Get yer polls, bets, numbers & pretty graphs! Elections 2008

Checking in late for the party, I am, but sure looks like good news!

60% of the votes counted, and Obama is leading Hillary by almost as much as McCain is leading Huckabee! And that despite Hillary's flurry of last-minute campaigning in the state!

(It's Obama 56%, Hillary 43% now; and McCain 54%, Huckabee 38%.)

And yeah, those exit polls!

MSNBC: "[Obama] cut deeply into Clinton's political bedrock in Wisconsin, splitting the support of white women in Wisconsin almost evenly with the former first lady and running well among working class voters in the blue collar battleground, according to polling place interviews."

Hawaii should be reporting in another hour or so. It is 5pm there now.

G'morning Nimh.

'Mornin!

Interesting nugget from the exit polls:

Did either of these candidates for president attack the other unfairly?

37% Neither did
26% Only Hillary Clinton did
27% Both did
6% Only Barack Obama did

Or, to put it another way:

Did either of these candidates for president attack the other unfairly?

53% Hillary Clinton did
43% Hillary Clinton did not

33% Barack Obama did
63% Barack Obama did not

Another one:

No matter how you voted today, how would you feel if Hillary Clinton wins the nomination:

68% Very/somewhat satisfied
31% Very/somewhat dissatisfied


No matter how you voted today, how would you feel if Barack Obama wins the nomination:

82% Very/somewhat satisfied
17% Very/somewhat dissatisfied


There really seems to have been some kind of trendbreak after Super Tuesday.. (so far, I mean..)

I'm a proud member of that 17%. Ron Paul 2008!

The best two local results for Obama so far, I think, tell a story of two very different constituencies:

Dane County (incl. Madison - thanks for the heads up, Rjb), with 58% of precincts reporting:

65% Obama
33% Clinton

Milwaukee County, with 69% of precincts reporting:

64% Obama
35% Clinton

There's also two counties he's getting 60% in: Eau Claire and Door. Anyone familiar with the scene (Soz?) who can tell me something about that?

They just don't comprehend how sick and tired the voters are of negative campaigning. It has backfired on Clinton every time she's attempted it.

85% reporting and the numbers haven't changed much. This one is not going down to the wire.

She canceled an event in Eau Claire. Prob'ly didn't go over well.

Cycloptichorn

Hawaii Democratic Caucus

71% precincts reporting

Obama 75%
Clinton 24%

Wisconsin Primary

100% precincts reporting

Democrats

Obama 58%
Clinton 41%
Edwards 1%

Republicans

McCain 55%
Huckabee 37%
Paul 5%
Romney 2%
Thompson 1%
Giuliani 1%

OK, so I believe it now! :-)

I refused to go to bed late last night -- got all ready and then came downstairs at 9 to see if CNN did one of their "this is a blowout according to the exit polls" projections. The "CNN projection" screen came on, and I got all excited, but it was for McCain. Then went to bed. Just getting caught up. Looks like good news all around!!



I was in Dane county of course -- can't think of what Eau Claire County is (close to MN? I remember there was something about ads in MN outlets that are seen by Wisconsinites). Door County is a vacation/ touristy place -- skiing in the winter, lakes and leisure activities in the summer. Quaint. Unions, maybe? (Year-round residents would tend to be in service industries I'd think.)

Let's see. My notes. Obama did well in WI last night, gaining ground against Clinton in some demographic areas where he had been trailing:
lesser educated, lower income.
The Teamsters union endorsed Obama after polling their membership. They have no money, this shrinking union representing some 1.3 million blue-collar or less workers in the U.S., but 60,000 of them are in Ohio.
The Teamsters tend, I believe, to toe the union line. Ohio could turn out to be close. Cyclo or Nimh may have posted a graph about that.

The secret weapon Obama has in Ohio is Sozobe.

In TX, Clinton is relying on the Hispanic vote, which is a very significant portion of the population. If she gets more than, say 50%, she could win the state. 25% of the voters there are black. Obama should do well but if Clinton can take just some of them, say 15%, her cause is helped. But that is something like two times the % she got in CA.

There are expected to be some 2,000,000 folks participating in the Dem primary in TX. As has happened elsewhere, many will be participating for the first time, which probably favors Obama.

One refrain I heard was "I am fed up with Bush and I am fed up with the status quo."

126 Dem delegates will be awarded in the primary (but in a weird process that might favor Obama) and 67 will be divided in a caucus-like event later on March 4th.

All in all, it seems to me that the stars are alligning in Obama's favor right now. But things can change.

:-D

I'm trying to corral the Columbus Deaf vote. Not huge, but something, and moderately successful so far (just sent something out to a Deaf listserv with several hundred subscribers).

About to get my yard sign -- nobody in the neighborhood has one (at all, any candidate) so that's a little nerve-wracking, but worth it.

Bringing supplies to the Columbus Obama office (they need paper towels, among other things).

Haven't heard back re: that job so probably don't have it but oh well.

Talking to main person in the [My area] for Obama group about setting up a local fund-raiser/ informational meeting.

Building up some steam...

Obama opens up lead on McCain in swing states; Hillary's electability fading

A new spate of state-level Rasmussen polls that match up both Obama and Hillary Clinton against McCain has yielded impressive results for him - and bad ones for her.

Rasmussen polled ten states this month, all roughly in the swing state category (Oregon was the most Kerry-friendly one, Missouri the most Bush-friendly). In all but two of those states, the polls have Hillary trailing McCain, sometimes significantly (-9 in Nevada, -14 in Colorado).

In all but two of the states, these polls have Obama leading McCain, sometimes overwhelmingly (+20 in Minnesota, +16 in Florida, +13 in New Hampshire).

Rasmussen polls help Obama look newly strong against McCain

These new Rasmussen polls certainly change the colouring of my Excel sheet. In the case of the column in which I "balance" the available opinion polls for that state, in which I give extra credence to recent polls, they make a drastic difference:




(Click to enlarge, you may have to click once more when the window opens to get it in proper size)

For example, the Rasmussen polls in Michigan, New Hampshire, Nevada and Colorado are the first polls out in these last four/five months that match Obama up against McCain in those states, so they alone determine the "balance" column - and make Obama look pretty good there.

For other states, some comparisons are possible. Unfortunately, there's just one state that Rasmussen polled Obama's match-up numbers in previously as well: Pennsylvania, which it polled last month too. Since then, Obama turned around a deficit of 8 points into a lead of 10 points.

This may seem too drastic to be plausible, but keep in mind just how volatile comparisons over time of leads/deficits, rather than of each candidate's individual polling, by definition are. Basically, the margin of error doubles in size.

New SUSA and Quinnipiac polls more ambiguous

Is such a drastic turnaround in Obama's numbers reflected by other pollsters too?

Quinnipiac polled the Obama vs McCain match-up in four states this month; it has the two candidate roughly tied in Pennsylvania, Ohio and Florida, and Obama up 7 points in New Jersey. Unfortunately, those are all states where they last asked this question in October, if even that. Compared to back then, Obama's had to retreat somewhat in PA and OH and stayed stagnant in FL, but this is no surprise; that was before the McCain boomlet in match-up polls in December-January, which was already flagged in my graph on the national match-up polls.

Survey USA has so far asked the match-up question in six states this month, with varying results. In New York, Obama almost quadrupled his lead over McCain, from 6 to 21 points; in Washington state, he almost doubled his lead, from 9 to 17 points; and in Virginia he turned around a 12-point deficit into a 6-point lead. But in Iowa, the drastic 17 point lead he opened up over McCain in January shrunk again, and in Kentucky he went from being down 19 points to being down 29 points.

What these numbers might show, beyond the inherent volatility of comparing leads in match-up polls, is the effect of Obama's campaigning. His huge 17-point lead in Iowa in January will have reflected the extensive campaigning residents had seen of his there. Now, he's made great advances in Virginia, Washington and New York - three states in which he's campaigned this month. It could be coincidence, but it would fit the overall pattern that whereever Obama campaigns hard, his numbers rise in the polls significantly - even, note, if it's in states in which his opponent, be it Hillary or McCain, also campaigned.

So what's the score?

Taken together, the SUSA, Rasmussen and Quinnipiac polls (with three polls by other pollsters thrown in the mix) have switched the colour in my "balance" column for Obama vs. McCain significantly. Here's the new "balance column", based on these last polls as well as all the previous polls since October:





Compared to before this month:

• Rhode Island, New Jersey, Michigan, New Hampshire, Nevada and Colorado, all previously blank, now show up as "strong Dem" for Obama;
  
• New York, Washington State and Minnesota change from "lean Dem" to "strong Dem";
  
• Pennsylvania and Florida switch from "lean Rep" to "lean Dem";
  
• Ohio and Virginia move from "strong Rep" to "lean Rep";
  
• Indiana, previously blank, now shows up as "strong Rep".

So what's the score? Given the polling numbers available on state-level for this match-up, Obama would win, in comparison to Kerry's 2004 score, Iowa, Nevada, Colorado and Florida, and lose, implausibly, Massachusetts. Not bad, and certainly better than what it looked like even just a month ago.

Note the column in which I've listed how many polls the colour is based on though! If there's just one, well, then it means less than if there are five, obviously.

continued from above..

What about Hillary?

Looking at Hillary's prospects, it's again the Rasmussen polls that are driving the picture. As noted above, in the ten states Rasmussen polled this month, Hillary trailed McCain in all but two, and sometimes significantly (-9 in Nevada, -14 in Colorado).

In nine of the ten states polled by Rasmussen, Hillary matched-up worse against McCain than Obama. The single exception was Missouri, the only Southern state apart from Florida in the selection.

This is clearly very bad news for her campaign, especially since these are all swing states. But a deterioration over time can not be verified directly, because of how Rasmussen didn't poll the same question any time recently in most of these states. In Pennsylvania, the only state which Rasmussen polled last month too, she actually made up a few points, even if she still trails McCain. Compared to last November, she's stable at a moderate distance behind McCain in Florida, while she went from trailing closely to being trounced in Colorado.

No comfort from Survey USA

Survey USA's match-up polling this month, however, can be compared over time, and is barely more encouraging. It shows her leading McCain by 11 in her home state New York, which is the smallest lead for her in this chart yet; her lead decreased from 39 points in October to about 20 points in November and December to 13 points last month and 11 now. It shows her tying McCain in Washington, which is in line with earlier SUSA polling of this race; trailing McCain by 3 points in Virginia, which is also roughly in line with the up-and-down of previous SUSA numbers; and trailing McCain badly in Iowa, Kentucky and Indiana. The Iowa numbers represent a distinct deterioration, and the Kentucky ones a gradual decline.

On top of that, SUSA too has Hillary matching up distinctly worse than Obama in the swing states Iowa and Virginia, as well as in Washington - and in her own home state New York! She does better than Obama in Indiana (barely) and Kentucky (greatly), but since she still trails McCain badly enough herself too in both those states, the relevance of that is questionable.

Quinnipiac polling on her performance against McCain is more ambiguous. It has her losing narrowly in Ohio and Florida, and winning comfortably in New Jersey and Pennsylvania. In Ohio, Florida and New Jersey that puts her score about equal with Obama's, while she does better than him in the Pennsylvania poll.

Finally, three other polls have her losing by 5 in Wisconsin, tied with McCain in Pennsylvania, and winning by 11 in Rhode Island. Her Wisconsin score is slightly worse than Obama's, while the other two have both candidates doing about equally well.

So what's the score?

So what about the score for her? Here's my current "balance" column for Clinton vs. McCain:





In short, given the polling numbers available on state-level for this match-up, Hillary would not win a single state that Kerry lost; while she would lose Washington and Oregon, with Maryland and Minnesota on the brink.

Again, note the column in which I've listed how many polls the colour is based on, though. And realise that some of these polls are by now quite dated. But you deal with what you have, and please - do click on the thumbnail in the post above of the more detailed table. That way you can see all the individual polls, and make your own judgement about what the overall picture of all these polls is.

I don't know exactly how implausible that really is.

When McCain ran in 2000, MA was his biggest win in the primaries (he took something like 65% of the votes). For whatever reason, people in MA really like him. I have no idea why.

And Obama's campaign is errily similar to the campaign that was run by our Gov. "Keep Hope Alive" Patrick. Patrick has only been in office for a year now so it's hard to gauge how well he's really doing but he hit a lot of bumps at first and most of his campaign promises have been road-blocked by career politicos in the State House (~80% of the MA State House are Dems as Patrick is...). With a lot of people I've talked to around here the view on Obama is that he'll run into the same sort of thing if he's elected so there is a fair amount of skepticism on his "Change" message.

I will say though, that most of the people in the area that I've discussed politics with aren't really happy with any of the choices at this point.

Interesting, Fishin'.

Here's the last installment of my "article," so to say:

The electability comparison: the regional strengths and weaknesses of Obama and Hillary

Overall, then, the numbers are quite clear - the electability gap Obama had faced on state levels up through December has disappeared, even reversed. There are still regional differences though.

In the month of January, the performance of Hillary and Barack in the match-up polls had already turned more or less even, overall -- but it wasnt even across the board. Obama was matching up better than Hillary against McCain in Maryland, Illinois, Iowa, Washington and Oregon. Hillary was matching up better in Massachusetts, New York, Ohio, Oklahoma, Alabama, Kentucky and North Carolina.

Clear regional patterns there: he outperformed her in the Northwest, she outdid him in the South. Perhaps worryingly, almost all the states he outperformed her in were already blue states, while she outdid him in a number of red states. If you look at polls pitting Hillary and Obama up against Huckabee as well (I've skipped those data here), these patterns became even clearer.

Now in February, the balance has become thoroughly slanted in Obama's favour. Obama matched up better than Hillary in states across the country - NH and NY; MI, WI, MN and IA; CO and NV; WA and OR; and VA and FL.

The two did about equally well in RI and NJ; PA, OH and IN; and MO. And she did better than him in KY. (That's it.)

Again, because of the inherent volatility in individual polls, the differences that exist in methodologies, and the particular up and down that applies when comparing leads, one shouldnt attach too great a significance to the datapoint per individual state. But overall patterns can still catch your eye.

For example, the West is really not friendly territory for Hillary, California excepted. Obama does much better there. His relative strength now also extends into the Western reaches of the Midwest (IA, MN, WI and IL - and the one poll in February that still matched up the Democrats against Huckabee as well also showed Obama as the bigger winner in Wisconsin).

In the South, data is getting more mixed now (see VA and FL), but I'd still give Hillary the electability edge there, with Kentucky as the prototypical example. In the Northeast the balance is increasingly even.

The map also looks more favourable for Obama strategically now. Of the 12 states in which he matches up better than Hillary, nine are among the swingiest of swing states, where Bush and Kerry were wthin five points of each other: OR, MI, MN, NH, WI, IA, NV, CO and FL. Whereas of the seven states in which she holds her own, just two are in battleground central (PA and OH).

Make of those regional differences what you will, but they're interesting. I'll follow up at the end of the month.

Great analysis, Nimh.