Wednesday, October 31, 2012

California Absentee Analysis 10/31

There are now 2.8 million California VBM ballots in, around 20% of the total votes. Right now the ballot distribution is 43.3%D/36.8%R/19.9%I. That's a 6.4% Democratic advantage, 1.2% above the 5.2% last Wednesday. In the last week, the Democrats have had a 7.5% ballot advantage. None of these numbers should scare Republicans. The 2008 exit polls had Democrats with a 12% advantage. So Republicans would like to keep it below 8% and Democrats would like to get it above 11%. Right now it looks like Democrats will have to hope for a big advantage in election day turn-out. It's possible, but there's no way to know.

Simpler Math

There's a national-state poll discrepancy. Mitt Romney leads by 0.75 points in the national average and trails by 1.11 points in the state averages. That's nearly 2 points. Right now there 6 states where Mitt Romney trails by 2-3 points in the RCP state averages. So a 2 point bias could be the difference between winning and losing. I've weighed in on it. Now Sean Trende and Nate Silver have.

The national total votes are the sum of the state total votes. It seems silly to remind people of that but some people ignore that when looking at the numbers. I think they do that because they only see the poll numbers, not the raw number of votes. So people grasp for explanations.

Romney is Doing Extraordinarily Well in States That Aren't Being Polled
The explanation for this is that Obama has an amazing campaign and ground game and that Romney will gain more where Obama isn't contesting. There are a few problems with this. First, it's anecdotal. No one knows if Obama's campaign is superior or how much that'd impact the numbers.

There are holes in this argument. As my chart shows, Romney has improved more in state polls for swing states than in other ones. There could be good reasons for this. As I noted, McCain maxed out the White vote in the South, limiting Romney's potential increases. Yes, Obama isn't competing in these states, but then neither is Romney. Romney needs to sell people he's a better alternative. There's only so much you can do without delivering a message. The big red and blue states (Texas, Georgia, California, New York, Illinois, New Jersey) are getting polled and those results are included. Delaware and Wyoming aren't swinging the election that much.

Exceptional Turn-out in Red States
The problems with this is that the pollsters aren't weighing red states heavier and no one is talking about red state turn-out being higher. It's likely to lag since evangelicals that have never warmed to Romney can leave their Presidential vote blank because they know Romney will win their state. Even increasing turn-out by 10% only in red states only moves the totals to Obama +0.82, a small percentage of the discrepancy.

The National Polls Are Wrong
The explanation for that is that state polls did better in 1996 and 2000. Yet in 2004 and 2004 the national polls were far more accurate, indicating that they changed methodology due to that weakness. The national pollsters tend to stay more stable. There are plenty of new state pollsters with unknown track records entering and leaving each cycle. The national polls could be wrong but no one has given an explanation as to a methodological flaw that doesn't exist in state polls.

That doesn't mean that the national polls can't be wrong, but there's just as much of a chance the state polls are wrong, or we meet in the middle. Remember that they all contain a margin of error, so the state polls and national polls can probably meet anywhere within Obama +2 to Romney +2 and there wouldn't be anything wrong with the polling.

What is wrong, however, is assuming Romney wins by roughly a point and the state polls being right. That's even simpler math.

Tuesday, October 30, 2012

Mittmentum in California

In 2008, Barack Obama won California 61%-37%, a remarkable performance above what any California Democrat could do statewide. If we use the same 98% the Presidential race got in 2008 the average statewide race is 54% Democratic 44% Republican. The typical Democrat would win by around 10 points. Barack Obama's 13.75% lead in the average of the last four polls tells us that Mittmentum isn't nearly as strong here as it'd appears to be and that Obama is still doing very well.

One piece of conventional wisdom is that because the President will win California he'll dominate the congressional districts. While he will dominate very Democratic districts, he won't dominate the swing districts.

Using the 2010 senate race as a guide the percentages below would be about what you'd expect in a race where the Democrat wins by 13.75%.

I project Mitt Romney will win 5 competitive districts, Barack Obama 3, and 4 will be roughly even. The President will win CA-9, 41, and 47, Jerry McNerney's district and two open Southern California districts. Barack Obama's coat tails might be enough to bring those home.

Mitt Romney may have coat tails too. Four of the five congressional districts I project him to win are occupied by Republican incumbents. Democrats failed to win any of the 8 Republican congressional districts that Barack Obama won in 2008. If they couldn't unseat Republican incumbents when Barack Obama won the district their chances should be less when he loses it.

The four districts that project to be roughly even have three Democratic incumbents. Since incumbents usually beat their Presidential candidate, John Garamendi, Lois Capps, and Jim Costa should be favored to win. The last district, CA-26, is an open district contested by Republican Tony Strickland and Democrat Julia Brownley. There appears to be no favorite there.


Democrat Raul Ruiz is out with an internal poll that shows him winning by 6 points. This is difficult to believe for the following reasons:

1. Ruiz concedes Bono Mack is winning by 6 points with those who've already voted. That makes sense because Republican VBM returns are currently R+5. My estimate is that 25% of the electorate has voted.

2. So Ruiz has to be winning the rest of the voters by 10 points to be up 6. Considering that 33% of voters have no impression of him, I have to wonder where his votes are coming from.

3. In 2008 Obama won the old district by 5. Bono Mack won by 17, a 22 point difference.

4. Obama won the state by 24 in 2008, but four consecutive polls taken this month by reputable independent pollsters put the margin between 12 and 15 points.

5. If Mitt Romney is doing 10 points better statewide, he's averaging doing 10 points better in each congressional district. President Obama has lost the most with White voters, so he's probably lost more than 10 points in Whiter districts. They have Obama leading by 7. That means Obama would've gained 4 points here, 14 points away from what's actually happening. It's a statistical impossibility for Mitt Romney to cut the lead by 10 points and lose 4 points in a congressional district.

Barack Obama isn't up 7 points in this district. Even if he were, Mary Bono Mack is unlikely to get less of the vote than Romney. Did I mention that Democrats have never won a seat where the majority of the district was in Riverside County?

Monday, October 29, 2012

California Absentee Analysis 10/29

With another 370k VBM ballots in we've cleared 2.1 million. This is now 14.6% of the total anticipated ballots. Since Whites are more likely to be VBM, the total ballots in CA-10, 24, 36, and 52 is at or above 25%. The good news for Democrats is that their electorate advantage went from 5.9% to kkkkkn

CA-16 continues to move leftward, while CA-21 moves further to the right. The only other congressional districts with movement are CA-41 and 47, both moving slightly more Democratic.

Howard Berman is a Democrat!

Ladies and gentlemen I bring you Top Two, where candidates with seriously partisan voting records need to pander to the other side. This Brad Sherman flyer takes the cake.

It's interesting how Sherman trots out the local Valley politicians while Berman's endorsement are heavily Republicans from Capitol Hill. It reinforces the image of Sherman is local and Berman is Washington. Next thing you know we'll be hearing that a vote for Howard Berman is a vote for Nancy Pelosi as Speaker.

The Southern White Vote

One theory floated around about the swing state-national discrepancy is that Mitt Romney is doing a lot better in under polled red states. Sean Trende of Real Clear Politics pointed out on Twitter why this is unlikely.

John McCain maxed out White support in Alabama, Mississippi, and Louisiana, getting at least 85% of the vote. While Romney can improve his share of the White vote in other states, that's unlikely here. He probably came close to maxing his support in South Carolina, Texas, and Georgia. The only way for Romney to improve in these states is to grab support from Blacks, something he's unlikely to do. As our Democratic friends constantly remind us, the south is becoming less White each year. So it's unlikely we'll see much of an increase in the White electorate.

If you're looking for states for the President to improve in, Alabama, Mississippi, and Louisiana are probably good bets.

Sunday, October 28, 2012

Simple Math

It's simple math. When you add up all the votes in the states, you get the national vote. Well, I've done that, with the average October poll. When there was none, I made up a number that was in line with how much a state should move based on the national polls.

The result is Barack Obama leads by 0.81%. Wait, you say. The RCP average says that Mitt Romney is leading by 0.90%. How can they both be right? They can't. If the RCP state averages are right, then the national average is wrong. If the national average is right, then the state averages are off by an average of 1.71%. That means that Romney should be leading in any state where the RCP average is below Obama +1.5% and should be in at least some of the states where Obama leads by 2.5%. And there are four of those.

If you believe the national polls are close to accurate and they were within 0.3% of the 2008 margin, then the race in swing states is much closer than it appears.

Saturday, October 27, 2012

California Absentee Analysis 10/26

Here are the latest absentee numbers. The vote is now around 12% of the expected total vote, up from 10%. Democrats bump up from a 5.8% ballot lead to a 5.9%. That's positive movement on their part but a far cry from the 8-9% they need to break even and the 12-13% they need to pick up seats.

The biggest changes were in the Central Valley where CA-16 went from a D+2% to D+5% turn-out advantage but CA-21 went from D+16% to D+11%. Democrat Jim Costa and Republican David Valadao are considered pretty much shoe-ins, so these shifts aren't surprising. The absentee numbers here have been low so far. Hence the ability to have a big one day shift.

Based on my calculations, CA-10 already has over 25% of the total overall ballots for the whole election. The spread is even, 42%D/42%R, just as it was when Jeff Denham had his big primary victory. CA-24, CA-36, and CA-52 are also over 20%.

CA-52 had a big Republican turn-out advantage in the primary, similar to CA-36. While Rep. Mary Bono Mack turned that into a 16.2 point win, Rep. Brian Bilbray eked out a 2.5 point win. Even though Bilbray has another big turn-out advantage, the model isn't kind to him. This is contrary to almost every other district in the state where Republicans exceeded any turn-out advantage, either by picking up Democratic votes or winning a good share of independents. Bilbray retains the biggest Republican edge in competitive seats in the state. So I'm not sure how he can do better.

Politico has an article today about how well Democrats are doing here. This surprised me because it's contrary to what I'm hearing. Now Republicans might be blowing smoke up my butt, although I can't think of why they'd tell a reporter we're doing badly and tell me we're very competitive. What I do know is this:

1. Despite cries of Republicans getting beat in June, the opposite happened. Republicans got more votes in almost every competitive district.

2. Absentee turn-out in June was D+5.5%. The final vote was D+5.3%. It's currently D+5.9%. That doesn't sound good for Democrats.

3. Twelve of the 16 internal polls released have been released by Democrats. Releasing internal polls is usually a sign the candidate is losing and trying to convince people otherwise.

4. Those internal polls have mostly been lackluster. They've shown Democrats with big leads is some, but haven't shown a lead in either CA-7 or CA-10.

5. The one independent poll out there showed Rep. Brian Bilbray with a comfortable lead.

The truth? I suppose we'll see November 6.

Thursday, October 25, 2012

California Absentee Analysis 10/25

Absentee counts jumped another 300k and we are over 10% of all the California votes for the first time. It was a good day for the Democrats saw the absentee spread go from 5.2% (42.7%D/37.5%R) to 5.8% (43.0%D/37.2%R). This is still nowhere near where the Democrats need it to be to have a good election. I've been unable to confirm that Republicans will be a lot closer with VBM than on election day, although I got a tip that VBM so far has skewed older than it should. Unless I can figure out another way to adjust the numbers I'll continue to report the results absentees only would produce.

f CA-16 remains as the lowest number of absentees turned in, but a big Democratic jump isn't unexpected. The other changes were relatively minor, but CA-24 and 36 moved 2 points to the right.

Democrats gained ground in the two swinish distress, SD-27 and 31.

Most of the assembly districts were unchanged.

Wednesday, October 24, 2012

California Absentee Analysis 10/24

Ballots in California are counted by the counties and then submitted to the state. The counties don't submit to the state on a uniform basis, so one county might submit one day but not the next. Districts that are in more than one county could have a Republican county submit but the Democratic county does not. So the absentees will move more Republican that day. Keep that in mind when looking at day to day shifts. Overall, the Democratic absentee advantage declined from 5.6% to 5.2%. Considering that 5.6% was a bad number for Democrats, this is the wrong direction to move in.

Even though overall absentees moved to the Republicans, more swing districts moved Democratic. The most significant shift is in CA-3, which moved 4 points more Democratic since yesterday. CA=10, 16, and 47 moved one point more Republican, while CA-21, 24, 26, 36, and 52 moved one point more Democratic. Republicans are still projected ahead in 11 of 12 districts, but CA-24 and 52 are razor thin leads.

The only change is SD-27 moves 1 point toward the Republicans. Newcomer Todd Zink had to be considered a big underdog in this swing district, but there have been nearly 1,700 more Republican ballots submitted thus far.

Few moves here, although AD-9 is looking increasingly competitive.

Tuesday, October 23, 2012

California Absentee Analysis 10/23

California absentee numbers are coming in and they can tell us how the general election will go. It’s just a question of knowing how to interpret them. Since nearly half of the ballots in California are absentees these days.

Let’s start with the primary. The absentee electorate was 43.9% Democratic/38.4% Republican. The actual electorate was 42.7% Democratic/37.4% Republican. So the relationship between the absentee and actual electorate is very strong. Registration in the state is D+13.2%. Since the electorate was D+5.3%, Republicans did very well in the election.

While Republicans did better than expected, most observers expected the electorate to be closer to registration in November. So far that hasn’t happened. In fact, it’s 42.8% Democratic/37.2% Republican, about the same as June. The good news for Democrats is that this is probably only about 10% of the people who will vote.

While the electorate is expected to be more Democratic, there’s no reason that the electorate that’s sent in their votes so far won’t be the same as it is now. We know these people have voted and don’t know about others. So, from here to the election I’m going to speculate how the people who’ve sent in their ballots actually voted.

The first column gives the partisan advantage in the primary. There were 5% more Democrats in the CA-3 primary. The next column gives the result. John Garamendi beat the Republicans in the election by 2.9%. Republicans did 2.1% better than the party difference. Currently there is also a 3% Democratic turn-out advantage. If the same principle holds up, Garamendi should be leading by 0.9%.

To do so, I’m going to assume a similar relationship between the electorate and the vote in the general election as there was in the primary. Since the overall electorate looks strong for the GOP statewide, so do the congressional districts. Republicans got more votes in 10 of 12 competitive districts in June and calculate to be ahead in 11 right now.

The Senate districts are fairly on track with what you’d expect since turn-out looks similar. SD-27 and 31 figure to be close. Republican Todd Zink looks good in SD-27 with the current R+5 turn-out, but fellow Republican Jeff Miller may need more of an advantage in SD-31. If the GOP wins SD-5 and 27, while the Democrats win SD-31 and 39, the Democrats will have a 27-13 advantage, enough for a 2/3 vote.

Strong Republican turn-out in assembly districts translates to Republicans winning 11 of the 15 districts that are considered potentially competitive. That looks better than it probably is, since the June results indicate less of these districts will be competitive than people think. If these are the results, Democrats will have 48 of the 80 seats, far short of the 2/3 they need to pass new taxes. They need 54.

I want to reiterate that this analysis makes a number of correlative assumptions which might not actually happen and that 9% of the electorate might not reflect the final electorate.

That said, a party ID advantage of 5-6%, like we're seeing now and saw in the primary, is extremely bad news for the Democratic candidates. They need electorates that are much closer to party registration to pull out wins. If this holds up expect more Republican victories than are being predicted.

Wednesday, October 17, 2012

California online registration

There have been over 540k online registrations since they started it. That's a 3% registration increase. Registration had only increased by 300k in the first 9 months of the year prior to that.

1. Many of the applications are registration updates, so they aren't all new voters.

2. It's likely that a higher percentage are invalid than normal, as registering fraudulently is easier.

3. In the last 45 days, registration went up around 300k in 2010 and by 1.1 million in 2008. So a registration jump is normal.

This has the potential to be painful to one party or the other. Word is that Democrats are benefitting heavily, but the SoS has yet to confirm that. These reports frequently don't go up for several weeks after the deadline date, so accurate counts might not be available before the election.

Monday, October 15, 2012

CA-9 and 36: Democratic Polls Look Good For Them

The Democratic Congressional Campaign Committee has put out two polls showing their party winning CA-9 and CA-36. Let's take a look.

Barack Obama won California by 24 in 2008. SurveyUSA polled the state in September and showed Barack Obama with a 22 point lead. They polled again after the debate. Barack Obama had a 14 point lead. While one poll shouldn't be considered definitive, Romney's gain is consistent with his gains throughout the country. If the President's lead has indeed dropped 10 points then California congressional districts will also drop by an average of 10 points. After all, the statewide vote is the sum of the 53 congressional districts' vote.

The President isn't expected to drop much with minorities so he'll likely drop by a lot more with white voters. Neither of these districts is majority minority, so it's pretty safe to assume that these districts will drop more than the majority minority districts to get to the result.

2008:Obama +15%
Expected 2012: Obama +5%
Democratic poll: Obama +11%

2008: Obama +3%
Expected 2012: Romney +7%
Democratic poll: Obama +5%

Obama winning CA-9 by 11% is possible, although it's probably high. It should be close here, especially since this is a majority White district.

If Barack Obama wins California by 14 points, he isn't winning CA-36, let alone by 5 points, even if Riverside County is trending Democratic. So let's adjust the polls based on how much the Presidential number is off.

Democratic poll: McNerney +9%
Adjusted poll: McNerney +3%

Democratic poll: Ruiz +3%
Adjusted poll: Bono Mack +9%

I'm not one to approve of polls just because the results look like what I think they should, but you tell me which looks right.

Sunday, October 14, 2012

ABC-WaPo: D+9 Electorate

I'm not one to buy conspiracy theories, but ABC News-Washington Post has a poll that has a D+9 sample, their biggest Democratic edge of the cycle and bigger than 2008. So while they are showing more Democrats and fewer Republicans going to the polls, they're also showing the enthusiasm gap swinging toward the GOP. Thus, even though Republicans are more enthusiastic they've decided to sit the election out and stay home. It's like the pollsters want to convince Reps their candidates can't win so they won't go to polls.

How Swingy are Swing States

Mitt Romney is winning. At least that's what the national polls say. There hasn't been a poll taken entirely after the October 3 debate that has had Barack Obama leading. Yet Mitt Romney doesn't appear to be winning the swing states and getting the majority of the electoral votes. Why not?

I looked at Mitt Romney's swing state v. national performance two weeks ago, before the debate, finding that Romney actually improved by more in the average swing state than in the national poll. At the time, President Obama led by 3.7 points. Now the election has moved more than 5 points toward Mitt Romney, with the former Massachusetts governor leading by 1.4 points in the RCP average.

Mitt Romney led in no swing states on October 1. He leads in North Carolina, Florida, and Colorado now. Yet that'd still give a 294-244 electoral college loss. What gives?

As the above chart shows, Romney has gained dramatically in swing states, although not by as much as he has nationally. Now, he's running 8.7% ahead of McCain nationally and 7.9% in swing states. Of course, all swings won't be the same. If he was getting an 8.7% swing in all swing states, he would add Virginia and Ohio, but not Colorado. He'd still lose 272-266. Still, the average gain isn't too dissimilar to the national gain. He won't necessarily gain the same in swing states as he does nationally.

The above chart projects out state by state vote based on margins and projected turn-out. Barack Obama leads by 1.8% if we look at all the state polls. That's incongruous with a Mitt Romney lead of 1.4%. This makes more sense when we realize that the numbers used for the above chart include 15 red states and 8 blue states which haven't been polled since the debate. If those states were adjusted based on expected post debate swing Romney would lead by 0.4%. That's still 1.0% behind his national number but such a variance could easily be explained by different pollsters, different days, and even that a poll that has Romney ahead by 2 points is actually 1.7.

Romney's average move of 7.9% in the swing states is below his aggregate move of 7.2%, because his move is larger with small states than bigger ones like Pennsylvania, Ohio, and Florida. Even so, he's around 50%-50% in swing states. Because they haven't been polled lately, or as often, Mitt Romney isn't gaining as much in the red and blue states as he is in the swing states. Again, however, that probably wouldn't be the case if there were more recent polling data for many of them.

Thus, we can't assume that Mitt Romney winning the popular vote will mean an electoral college win. It's likely it will, but wouldn't be inconsistent if he didn't.

Thursday, October 11, 2012

CA-30: What's Howard putting in the water?

Howard Berman is racking up Republican endorsements:
Buck McKeon:
Republicans in Congress know that Howard Berman is one of the few Democrats we can partner with to advance bipartisan, commonsense legislation to deliver meaningful results for all of our constituents.
I don't get it. I've met Buck McKeon. Unless Howard Berman commits to reversing the defense cuts, Buck isn't going to be interested in weighing in.
The list:
Reps. David Dreier, Wally Herger, Dan Lungren, Elton Gallegly, Buck McKeon, Ed Royce, Jerry Lewis, Ken Calvert, Mary Bono Mack and Darrell Issa.
Interesting list. It has four retiring Republicans, three in safe seats, two in competitive seats, and one in a seat that's safe but Democrats fantasize is competitive. Berman is well known for his relationships on the Hill. So I can see retiring Republicans doing a favor for a friend.

But what about Dan Lungren? Why is Dan Lungren endorsing him? If I were Dan, I'd want an endorsement back from Howard, something that Berman can't give.

Of course endorsements on a web page don't win votes. Berman needs to use these endorsements to convince Republicans and right leaning independents to vote for him.

Monday, October 8, 2012


Pew has come out with an electorate that has swung 13 points towards the Republicans. The electorate was never going to be D+10 and it won't be R+3 either. The disturbing thing is how they can show such a huge change overnight. Maybe those people who've yelled party ID doesn't matter will change their tune.

Monday Potpourri

Interesting article on top two and how it's pushing some candidates in one party races to move to the center on issues. I expected this and any candidate who doesn't try to go after the other party is likely to lose.

Fresno Bee for David Valadao.

The very wealthy Pete Stark not only collects Social Security but his kids do too. And he thinks it'd be terrible if the government stopped giving money to rich people. Of course giving money to rich people means the Social Security trust fund will run out of money and increases the chances that those that need it in the future won't get it.

Stark won't debate Swalwell, but instead accuses him of supporting the Ryan budget and all but calls him a Republican. Swalwell is a Democrat and doesn't support the Ryan budget or probably almost anything else Republicans propose. Stark makes me want to contribute to Swalwell and I am definitely a Republican.

Friday, October 5, 2012

Suddenly everyone is a Republican

We Ask America is an independent owned by the conservative leaning Illinois Manufacturers Association. Their polls have been among those with the highest Democratic party IDs, however. Does that make them non-partisan or just funky? Well, last night they did three polls. Florida went from a D+5 electorate to R+5. Virginia went from D+4 to R+2. Ohio actually went from a D+2 electorate to D+4. Such a shift is unbelievable and the Florida electorate is more Republican than it's ever been. Party ID matters and party ID that favors Republicans too much matters as much as party ID that's too Democratic.

Thursday, October 4, 2012

Outside Congressional Spending in California

There are supposed be to 12 competitive congressional districts. When outside groups come in that means it's a competitive race or the group thinks it'll be competitive if they advertise. That second possibility shouldn't be dismissed and we should never assume that a PAC knows something we don't.

There are a lot more Democratic groups than Republican ones. That fits since the Democrats have been outspending the Republicans. The deluge of dark money the Democrats warned us about hasn't materialized. Unless they were talking about their own dark money.

The first interesting thing to note is that no outside groups have gone into CA-3, 16, and 21. We can assume that no one thinks it's worthwhile to invest. The challengers in these districts aren't regarded as strong as in other districts.

Everybody and their brother is in CA-7. I still believe Lungren is a favorite but others seem to disagree with me. The speculation has been that outside groups would go into where the parties don't. And sure enough, the DCCC isn't in CA-7. Still, you have to wonder if it's overkill. Do the Democrats pick up this district but fail in others due to light spending? A low number of groups, of course, doesn't mean little money. Some of these do have significant spending.

For the most part Republicans don't see a need for much of a defense in CA-36. The Democrats don't see a need to defend CA-9 and have decided to ignore Democratic leaning open seats. These districts do have a lot of Democrats but they are light voters. Considering the Democrats' lack of historical success in Riverside county I'd think they'd put money in CA-41.

Wednesday, October 3, 2012

Republicans Becoming Independents?

Democratic pollster Stan Greenberg claims the lack of Republicans in recent polls is due to the Republicans becoming independents.That's the idea generally floated on the left to explain that the samples are fine. There are problems with it.

1. Republican registration is increasing compared to Democratic in many states. What you're seeing here, e.g. a drop from 37% to 28% in Ohio, implies that 25% of the people left the party. You'd think you'd see some drop in Republican registration would show up.

2. This is true in California, but Democrats are also declining and the Republican drop is 1% of the total from 2008 from 31% to 30%. It's not 9%. Drops that sharp don't happen.

3. They are voting Democratic. We can take the 2008 Virginia voting and move the people that were Republicans and Democrats but are now independents according to Greenberg using this CBS/New York Times poll. If they voted 92%-8% Republican, Romney would lead 56%-44% with independents. Yet he still trails by one in this CBS News poll. So they'd not only having to be leaving the Republican party, but half of these McCain voters are now voting Obama. Anyone have the sense that this many Republicans have moved to the left in the last 4 years and are now voting Obama?

4. If they didn't leave in 2010, when Republican party approval ratings were much lower with Republicans why leave now that you've gotten all the people you wanted in there elected?

It's very possible that a small percentage have become independents, but nothing this massive. They can't say that Republicans aren't planning to vote, since their polls generally include 85-95% of the registered voters in their sample. And in most polls, Republicans do better with likely voters. The polls simply are missing these Republicans.

Monday, October 1, 2012

Swing States vs. National Poll

The Washington Post reports the widely held perception that Mitt Romney is underperforming in swing states compared to the national poll. According to their poll, Obama leads by 2 nationally and 11 in swing states. Of course they polled swing states as an aggregate and electoral votes aren't awarded that way.

The chart above shows the average Obama lead in each swing state and how much of a change it is since 2008. So while Obama leads Romney in all swing states in their poll, he leads by an average of 4.7 points in the swing states. The move is slightly more than how much Romney has improved over McCain. So Romney is not really doing any worse in swing states. If the margin is Obama by 2, as the Washington Post has the race, the national poll has moved 5.1 points, more than the 3.8 swing state change.

Obama overperformed in swing states in 2008. Romney will have to improve in swing states by more than the 7.1 points McCain lost by.