A Man in the Arena

Why McCain needs to invoke the spirit of Theodore Roosevelt

As I have said elsewhere, it is obvious from the polls that that McCain’s attempt to run on a hard right tax cutting agenda isn’t working, and more importantly is something that McCain doesn’t believe in. In reality McCain is a right of centre populist who believes in free trade, school vouchers and attacking wasteful spending but also believes in regulatory reform, global warming, expanding access to health care and is worried about inequality. He also supports issues such immigration reform, energy independence and balanced budgets which are neither right nor left. Just as his best moment on foreign policy came when he directly confronted Ron Paul, his best two domestic policy moments came when he attacked Romney (and implicitly Giuliani) over the ‘profit not patriotism’ line and Huckabee over his regressive tax plan.

The Democratic lead on economic issues and their lead in the generic ballot means that there is simply no popular appetite for further economic conservatism. Obviously, Dick Morris is the last person who should be giving policy advice and his stuff about oil speculators and companies is simply demagogic nonsense. However, his contention in a recent interview that, ‘McCain can win by entering the democratic primary. He has to be a democratic opponent to to Obama. He should be a populist, and win over (working-class) Hillary voters’, is correct. It is also self-evident that the Mitt Romney and Fred Thompson wing of the Republican Party, who seem to have hi-jacked the McCain campaign, seem hell-bent on blaming everyone else, including the very people that McCain needs to win over, for his predicament. For instance, even the normally reasonable Irwin Seltzer seems to have gotten into his head that the one person to blame for Bush’s failures in the past two years is not Bush himself but Joe Lieberman.

So what should McCain do? I think the first thing McCain needs to do is to stop trying to be a third rate Barry Goldwater. Even though Obama has tried to give the illusion of moving to the centre there is plenty of room for McCain to come up with a tax plan that is a bit more progressive and which closes more loopholes, while still keeping economic conservatives on board. McCain should also stop hedging, instead of talking about building a few more nuclear power plants in the next twenty years he should talk about building a whole generation by 2012, a goal that is achievable if foreign technology is allowed to be imported into the US and Yucca Mountain is opened up. As I have said before McCain also needs to make overtures to those on the DLC wing of the Democratic Party. Although this may seem radical, it would be no different from the Republican recruitment of hawkish Democrats, such as Richard Perle and Paul Wolfovitz in the mid to late 70s. I am not suggesting that McCain should go overboard or do anything silly, but moving the Republican party to the centre, by enabling conservative Democrats to vote for McCain, would do more to help protect the idea of free trade and market solutions than anything the NRO wing of the GOP have come up with so far.

Obama’s lead returns to 5.90%

Obama’s bounce persists

My new national projections are:

Barack Obama 48.95
John McCain 43.05

Unlike my previous projections, there is no way that these results can be spun. Although, in time McCain’s ratings might start to catch up with those of Obama this is completely unecessary risk. McCain’s campaign has been poorly run with it increasingly reminding one of the dog in James Thurber’s short story, afraid to attack Obama because of a perceived backlash, afraid to move to the centre because that would offend the ‘base’ and most importantly afraid of mentioning the war in Iraq, even though McCain actually leads on this issue in the polls. Indeed, in terms of economic policy you could be forgiven for questioning where the imprint of Thompson and Romney ends, even though McCain has had a distinguished record fighting for the interests of hard working Americans. On domestic issues McCain desperately needs a declaration of independance.

NOTE: My methodology for my detailed polling projections involves collating all the available polls of likely voters and then filtering them through Samplemiser, a Kalman filter. I only used polls of likely voters and I get my data from pollster.com.

Huckabee leads McCain by 1.29%

The Palmetto State Looks to be edging Huckabee’s way

Taking all the polls ending on January 6th to today inclusive, my projection is: Mike Huckabee 27.91, John McCain 26.62, Fred Thompson 15.37, Mitt Romney 13.19, Ron Paul 3.21 and Rudolph Giuliani 3.04. 10.66% of voters are still undecided.

Looking at the filtered data I am struck by the fact that many of the undecideds are going to Fred Thompson. Although I don’t like to make Kerry-esque predictions that cover both eventualities, or to contradict my own projections, my gut feel is that Fred Thompson will get around 21% of the final vote, because he is the only candidate who has a strong (-0.51) negative correlation with undecideds (indicating that they will probably go to him). However, this won’t be (in his words) enough to get him a ticket to the dance in Florida. Given that Mike Huckabee’s dismal results in Michigan, and his decision to imitate the rhetoric of Richard Santorum, has eliminated his claim to be the electable face of Social Conservatism, I predict that two candidates left standing will be John McCain and Mitt Romney.

McCain leads Huckabee by 10.26%

The Senator from Arizona leads Huckabee by over 10% in national polls

Taking all the polls ending on December 30th to January 13th inclusive, my national projection is: John McCain 31.46, Mike Huckabee 20.20 Rudolph Giuliani 11.66, Mitt Romney 11.43, Fred Thompson 10.42 and Ron Paul 3.4. Just over 11% of voters are still undecided.

Some reflections on Michigan

TPT reflects on Romney’s victory

1. I really should not get involved with trying to second guess my projections. My decision to project Romney ahead but predict a McCain victory was a ‘Kerry-esque’ stunt too far. Although political betting is about anticipating changes in polling figures, there is a limit to the value that such analysis can add on election night.

2. This is a great night for Romney and a bad one for McCain, Huckabee and Giuliani. South Carolina is pretty much a three-way contest between Huckabee, McCain and Romney, with the latter two needing to come second to stay in the contest (Huckabee needs to win or he is finished).

3. McCain really needs to find his voice on economic issues. Although he is constrained by the need to appear more conservative than he actually is, he needs to go after Huckabee for his support of the ‘fair’ tax and Romney for his tax-dodging and incompetent management of a public works project as Governor.

4. McCain also needs to stop being so complacent. He cannot just hope to coast through the primaries. He is already taking hits from Rush Limbaugh and the people who are attempting to smear his record are starting to get some attention from the mainstream media.

5. Although the number of crossovers were reduced by the cold weather, McCain (and the Republican) party needs to move to the centre on economic issues. They are not going to hold every single state so they can’t simply rely on national security and Iraq.

6. Giuliani is dead. He’s finished behind Paul in Iowa and Michigan and behind Huckabee in New Hampshire. The fact that he fell behind the non candidacy of Fred Thompson was the final nail in the coffin.

7. If Huckabee is nominated he’s going to get crushed. While McCain has had to tone down his economic populism, Huckabee went in there with all his guns blazing. 16% of the vote indicates that if he gets the nomination the Republican party will get creamed.

Romney leads McCain by 0.95%

TPT’s final predictions for Michigan

With the Detroit News and Mitchell having produced polls that end on the 12th, my projection is: Mitt Romney 29.57, John McCain 28.62 Mike Huckabee 15.46 Ron Paul 7.04, Fred Thompson 4.42 and Rudolph Giuliani 3.04. The undecideds have shrunk to 11.85%. My guess is that pollsters are underestimating McCain’s support from Democrats and Independants, so I’m going to cross my fingers and give John McCain a 55% chance of winning tonight. It goes without saying that if Giuliani comes behind Ron Paul and Fred Thompson (or even one of those candidates) then he should pack up and endorse John McCain (though I can see Dick Morris getting him to endorse Huckabee). For the record my projections with the undecideds are (assuming that Duncan Hunter gets 0.5% of the vote and the other undecideds split proportionately) are: Romney 33.38, McCain 32.31, Huckabee 17.45, Paul 7.95, Thompson 4.99, Giuliani 3.43 and Duncan Hunter 0.5.. My guess is that Hillary has a 65% chance of getting the 50% she needs for a moral victory and a 75% chance of winning outright against uncomitted.

McCain leads by 4.54%

The contest in New Hampshire is still very close

Inputting the 10 (!) polls that have been published today the following projections are produced: John McCain 33.52 Mitt Romney 28.98 Mike Huckabee 11.43 Rudolph Giuliani 8.83 Ron Paul 7.31 Fred Thompson 2.86. It is no exaggeration that the future of both the Republican party, America and possibly the West hangs in the balance.

McCain leads by 4.86%

… but is it enough?

Updating my collection to take account of the last 4 polls and to delete overlapping tracking polls (and to add non-overlapping historic polls) I have come up with the following projections: McCain 32.91 Romney 28.05 Huckabee 11.28 Giuliani 8.77 Paul 8.31 Thompson 2.28. Although some of these polls were conducted over the past few days the Rasmussen poll agrees with the idea of McCain having a 5% lead. A 0.36% bounce from Iowa is pretty underwelming, although I do hope that it will increase in the next few days.

McCain leads by 4.04%

The Senator from Arizona leads Romney in the last of the pre-Iowa polls.

Incorperating the last of the Iowa polls into a compilation of polls from December onwards produces the following projections (courtesy of Samplemiser): McCain 32.2 Romney 28.16 Huckabee 11.7 Giuliani 8.8 Paul 7.78 Thompson 1.73. The race is extremly close and McCain should not assume that it will be a walkover. In fact I’m going to guess that McCain will (unfortunately) not win by more than 5%. The raw data is here.

Updated Projections: Huckabee leads by 5%

Huckabee leads Romney, McCain fallls into fourth

Inputting the newest ARG and Zogby polls, deleting the Zogby poll that ended on January 2nd and replacing it with the one that ended on the 29th the final projections are produced: Huckabee 30.15 Romney 24.88 Thompson 11.76 McCain 10.50 Paul 8.37 Giuliani 6.76 (as of 4:15pm GMT). McCain seems to have to have lost most of the boost that he has gained from the DMR endorsement but he is still (for now) ahead of Giuliani and Paul. The updated projections are here.

Note: I initially forgot to remove the Jan 2 poll, as a consequence Huckabee’s lead is 5.27% not 5%.

Iowa comes down to the wire for the GOP

…but Huckabee has the edge.

Inputting the Zogby tracking poll to my collection (and ignoring all previous tracking polls because they overlap with the last one) produces the following projection (as of Noon GMT): Mike Huckabee 28.23 Mitt Romney 26.41 John McCain 12.13 Fred Thompson 11.91 Ron Paul 8.84 Rudolph Giuliani 6.89 (Duncan Hunter and Alan Keyes are not included because their support is marginal and they have not be included as an option in all the polls). The margin of error is just under 2% so this is a statistical dead heat for first and third. The numbers undecided have gone down to 5.59% and my prediction is that they will go to Mike Huckabee and Ron Paul (I would guess in a ratio of 4:1) since they are the only two candidates to have a negative correlation with the undecideds. I will update the figures if new polls come out later today but I have a horrible feeling that Romney will beat Huckabee and McCain will be pipped by Thompson.

(The raw data and the Samplemiser projections are here)

Where does the value lie in the Primary betting?

Why John Edwards and John McCain represent value bets

 

After looking at the polls I have decided that John McCain should be the Republican frontrunner. All the other Republican candidates have major flaws and I am beginning to get the feeling that people are starting to coalesce around McCain as a candidate who can unite the Republican party – and possibly even America as well. The only question that remains is whether McCain can do well enough in Iowa to give him a strong shot at New Hampshire, which is a must-win state for him. Although his mini-surge in the polls has raised the bar for him (with a finish in double digits ahead of Paul and Giuliani an absolute minimum) I believe that he can pull this off. My guess is that McCain’s chance of winning the nomination is at the very least 35%, while intrade.com has him at 23.1-23.6.

On the Democratic side I believe that John Edwards is a value bet. My reasoning is that a large proportion of Democrats distrust Hillary Clinton, or believe that ultimately she is too divisive. At the same time it is blindingly obvious that Barack Obama is too inexperienced, too immature and that he will be eaten alive by the Republicans. Despite doing the best to run his campaign in the ground he represents an effective ‘none of the above’ option, as well as the memory of a smart, charismatic Senator who was hawkish and managed to mix populism with a upbeat centrist message as late as 2004 (which helps him attract centrist voters). I believe that Edwards has at least a 1 in 3 shot at Iowa and that if he wins he can go toe to toe with either Obama or Clinton. My prediction is that he has at least a 20% chance of winning the nomination while Intrade has him at only 7.0-7.8.

Neither of these bets are ‘sure things’ (and indeed I guess that the probability of neither of those bets coming off is more than the 50% a probability tree would imply since Edwards might bleed some support from McCain in New Hampshire). However, if you believe in value betting they are very good bets. The only other bet that might be useful is putting some money down on Fred Thompson at 2.4-2.9 just in case he comes back from the dead.

Huckabee leads by 2.24% in Iowa

 Huckabee is first while Giuliani is last 

Combining the last three polls that ended on the 30th and putting them into my collection produces the following projections: Huckabee 29.71 Romney 27.47 McCain 12.34 Thompson 8.84 Paul 7.75 Giuliani 6.47.

Romney leads in Iowa with McCain in third

Romney leads Huckabee by 0.42%.

Incorperating the CNN poll into my collection and filtering it through Samplemiser produces the following projection;

Romney 28.32 Huckabee 27.9 McCain 11.11 Thompson 8.29 Giuliani and Paul 7.69 each

. This strongly suggests that Fred Thompson is really out of contention and we can expect to see him withdraw (and endorse McCain).

Iowa polls remain in flux

Can we really trust the last few polls?

Adding the Research 2000 polls to the mix and then filtering them produces the following projections: Huckabee 31.23 Romney 26.76 Thompson 12.68 McCain 11.17 Giuliani 6.01 Paul 5.05. Certainly, the recent polls have confirmed that the contest between Huckabee and Romney is pretty tight and that McCain has been boosted into at least fouth place. However, unless Thompson’s Christmas adverts have greatly boosted him, the last three post-Christmas polls contradict the ARG polls.