Republican Seats possible: 1
Republican Seats that are very vulnerable: 0
Republican net pick-up count: +9

Republicans should pick up either five seats at 7 and four at 7:30 or vice versa to be up 9 seats. Ohio is a huge pick-up opportunity for the GOP, with five seats that look like a strong possibility to flip. At this point Republicans shouldn’t have lost any of their own seats. In the nine states that close before 8 PM, only OH-12 is considered to be potentially vulnerable and that seat is considered highly unlikely to flip.

8:00 PM Polls Close: Alabama, Connecticut, Delaware, Florida, Illinois, Maine, Maryland, Massachusetts, Mississippi, Missouri, New Hampshire, New Jersey, Oklahoma, Pennsylvania, South Dakota, Tennessee, Texas
Democratic Seats possible: 48
Democratic Seats that are very vulnerable: 28
Republican Seats possible: 8
Republican Seats that are very vulnerable: 3
Republican net pick-up count: +23

There are a lot of states here and a lot of vulnerable seats. The key states to watch are Illinois, New Hampshire, Pennsylvania, and Tennessee. Illinois has four strong possibilities for the GOP, while there nine possible, and six very vulnerable.

Three of the five most vulnerable GOP seats, IL-10, FL-25, DE-AL, are possible flips, but a strong GOP showing should be able to overcome that. At this point the Democrats will have had 74 possible losses and 42 very vulnerable compared to 9 and 3 for the Republican Party.

If it’s going to be a GOP landslide the Republicans might already be at +39 before 8:30 PM eastern.

8:30 PM Polls Close: Arkansas
Democratic Seats possible: 3
Democratic Seats that are very vulnerable: 2
Republican Seats possible: 0
Republican Seats that are very vulnerable: 0
Republican net pick-up count: +25

Arkansas has to be different. While picking up two seats here isn’t mandatory for the majority, the Arkansas seats are very important for the GOP. Despite voting Republican for President, Democrats have continued to hold these seats for years. If they don’t flip now, they might not for a long time.

9:00 PM Polls Close: Arizona, Colorado, Louisiana, Michigan, Minnesota, Nebraska, New Mexico, New York, Rhode Island, Wisconsin, Wyoming
Democratic Seats possible: 32
Democratic Seats that are very vulnerable: 17
Republican Seats possible: 5
Republican Seats that are very vulnerable: 1
Republican net pick-up count: +34

I consider it very likely that people will be using “Speaker Boehner” at this point. While they should be at least +34 to be on pace, I expect them to be over 40 when these votes are counted. Arizona, Colorado, Michigan, and Wisconsin are states where Republicans could pick up multiple seats. It’s possible, but unlikely, that all three Democratic New Mexican congressmen could go down.

New York has 26 Democratic districts and there are a multitude of possible Republican pick-ups here. The New York Republican Party has been very ineffective. It’s possible the GOP might only pick up 1 or 2 New York seats.

If Joe Cao defends the most Democratic seat held by a Republican, it’ll be a GOP landslide.

10:00 PM Polls Close: Hawaii, Idaho, Iowa, Kansas, Nevada, North Dakota, Utah
Democratic Seats possible: 8
Democratic Seats that are very vulnerable: 4
Republican Seats possible: 2
Republican Seats that are very vulnerable: 1
Republican net pick-up count: +37

These 7 states have only 20 congressional districts. There is a strong pick-up possibility in Kansas and a 3 other good ones, but this isn’t a group of states that Republicans should count on.

11:00 PM Polls Close: California, Oregon, Washington
Democratic Seats possible: 10
Democratic Seats that are very vulnerable: 5
Republican Seats possible: 5
Republican Seats that are very vulnerable: 1
Republican net pick-up count: +39

If Republicans are counting on the Golden State or the rest of the west coast they could be in a lot of trouble. Democrats only picked up one Republican seat in 2006 and 2008, so this list of vulnerables may be high. On the other hand, there were six California/Washington districts the GOP retained by less than ten points in 2008. Two are considered safe, while three of the four other seats are thought to be Democratic long shots. That said, if California bucks the national trend and goes Democratic, Republicans could lose the majority they thought they won earlier in the evening.

Alaska closes at midnight eastern. There’s only one House seat there and the Republicans are holding it. A recent poll showed Don Young with a 33 point lead. If his seat is even in discussion Republicans won’t be close to a majority.

Keep in mind that this is the pace the seats need to flip, not a prediction. If Republicans are +37 before the 11 PM returns come in, they should get to 39, but that's in no way guaranteed. If you're a Republican you want the GOP to be ahead of the pace each half hour. If you're a Democrat you want them below it.

Thursday, October 7, 2010

Hawaii-2 In Play

Earlier this year the special election for Hawaii's 1st congressional district made the news. Charles Djou, the Republican, won, but he only had 39% of the vote. The two Democrats had 58%. So it was widely assumed that when Djou ran against only one Democrat in November he'd lose.

Hawaii has two congressional districts, the 1st in Honolulu, and the 2nd which covers the rest of the state. The districts are Cook D+11 and D+14, respectively. They probably aren't that Democratic. The 1st was D+17 in 2008, but D+5 in 2004. The 2nd was D+20 in 2008, but D+8 in 2004. Barack Obama is a native of Hawaii and he likely got a lot of additional votes that wouldn't normally go to a Democrat. At D+8, Hawaii-2 is probably at the limit of possible Republican pick-ups this year. Republicans are running John Willoughby As of the end of August, he'd only raised $16,000. You can see why no one is watching this race.

A few days ago dailykos/PPP came out with a poll that showed the race virtually tied. This is a bit of a surprise, but nothing much should surprise us this year. After all, Djou is a sitting congressman, has raised a good amount of money, and it's a Republican year.

Duke Alona, the Republican Lt. Governor, has virtually the same percentages in the district.

The closeness of the percentages is a bit eerie, but clearly if you're voting Republican you're voting for both candidates and if you're voting Democrat, you're voting Democrat across the board. dailykos/PPP also came out with a poll covering the entire state. Looking at the number of likely voters surveyed it appears that PPP surveyed a fairly equal number of voters in each district. By knowing the numbers overall and the numbers for Hawaii-1, we can guess the numbers for Hawaii-2 with some confidence.

I've also included the senate race. Daniel Inouye, the long time well respected senator, picks up Alona/Djou voters across the board. He seems to pull in a similar number of Alona Democrats, Republicans, and independents in each district. It's likely that the Alona/Abercrombie voters are doing this in Hawaii-1.

We know that the Alona voter is almost exclusively voting Republican for Congress in Hawaii-1 and will vote about 61%-39% for Cavasso. We've seen that Hawaii-2 is virtually the same in the voting patterns. In fact, the Alona voter is probably 65%-35% for Cavasso.

So we've seen two different Alona voters. The first will vote Alona/Cavasso/Djou and the second will vote Alona/Inouye/Djou. Those two voters seem to be identical in Hawaii-2 voting Alona/Cavasso and Alona/Inouye. As we saw above, the breakdown by party of these two groups is pretty much the same in both districts. In Hawaii-1, they all are voting Djou. Logically, if they're doing that, they are likely to vote Willoughby in Hawaii-2. The Alona voter breaks down almost identically in both districts for their senate vote. They should do the same for House vote.

dailykos/PPP either didn't survey this race or didn't publish the results. Either way I'm betting that this survey would show a tight race if the question were asked. Since dailykos and PPP are both Democratic, we know that they aren't bending the numbers in the Republicans' favor.

The Demographics

Democrats were certain after 2008 that they'd overwhelmingly win every election for now on. After all, "demographics are in our favor." I'm sure they know this because they read it on dailykos. Now that it appears Democrats not only won't overwhelmingly win this election, they'll also get trounced, they've changed that for 2012 and beyond.

They use Obama's strong showing with Hispanics and the 18-29 group as evidence.

Hispanics aren't leaving the Republicans at all. Republicans took only 22% of Hispanics in 1996. This increased in 2000 and again in 2004 before dropping to 32% in 2008. I can give a lot of reasons why some Hispanics prefer the Republican Party, not the least of which is that the average Hispanic is MORE socially conservative than the average Republican. There are a number of economic reasons Hispanics could favor Republicans, but that's another debate.

There's no doubt that the current immigration debate favors the Democrats, but that's unlikely to be an issue long term. Eventually the border will be secured and immigration reform will happen. Puerto Rico can provide an excellent example of Hispanics when the immigration issue is removed. Puerto Rico has long been dominated by the NPP, most of whom are Republicans. The current governor, senate President, and House speaker are all NPP and all Republicans. If Puerto Rico gets statehood, the GOP is likely to take the majority of the seats. As Puerto Rico shows, if you remove the immigration issue Republicans will do just fine with Hispanics. Certainly Republicans will eventually have problems if they keep getting 32% of Hispanics. How Hispanics vote when they are a significant force is many years off.

Let's see how Republicans do with this group this year.

Youth Vote
Obama took 67% of the 18-29 vote. That's significant, but it's questionable if that's sustainable over time. When you include only Republican and Democratic votes Bill Clinton got 60% of the 18-29 year olds in 1996. In 2008 Obama got just 53% of this group, the same percentage overall. Some people, as they get older, become Republicans for a variety of reasons. There's no way to know how future generations will vote. People that came of age in the 60's and 70's were overwhelmingly Democrats. The generation that came of age in the 80's skewed Republican. In 1984 Reagan took 59% of the 18-29 vote. He took 59% of the vote overall. Of course, the young people who came of age in 1990s and 200s were a lot more Democratic.

The best way to correlate this is that Ronald Reagan and Bill Clinton were popular with young people, while George Bush was unpopular. Young people are heavily influenced by the President who is there when they come of age. Long term there's no reason to think Republican Presidents will be unpopular or Democrats popular. Democrats had no charismatic leader from 1963-1993 and none between Clinton and Obama.

It's also questionable whether Obama's success with the youth vote will carry over to other Democrats. In the 2009 gubernatorial race, Creigh Deeds got 44% of 18-29, 44% of 30-44, and 41% of 45+. His skew was fairly flat.

Democrats took 56% of the 65+ vote in 1992, 53% in 1996, 52% in 2000, 48% in 2004, and 46% in 2008. It seems that more Democratic voters are dying and they are replaced by Republican voters. If this trend continues, the next 10 years look fairly rosy for the GOP.

It'll be interesting to see how the 18-29 age group and Hispanics vote this year. When Barack Obama is on the ballot again the electorate will surely be different, but he's only on the ballot one more time if he wins in 2012. Maybe he'll be on again if he loses, but if he loses the magic is likely gone. The 2008 vote might have been an outlier. I don't think we can make any assumptions on this year's electorate or the 2012 electorate.

Wednesday, October 6, 2010

Does How Republican The District Is Matter?

It sounds logical that the Republicans won't flip Colorado-7, which Obama won by 19 points, without getting Colorado-3 and Colorado-4, both of which McCain won. I decided to look at 2008 results to see if this were true. I included all districts where the winner won by 15 points or less. Districts that had a wider margin likely weren't competitive for reasons like a poor challenger. So it didn't matter how many people voted for McCain or Obama. Even though McCain won Oklahoma-2 by 32 points, it's unlikely to be competitive.

I could use 2006, also a Democratic year, but using 2006 congressional and 2008 Presidential are using different electorates.

Clearly there were a lot of safe seats in this group. That's not a surprise. The Democrats won 6 of the 15, only 2 of which were long time Democratic seats. [Alabama-5 is in red because Griffith changed parties.] They did drop a Louisiana seat they picked up only a few months before. Texas-22 was a 2006 pick-up that flipped back. Only 1 win was over a Republican incumbent, however.

Democrats did slightly worse in this group. They were able to defend several recent pick-ups, but lost Kansas-2, a 2006 pick-up. No Republican incumbents lost.

The parties split the districts that were Presidential toss-ups.

The Democrats did better in Democratic skewing districts, but still managed to lose 7 of them.

Democrats still lost in districts Obama won by significant margins and they had a competitive candidate. The Democrats knocked off a bunch of Republican incumbents here.

In Pennsylvania, Obama won the 6th district by 17 points, but the Democrats were unable to take the seat. Yet they won 3 other competitive Pennsylvania seats that McCain won. While they didn't knock off many Republican incumbents in McCain districts, a number of those districts were recent pick-ups where they knocked off Republicans in 2006. Republican skewing open seats were much more vulnerable than Republican incumbents in Republican seats. The bad news for Democrats is that they have significantly more Republican seats than Republicans had Democratic seat in 2008.

If this provides a rule for 2010, Colorado-3 and Colorado-4 should be more vulnerable than Colorado-7. The question is how far the Republican wave will go into seats Obama won and whether the Republicans will take a higher percentage of Republican skewing seats than the Democrats did in 2008.

Tuesday, October 5, 2010

Monday, October 4, 2010

The Battleground

There's been a lot of talk that many key races this year will be fought in the frost belt, with New York, Pennsylvania, and Ohio having more than any state. One area that is a microcosm of that is the Philadelphia metropolitan area. While the two Philadelphia House seats have never been in play, many others are:

The Senate
Pennsylvania - This is the key area for this race. Sestak will get Philadelphia, Pittsburgh, and some smaller urban areas throughout the state. Toomey will get the rural areas and likely grab the Reagan Democrats in western Pennsylvania. The Philadelphia suburbs used to be Republican, but the area has changed. The counties around the city voted for Obama by 10-20 points. Joe Sestak's district is here. Sestak would need a big turn-out in this area and to hold on to their votes. It looks like Toomey may hold his own here. Toomey has a nice, but not solid lead, according to the polls. This should be the 4th Republican pick-up after North Dakota, Arkansas, and Indiana. A solid win here would mean bigger gains are possible. A struggle here likely would mean Wisconsin, Colorado, and other possibilities are in trouble.

Delaware - Some people, including me, think this race won't be competitive. That's not what makes it interesting. Christine O'Donnell is a huge symbol of the tea party. Whatever showing she gets will largely be due to tea partyers. This'll be a sign of what they can accomplish.

The House
Pennsylvania 7 - This is Sestak's open seat and one that looks to be a sure pick-up for the GOP. They have a strong candidate who has done well fundraising. Before Sestak, Republican Curt Weldon held the seat for 20 years. This is a D +3 district. While Republicans expect to pick up many Republican seats, this will show they can pick-up Democratic seats.

Pennsylvania 8 - Similar to the 7th, but this has incumbent Patrick Murphy. Like Sestak he took the seat from a Republican in 2006. This is a D +2 district. Mike Fitzpatrick is ahead in the polls. This exemplifies Republicans beating incumbents.

Pennsylvania 6 - Obama did better in this district than he did in the 7th or 8th, but Republican Jim Gerlach held on by 1 and 4 points the last two elections. Gerlach should hold the seat fairly easily in this environment, but it's the type the GOP could lose if the tide isn't strong enough.

Pennsylvania 13 - This used to a swing district but the 2002 redistricting turned it solidly Democratic with Obama taking 59% of the vote. Allyson Schwartz's seat is considered safe, but I drove around the district last night. I saw a lot more Adcock yard signs than Schwartz. We haven't seen polling here, but if the wave is big enough even this district could fall.

Delaware - Delaware is heavily Democratic and likely needs a Mike Castle type to pick off enough Democratic votes to win. The GOP has held this seat for the last 18 years and 34 of the last 44. It should be a loss. A win here would could mean the Republicans don't lose any of their own seats.

New Jersey 3 - Similar to the seats on the other side of the river, this suburban Philadelphia went to Obama, by a small margin, and was held by a Republican until the 2008 switch. Former Eagle Jon Runyan is running for the Republicans. In a strong GOP year, he should win just by being an ex-Eagle alone.

New Jersey 12 - A touch to the north in Princeton, Republicans haven't had this one in 12 years. Obama got 58% of the vote. Like Pennsylvania 13, this would mean ousting an entrenched incumbent.

New Jersey 2 - The southern portion of New Jersey. Frank Lo Biondo has held the seat longer than Jim Gerlach, but he's never won a race by less than 20. Obama won the district after pushed squeaked by. In a Democratic year this would be a target. As it is now, the seat isn't mentioned at all.