Sept. 11 Commemoration

Let's have 1 minute of silence to remember the people who died on Sept. 11, 2001 in the terrorist attacks.

Voter Registration

While most of the media attention is focused on candidate misstatements and national polls, there is important stuff going on below the radar. In particular, voter registration. Both sides are doing it, but the Obama campaign is doing it on a scale no party has previously done. According to an article in Time, voter registration is up by over 400,000 in Florida, Michigan, and North Carolina, and up by over 100,000 in 10 other key states. Most of the people being registered are under 35, a key Obama demographic. These huge increases could determine the election in a number of swing states.

As an aside on voter registration, if you are one of the 7 million Americans living abroad (including students) and are over 18, you are eligible to vote in the state you last lived in. Click on the banner ad at the top of this page to go to VoteFromAbroad.org to register to vote.

Electoral-College Graph in a Different Format

A new site, evstrength.com, displays the electoral vote data in a somewhat different format than the "Electoral coll. graph" listed above. It makes the strong, weak, and barely states more visible.

Polling and Partisan Identification

A series of recent polls have shown John McCain gaining ground on Barack Obama. No doubt part of that is real due to the Republican convention and especially the speeches by Sarah Palin and McCain himself. But there is also a troubling issue here relating to partisan identification about which pollsters themselves strongly disagree. There are two schools of thought. Gallup, SurveyUSA, and many others say in effect: "If we sample 1000 people and 500 are Republicans and 500 are Democrats, that's the way it is." They just report the raw numbers--say 50% McCain and 50% Obama. Other pollsters, such as Rasmussen say: "I know there 42 million registered Democrats and 31 million Republicans in the country. Furthermore my own samples over the past 3 months show (for example) that 55% percent of the people I've called are Democrats and 45% are Republicans. So if I happen to get a 50-50 sample one day, I am going to weight the 500 Democrats as if they were 550 and I am going to weight the 500 Republicans as if they were 450. Thus even though the raw data said Obama 50% McCain 50% I am going to report that as 55% to 45%."

The argument for the first approach is that you are not fudging the numbers, just reporting what you measured. The argument for the second one is that apparently today we had a bad sample and pretending that half the country is Republican doesn't make it so. The argument about some of the recent national polls is that the number of people calling themselves Republicans is far higher than in previous months. A good Republican convention may cause more Republicans to plan on voting and may sway independents and even a few Democrats, but it almost never changes partisan composition. Polls that don't correct for partisan identification may thus undersample or oversample either party. Correcting for it eliminates this problem but introduces the problem of knowing when to change the partisan weighting. Rasmussen uses long-term averages over months to change the weighting.

If you are wondering about what's going on with these recent polls, here is an article from Seth Coulter about this subject and here is a reaction to the article from Nate Silver. Brian Schaffner has yet another article on the subject, this one applied to the recent SurveyUSA poll in North Carolina (where the partisan breakdown was 40-41 compared to 46-33 in August).

A different but related subject is dealing with the partisan pollsters. Firms like Gallup and SurveyUSA make their money by running polls for newspapers, TV stations, businesses, and others who want to know what the public is thinking. However, there are other firms that are basically consultants, generally to only one party, and whose job is helping to elect its clients' candidates. They may shade or manipulate the results, use biased questions, or selectively publish results (such as only those favorable to their clients). In general, they are best ignored. That is why we consistently omit certain pollsters here, such as Strategic Vision (R) and PPP (D).

What If the Electoral College is Tied?

Many people have asked: "What happens if the electoral college is tied at 269-269?" For example, look at the map of Aug. 24. If it ends up like this and McCain wins Virginia, the final result is 269-269. What next?

If no candidate for President receives 270 electoral votes (either due to a 269-269 tie or a third-party candidate has won a few electoral votes), the newly elected House chooses the President, with each state having one vote. Some people have asked what the House is likely to look like in January 2009. Lets start with the House as it is right now.

The blue states have Democratic majorities. The red ones have Republican majorities. Arizona and Kansas are split evenly. Click on the map for an interactive version.

Currently the Democrats control 27 state delegations and the Republicans control 21. Arizona and Kansas are evenly split and would probably be unable to decide how to vote. This assumes all representatives vote the party line. Any Democrat voting for John McCain would no doubt be punished swiftly and vigorously by Nancy Pelosi, which is probably worse than potential punishment from the voters in 2 years, by which time other issues may dominate the news.

Of course, it is the new House that gets to vote, not the old one. What changes are likely? Let's look alphabetically. Alaska only one congressman, currently Don Young (R), but he is facing a tough primary and the winner of the primary will face a difficult race against Ethan Berkowitz (D). If Young wins the primary, Berkowitz has a decent shot at beating him in the general election. Arizona is current split 4-4. If the Democrats hold their four seats and pick up the open seat in AZ-01 being vacated by Rick Renzi (R), they would control Arizona. It would certainly be embarrassing to have Arizona to vote against McCain for President, but there is a good chance this could happen. Thus we have two states that might go Democratic and we haven't even gotten to "B" yet.

Kansas is also split, 2-2, and there is a very competitive race in KS-02, with Rep. Nancy Boyda (D) facing state treasurer Lynn Jenkins (R). Jenkins could win this one, which would make Kansas go Republican. In Michigan, two Republican incumbents are facing very tough battles, Tim Walberg in MI-07 and Joe Knollenberg in MI-09. If the Democrats win both of these they will control the Michigan delegation. Mississippi is currently Democratic, but Travis Childers (D) is in a rematch with Greg Davis in MS-01. Childers already beat him twice this year in special elections, but the turnout will be heavier in the general election. If Davis wins, Mississippi, the state splits 2-2. In Missouri, Sam Graves in MO-06 has the fight of his life this year against former Kansas City mayor, Kay Barnes. If Barnes wins, the Democrats will control Missouri.

Nevada could also change. Jon Porter (R) in NV-03 has a real battle against state senator Dina Titus (D). If Titus wins, Nevada flips to the Democrats. New Mexico is another state where changing one seat flips the delegation. The race to watch here is NM-01, a D+2 open seat vacated by Heather Wilson in her abortive run for the Senate nomination. Initial polling here puts Martin Heinrich (D) ahead of Darren White (R) here, which means the state could flip.

Ohio currently has 11 Republicans and 7 Democrats in the House. Two of them, Steve Chabot (R) in OH-01 and Jean Schmidt (R) in OH-02 are facing very strong challengers and could lose. Two more seats, OH-15 (Pryce) and OH-16 (Regula) are open and the Democrats are fielding very strong candidates. Winning two of the four is a realistic possibility, in which case Ohio would not be able to vote. There is one Democratic-held seat in Ohio (OH-18) which is theoretically at risk, but incumbent Zack Space (D) lucked out when the Republicans failed to come up with a first tier candidate. The Republican, Fred Dailey, has never run for elective office before, so despite the R+6 rating, Space is favored.

So in the best case for Obama, Boyda and Childers hang on and the Democrats pick up Alaska, Arizona, Michigan, Missouri, Nevada, and New Mexico, while neutralizing Ohio. This would give them control of 33 state delegations. What's the best case scenario for McCain? Boyda and Childers could lose and one of the freshmen in Arizona (Giffords or Mitchell could lose). This brings it to 26 Democrats and 24 Republicans. Still enough for a narrow Obama win.

However, all this assumes that all representatives vote the party line. But consider Stephanie Herseth Sandlin (D-SD) and Earl Pomeroy (D-ND). The Republicans carried both states by over 20 points in 2004 and 2000 and probably will again. What will they do? Pomeroy is 56 and his been in Congress for 8 terms, so he qualifies as a senior guy who won't be pushed around easily. Herseth Sandlin is only 37 and this is only her second term. Furthermore, she knows very well that she got her job because many voters fondly remember that her grandfather was once governor. There will be immense pressure on the poor woman to vote the way her state did. If both of them vote for McCain, he becomes President. If they split and it becomes 25-25, then the House will unable to elect a President. In that case the Vice President elected by the Senate (with each state having two votes, one per senator) acts as President. The Democrats are likely to hold a minimum of 55 seats in the new Senate, so in that case Joe Biden would become acting President. This scenario would also mean that if Pomeroy votes for McCain, then Herseth Sandlin could make McCain President or Biden President depending on her vote. Ditto Pomeroy if Herseth Sandlin votes for McCain. And by the way, Rep. Mike Castle (R) is the sole representative from heavily Democratic Delaware, a fact Joe Biden will surely bring up if it gets this far. The whole thing could make 2000 look like a clean-cut victory.

With a close presidential election, half a dozen House races could ultimately determine who becomes (acting) President. For a discussion of about 60 key House races, click on the Hot House races link on the menu.

Today's Polls

We have eight new presidential polls today. There is good news and bad news for both candidates. For Barack Obama, the good news is that he is still ahead in Michigan and New Hampshire, two Kerry states he really can't afford losing. If he holds the Kerry states, he starts with a base of 252 electoral votes. To that he needs 1-3 more states. The good news for John McCain is that for the first time this year he is ahead in New Mexico, a state that Bush won in 2004 but which Obama was really counting on winning this time. His 49% to 47% lead is a statistical tie, but at least he is not far behind there, as he has been all year.

We also have three Senate polls as follows. Ted Stevens seems to be narrowing the gap in Alaska, despite his indictment and upcoming trial. Perhaps he is riding on Sarah Palin's coattails.

Finally there is also a poll in WA-08, a hotly contested race.

If you like this Website, tell your friends. You can also share by clicking this button