Comment, news and views from the US Elections.
Susbscribe to an XML feed of this blog at:
http://timesonline.typepad.com/uselections/rss.xml
 With early voting closing across the country, the numbers are looking good for Barack Obama - and for American democracy. Some states have seen  record levels of early voting - indicating that tomorrow's participation may surpass already towering expectations. In four key battlegrounds - Colorado, Nevada, New Mexico and North Carolina - early votes have equalled at least two-thirds of all ballots cast in 2004, while in Florida and Georgia more than half of the 2004 totals have been surpassed - 54 and 61 per cent respectively.
In all, almost 29 million Americans have already cast their ballots and among these, more identify themselves as Democrats than Republicans, in some states by large margins. In Florida, 46 per cent of those voting early are Democrats while 38 per cent are Republicans - in 2004, 44 per cent of early voters were Republicans and just 41 per cent Democrats. In Louisiana, 56 per cent of early voters are Democrats compared to just 28 per cent for the Republicans. This does not necessarily translate into a vote for their respective candidates - but the likelihood is that the figures will be similar. For a full breakdown of the statistics, read this report from Dr Michael MacDonald of George Mason University.
The blog X Curmudgeon also has a handy compilation of polling conducted among early voters - there are a number of organisations currently focusing on this category. (Note - these results are not from exit polls as is stated.) The results vary widely but all give Obama a healthy lead; the majority indicate margins well into the double digits (today's CBS/New York Times poll, for example, gives Mr Obama a mammoth 17-point edge).
These figures represent a sharp reversal from the early voting advantage Republicans enjoyed four years ago, when President Bush won all six battleground states. If they are borne out tomorrow, the Democrats could see that longed-for landslide - if not, American democracy will be a winner at least.
Our ever-informative friends over at the Gallup polling organization have an excellent look back at elections of years past and what the polls looked like with a week to go. Their conclusion is that late comebacks aren't without precedent, but they certainly don't happen often.
In fact, only twice since 1952 have the polls shown a candidate leading this late in the race who did not go on to win the popular vote.
Jimmy Carter famously led Ronald Reagan by 8 points with less than two weeks to go, but their final debate and the 1-year anniversary of the Iran Hostage Crisis fell in that span and certainly contributed to the Republican's momentum. Alternatively, George W. Bush led Al Gore heading into election day in 2000, only to lose the popular vote while winning the Electoral College and the Presidency.
Outside of those two instances, the polls have largely held up, although they tend to tighten considerably leading up to election day. Look for John McCain to try to create a game-changer in the next few days, albeit without the benefit of the hostage crisis.
 The gap is narrowing, insists the McCain campaign, pointing to Sunday polling data which suggests an erosion of the lead held by Democrat Barack Obama. But is that really true?
A cursory comparison between last week's galloping 14-point lead in a New York Times survey and the meagre 3-point edge seen in Gallup tracking data yesterday would seem to suggest it is. But in the numerical minefield that is the American political polling industry, one must dig beyond the screeching headline figures to get a clear picture - or at least the clearest available.
The key thing to be aware of is the multitudinous nature of American polling organisations. Unlike in the UK, where polls are conducted by just two or three large outfits, there are scores of smaller bodies conducting surveys on both nationwide and local levels. Some of these tend to be conservative in their methods, others more generous. Margins of error vary widely, as do polling samples. Whereas some surveys are a snapshot of a single day, others last up to a week, meaning respondents at the end will have been influenced by events that those earlier won't have. Even at the peak of Obama's surge last week, when some polls were trumpeting Obama leads of well over ten points, others, such as the LA Times survey, were pointing to far more modest advantages of seven to eight points.
As a result, polls of polls - such as the widely renowned Real Clear Politics average - have not seen the substantial change in the past days reflected in some individual surveys. After the release of the somewhat anomalous NYT poll last week, the RCP average sat at Obama +7. Yesterday and today, it is Obama +5.5. A drop, yes, but not a statistically significant one. What has narrowed is the spread of polling outcomes, rather than Obama's overall lead.
Polls of polls are useful to get a general picture of the race, but also have some drawbacks - primarily that the polling window from which they take their average tends to be rather large. This is particularly true for state polls, where surveys are not necessarily conducted so regularly and an average might be taken over a month or more - with the obvious consequence that some of the results are a little outdated. You can also argue that nationwide polls matter little, the race being decided not by the popular vote but in the battlegrounds of Ohio, Florida and North Carolina.
For real accuracy - or the closest approximation that one can get - it is best to compare like with like. Look at the most recent New York Times or Zogby poll, and compare it with the one before, and the one before that. This way, you will get a good sense of the overall trend if not the precise numbers involved. You can do this with polls of polls too, but check that they are always using the same surveys to be sure of a reliable result.
Lastly, remember that polls are only a yardstick, and do not always reflect the eventual outcome. This was demonstrated starkly during the primary race, when the results of more than one Obama-Clinton contest confounded all polling predictions, to the embarrassment of pundits across the nation. Extrapolating nationwide voting intentions from a sample of 1000 or so voters is not a foolproof method, and will vary depending on how the sample is defined. Gallup's recent tracking data is a good example - yesterday, among likely voters, Obama led McCain by just three points - a figure that incidentally has barely fluctuated over the past week. Among registered voters - he led by ten points, 52 to 42 - the highest support he has polled in the survey so far.
The message? Liberal ingestions of salt with each poll reading. It's an idea, but, particularly in a contest with an unknown racial component that will only become fully evident on November 5, no more than that.
Two maps, in fact. The first uses the latest poll in each state to determine whether that state is a battleground (if it is tied, or one candidate is leading by six points or less) or if it is a solid state for the Republicans or Democrats (if one candidate has a lead of seven points or more). In this scenario, Barack Obama has a clear lead though there is a great deal still to play for.
In the second map, there are no battleground states. It is assumed that if a candidate has a lead of any size, the state will fall to that candidate. If the state is tied in the latest poll, I have taken into account the previous poll in order to break the tie. This map shows just how agonisingly tight the race really is - the two candidates are tied, both one point away from the magic number of 270.
I have listed the results of the latest polls for all states below the jump. A couple of notes: the result given is from the latest poll, or if multiple polls are submitted on the same day, an average is taken. Most of the polls were concluded in the past week, though in some cases the most recent data available is from the previous week. Results are collated from Real Clear Politics and USA Election Polls.
Hat tip to the Washington Post for the map - click on the link to the side of the map to create your own.
Map one: Electoral college including battleground states
Map two: Electoral college, no battleground states
Click below to keep reading for a breakdown of polling data
Continue reading "State of the race: an electoral college map" »
 Barack Obama has all but eroded John McCain's poll lead as the Palin effect begins to wear off, according to the latest polls.
The Republican nominee had established a four to five point lead over his rival following the St Paul convention - and in particular Sarah Palin's introduction to the country as his vice-presidential pick.
But Obama has now pegged him back to just a one point edge, according to an average of the latest polling data by Real Clear Politics. The pair are now virtually tied at 46.1 to 45.1 per cent.
Meanwhile some surveys, including the Hotline/FD Tracking poll released today, show the Democratic nominee in the lead by 2 to 3 points. A round-up of four daily tracking polls by MyDD gives Obama a 1.5 point lead, but with margins of error typically around 3 points the overall impression is of a dead heat.
The data may reflect Obama's success in painting his rival as out of touch with economic concerns aas he contrasts McCain's previously upbeat assessment of the economy with the current turmoil on Wall Street. The Hotline/FD poll shows McCain's approval ratings on the economy sinking from 43 to 36 per cent in just four days.
The Obama camp has also been fighting back hard against a series of negative ads by McCain, accusing him of running a dishonest campaign based on smears and distortions.
But perhaps the most significant factor in Obama's resurgence may be that the conservative euphoria surrounding the young governor of Alaska is beginning to wane. This view is borne out in poll data cited by Newsweek, which shows Palin's favourability ratings dropping by 10 points in just three days as a series of unhelpful stories and a questionable performance in her interview with ABC's Charlie Gibson reinforce doubts over her experience and past political judgments.
 The latest polls show Barack Obama leading John McCain by 50 to 42 per cent nationally, but as always it will be in the battleground states that the race is won and lost.
Here's a battleground breakdown - using an average of nationwide and local surveys from Real Clear Politics. I've omitted states where one candidate has a lead of more than 17 points :
Virginia: Tied
New Hampshire: Obama +0.3
Colorado: Obama +0.4
Nevada: McCain +1
Ohio: McCain +1.2
Florida: McCain +3.1
North Carolina: McCain +4.2
New Mexico: Obama +4.3
Michigan: Obama +4.3
Minnesota: Obama +4.5
Pennsylvania: Obama +5
Iowa: Obama +5.3
Montana: McCain +5.3
Georgia: McCain +6.3
Indiana: McCain +6.7
Missouri: McCain +7
Oregon: Obama +7
Wisconsin: Obama +7.2
Washington: Obama +10.5
New Jersey: Obama +10.7
Texas: McCain +10.8
Mississippi: McCain +11.7
Arizona: McCain +11.7
Massachussetts: Obama +12.6
Maine: Obama +15
California: Obama +15.3
Kansas: McCain +16
Kentucky: McCain +16
New York: Obama +16
Louisiana: McCain +16.3
Arkansas: McCain +16.3
Using this data, I've created an electoral college map over on the Real Clear Politics site, which you can view here. Any state in which a candidate is leading by four points or more I've classified as leaning towards that candidate (light blue or light red). Any state in which a candidate is leading by seven points or more is classified as solid (dark blue or dark red). Toss-up states (no candidate leads by four points ore more) are shown as grey.
These classifications differ from those used by Real Clear Politics, which labels any state with a candidate leaning by less than 5 points as a toss-up, by between 5 and 10 as leaning, and over 10 as solid. You can see the original map here.
In my scenario, Obama leads in the electoral college with 260 votes to McCain's 200, with 78 a toss-up.
In the RCP scenario, Obama leads 228 to 185, with 125 a toss-up.
A candidate needs at least 270 electoral college votes to win.
This week's polls showing Barack Obama with small, but significant, single-digit leads among likely voters are certainly welcomed by Democrats, but recent history hasn't been kind to early frontrunners.
In fact, only one of the last five June election-year polling averages has correctly predicted the popular vote winner in November - Bill Clinton in 1996. Even then, the polls missed his win-margin by more than 9 percent.
 As hard as it may be to believe, Michael Dukakis (left) was leading the first George Bush by an average of 8.2 percent in June of 1988. Bush went on to win the general election by 7.8 points.
Mr Bush led the relatively unknown Bill Clinton by 4.9 percent In June of '92, but managed to lose in November by 5.6 percent.
June 1996 polls showed the incumbent President Clinton leading by a whopping 17 points, but even Bob Dole managed to close the gap to a more respectable 8.5 percent.
2000 was different only in that George W. Bush led by 4.7 percent in June, won the election, but lost the popular vote to Al Gore by 0.5 percent.
And finally, John Kerry led in the June 2004 polls by an average of 0.9 percent, but lost the popular vote, and the election, to the incumbent Bush by 2.4 points.
So, while Mr Obama's leads are certainly signs for Democrats to be optimistic, history points to caution. A lot can happen between June and November.
An air of dejection hangs over the polling department at Camp McCain. Pinched faces sit in nervous silence, each new set of numbers triggering a volley of angry barks before the gloom settles in again.
I am not privileged with access to said office, of course, and am therefore making this up. But that is what I imagine the current atmosphere must be like, if they put any stock at all in the proliferation of negative polls currently battering their candidate.
Rick Davis, John McCain's campaign manager, insists the Republican nominee is in a strong position, claiming that he is on course for victory in the battleground states of Wisconsin, Michigan, Missouri and Nevada. In an online strategy briefing, he asserts that Ohio is "a dead heat" and that Pennsylvania too could go Republican.
 But polling analysis, such as that conducted by the Real Clear Politics team over at Time magazine, suggests otherwise. It shows Barack Obama leading in the five most recent Pennsylvania polls by an average of 5.8 points, while in Wisconsin he commands a two-point edge.
RCP's thorough (and consistently accurate) breakdown of recent polling makes heartening reading for the Democratic nominee. Averaging out the results of surveys carried out since mid-May, it shows Obama beating McCain by 3.1 points in the November head-to-head. While such a margin won't see the Obama camp breaking out Havana's finest, it nevertheless represents a climb since his victory over Hillary Clinton became apparent last week and will reassure advisors that the campaign is moving in the right direction. Perhaps even better news for the Illinois senator is that all the surveys analysed project victory over McCain - bar Newsweek, which finds the two candidates in a dead heat. (RCP also has some useful electoral college maps here.)
But the path to the White House runs, as always, through the battleground states, which this year include the electoral collage heavyweights of Florida, California, Pennsylvania and Ohio. All of these favoured Clinton in the primaries, leading her to argue that her rival could not carry these states against McCain in the autumn. So far, at least, such predictions are proving incorrect, with RCP averages giving Obama a sliver of a lead - 1.3 points - in Ohio, a healthier lead in Pennsylvania, as already discussed, and a commanding 12.6 point edge in California. However he has yet to win over Floridians, who currently back McCain by an average margin of 8.3 points. Could his performance improve as memories of the vitriolic dispute over its rule-breaking primary fade? Perhaps. But the Sunshine state, with its large communities of Cuban exiles and wealthy retirees, has always been reluctant to vote Democrat - having done so just three times since 1952 - and ultimately Obama may do better to cut his losses here and focus on making the votes up elsewhere.
This, with his current push into Republican strongholds, seems to be exactly what Obama is intent on doing. Today, he embarks on a two-week tour North Carolina - which lasted voted for a Democratic presidential nominee 32 years ago - after which he is to head to Missouri, which has voted Republican in seven out of the last ten elections. He chose St Paul, Minnesota - the venue of the Republican convention in November - to declare his nomination victory last week before taking the battle straight to the GOP with campaign stops in Virginia, which last voted Democratic in 1964. (Political nerds can find an excellent breakdown of electoral history at www.270towin.com).
While this is a risky strategy, it could ultimately pay off. Having adopted such a tactic during the primaries, Obama has established robust voter turnout operations in many of these states and could yet pose a stiff challenge to Republican primacy. At the very least, as Obama aides quoted by the New York Times point out, it could force McCain to spend time and money campaigning in what should be safe ground, distracting him from fights elsewhere.
All this appears to be rattling the Republicans, with campaign insiders and party strategists using the press to urge a shift in strategy. Speaking to the Huffington Post, one Republican operative spoke of "a world of problems" facing McCain, suggesting the only was to finance a volley of negative ads painting Obama as a liability in an age of terrorism - a tactic the Arizona senator has so far disavowed.
Last week, Bay Buchanan - Reagan's treasurer and a leading conservative - delivered a scathing assessment of McCain's prospects, noting that his moderate policies alienated much of the Republican right while his opposition to trade protections would frighten off many of the so-called Reagan Democrats whose votes may well swing the election.
Writing in Human Events, she said: "In reality there is only one candidate. Barack Obama. In November he will win or he will lose. John McCain is relevant only in so far as he is not Barack Obama. The Senator from Arizona is incapable of energizing his party, brings no new people to the polls, and has a personality that is best kept under wraps."
 With Super Tuesday promising to whittle down both races (and perhaps, on the Republican side, produce a winner) thoughts are turning to possible pairings for the presidential head-to-head. Unsurprisingly, it is a subject fascinating the pollsters, who have probed every possible permutation to determine the likely outcome. We all know that polls can be wrong (especially after the New Hampshire debacle) but nevertheless this breakdown of national surveys on RealClearPolitics makes interesting reading.
It confirms what Democrats have feared all along- that a McCain candidacy would be tough to beat. Considered a liberal Republican (despite his plan to keep the US in Iraq for 100 years,) McCain has significant crossover appeal and could easily leach votes from the Democratic candidate, particularly if that happens to be Hillary Clinton, a divisive figure even among her own party.
When all the national poll results are averaged, McCain beats Clinton by a slim margin. But he is the only Republican who does - both Mitt Romney and Mike Huckabee lose by double digits. Barack Obama, on the other hand, not only routs Romney and Huckabee but also comes out on top against McCain (by a rather paltry one point, admittedly, but then, as George W Bush well knows, he could make do with even less.)
This is a message that has not been lost on Republicans, who have responded to Obama's surge by rallying behind McCain as the most electable candidate. It is a bitter pill for right-wingers, who dislike McCain almost as much as Mrs Clinton.
But, as Andrew Sullivan argues in this Times article, the prospect of another Clinton in the Oval Office might be the one thing to make them swallow.
He writes: "She (Hillary) has extraordinary negatives. She galvanises the conservative movement in ways no other Democrat can. Against McCain, she and she alone enables the Republicans to forget their deep internal divisions and unite. Nothing – nothing – unites them as she does."
Sullivan closes his argument with this salient question: "Will Democratic voters realise that he is now their best bet against McCain or will inertia and fear keep Clinton alive? One thing I’ve learnt in American politics: never underestimate the capacity of the Democratic party to screw it up."
 Moments after John Edwards quit the Democratic race today the blogosphere was buzzing with frenzied analyses of the likely impact on the remaining  contenders. By and large the commentary is as inconclusive as polling on the subject, with some suggesting Obama might gain a slight advantage, others leaning towards Mrs Clinton and more still sitting resolutely (and probably wisely) on the fence.
There are two key points to consider here: delegates already won and supporters who have not yet voted.
Edwards has not yet thrown his weight behind a candidate and may well not do so until he has assessed his potential influence. If the race stays as close as it is as present he may decide to hold on to the dozens of delegates he has accrued and direct them nearer the convention to the candidate who offers him the best reward (ie the VP spot on the ticket or a future Cabinet post.) On the other hand he is clearly ideologically closer to Obama and may endorse him sooner if it looks like Hillary is pulling ahead. I've written about this in an earlier blog post.
However with his delegates not bound to obey him, the bigger question is where his supporters in the remaining states will turn.
Here's a selection of the best comment on the web:
Continue reading "Edwards quits - but where will his supporters go?" »
Latest News
Confused?
Contact Us
categories
RSS Feeds
Recommended
Times Online
More of our blogs
Multimedia
|  |
|