Tuesday, September 30, 2008

The Verdict is Coming In

The public's verdict on the first presidential debate is beginning to come in and . . . . . . nothing much really happened. The needle has hardly moved. The temperature is still more or less the same. Obama's lead over McCain has inched up just a notch:

Obama's electoral vote lead remains at 317-221. Very little movement in the swing states was recorded over the past 48 hours:

Take a nap and check in again in a day or so and we'll see if anything changes . . .

Monday, September 29, 2008

Too Much Red Meat

John McCain spent too much time throwing red meat to the red state folks at the first presidential debate the other night and too little time trying to impress the undecided independent voters that will decide the election. Unfortunately for him, Barach Obama's lead heading into the debate was growing as we can see from the chart below:

Obama's share of the vote appears to have increased by 0.6 percentage points between Thursday and Friday last week and now stands at 2.2 percentage points. His lead in the Electoral College remains at 317 to 221. However, his leads in the close states have grown and McCain's margins in Florida and Nevada continue to thin:

The polls coming in during the upcoming 24 hours will give the first concrete impression of the first debate's effect. Obama should have something to smile about.

Saturday, September 27, 2008

Momentum Obama

The first presidential debate between John McCain and Barach Obama came and went earlier today (evening in the US) but it will be several days before the impact of the debate, if any, will be clear. On a purely stylistic level the debate was apparently "won" by Obama. As far as I was concerned however, neither candidate appeared interested in seriously engaging in a real discussion on the most pressing topic of the day - the $700 billion bailout of the US financial system. Neither candidate was willing to state their real feelings about the proposed bailout and instead equivocated and obfuscated their way to completely unsatisfactory responses. As Ross Douthat noted in the on-line Atlantic, in the end this election may be about nothing at all; if this is the case, it doesn't matter whether Obama or McCain win at least as far as policy is concerned. The reality may be that the combination of crumbling economy and massive deficits may limit the ability of either candidate to undertake any of their agenda once they are elected.

At any rate, back to the fun part, namely handicapping the state of the presidential race. Polling over the past 48 hours suggests that Obama has opened up a clear lead over McCain in the national popular vote:
This still somewhat narrow popular vote lead translates into a significant vote margin in the Electoral College:
Missouri, Ohio and Virginia, with a combined total of 44 electoral votes, appear to have flipped from McCain to Obama over the past several days. In addition, Florida has now moved into the swing state category:

The electoral map makes clear the spread of blue over the US in recent days:

If my perception is correct that Barack Obama "won" the first debate, the above map could be becoming increasingly blue over the next several days.

Thursday, September 25, 2008

Advantage Obama

Barach Obama has, according to my model, taken a narrow lead both in the popular vote and in the Electoral College vote based on a number of polls released over the past 36 hours. As has been the case in recent days, the state-level polls are suggesting a closer race than the national-level polls. The national polls released over the past 10 days suggest that Obama has a 3.4 percentage vote lead. However, the state level polls suggest, if anything, that John McCain has a slight lead. While the 14-day decay period for the state-level polls (compared with a 10-day decay period for the national-level polls) may explain part of the reason for this disparity, it is also clear that McCain is maintaining leads in states like Ohio that seem out of kilter with the national percentages. My model averages out these differences with a bit more weight to the state-level polls than the national polls.

Barack Obama has a 1.0 percentage point lead in the national popular vote:

This lead translates into a narrow lead in the Electoral College:

However, as the table showing those close states with a spread of less than four precentage points below makes clear, the race is electron-microscope close in Ohio and Colorado:

Below is the current electoral map showing the currently safe states, (shown as dark red for the GOP and dark blue for the Democrats) and the close states (shown as pink for the GOP and light blue for the Democrats):

Wednesday, September 24, 2008

Too Close to Call

I've added a time decay factor to polls in the model so that older polls have less weight than more recent polls. National polls are assumed to decay in value over a 10-day period while state level polls are assumed to decay in value over a 14-day period. This change gives the daily tracking polls much more weight than before, as well they should. I also found a pile of mostly state-level polls done by the American Research Group that have not been included in the list of polls put on the RealClear Politics website and included these polls in my model. It appears there is some criticism of the accuracy of polling done by ARG, but a review of the polling results done by this organization suggest that their results seem in line with polling results from other organizations. All of this extra effort produced more or less the same results as if I had not made the effort. However, as every statistician knows, such efforts lower the uncertainty in the model and so I feel much, much better about the model, even if nothing else seems to have changed. It's like taking vitamin pills - they make you feel better if they actually don't make you any healthier.

The popular vote share remains the same as several days earlier:
While the movement of New Hampshire into McCain's column gives the Republican candidate an 18-vote margin in the Electoral College:
The reasons for the discrepency between the national popular vote and the electoral vote are extremely close races in five states (Colorado, Missouri, New Hampshire, Ohio and Virginia) all of which are in the McCain camp according to my model:

In reality, it is more reasonable to view these five states as toss-ups and the election as well.

Sunday, September 21, 2008

McCain's Still Ahead

Well, sort of. I guess. Maybe. But I'm not sure. The national tracking polls have definitely been trending towards Barack Obama over the past five days but this national trend isn't necessarily in evidence at the state level. A good example of this is in Ohio where a plethora of polls suggest that there has been very little movement one way or other over the past 10 days. Another example is in Missouri where three polls taken by three different organizations over the last eight days show are within one percentage point of each other. This discontinuity between the national and state polls is also noted by Nate Silver on the 538.com website. Here is my take on the national popular vote shares between the two main candidates:

My model shows Obama ahead by 0.6 percentage points in the popular vote. This compares with an Obama lead of 2.2 percentage point spread in the 538.com model. Although both of our models showed McCain with a lead of about 2 percentage points just a few days ago, the 538.com model has been more sensitive to the polling movement at the national level than has my model and has been more quick to swing to the other direction. It's possible that the 538.com model puts more weight on the national polls than does my model, the latter which weights the national polls at about one third and the state polls at about two thirds. While Obama is ahead in the popular vote in my model, McCain still hold the same 10 electoral vote lead he has held for the past several days:

The reason for McCain's continued lead is an increasingly slender edge in Colorado, Missouri, Ohio and Virginia:

The four closest states have a total of 53 electoral votes, enough to give Obama 317 votes and a comfortable lead if all of these states flip. At this stage in the campaign, it looks like these four states may well hold the key to the election.

Saturday, September 20, 2008

Cooling Trend Continues

Although it has been little noted in the media the earth has been going through a cooling trend over the past nine months. During this period average global temperatures have declined by about .10 degrees Celsius (based on a 12-month moving average of the global temperature anomaly). Given the range of natural temperature fluctuations possible in the short term, most scientists would probably just dismiss this recent temperature movement as a random fluctuation.

While that view is very reasonable, I do consider that this is indicative of a long-term trend that will, in my view, produce a period of cooling for the next 30 years. As such, it is reasonable to keep track of this phenomenon as it unfolds so that more people can become aware that this is possible. The chart below shows the 12-month moving average global temperature anomalies for the actual data (from NOAA), for my own global temperature forecast based on the assumption of a cooling trend through 2038 counteracted by anthropogenic (human) warming caused primarily by greenhouse gas emissions and for Tamino's forecast as noted in other posts in this blog. Note that his forecast has been lagged one year so that it is directly comparable with the anomaly and anomaly forecast curves. Note also that my assumption of anthropogenic warming assumes that the climate sensitivity (or feedback effect) is zero and thus the position of the actual temperature anomaly above or below my forecast line potentially provides an indication of the actual climate sensitivity. The figure below shows the two forecasts and the actual temperature curve. The beginning date in the chart represents the approximate beginning of the current climate cycle:

Note how my forecast and Tamino's are, thus far, statistically indistinguishable. Tamino's forecast represents the IPCC consensus for an approximately 0.2 degrees increase in average global temperatures each decade. There are two things of note here - first, the actual global temperatures are now well below the consensus forecast as represented by Tamino's forecast and; second, the shape of my forecast and that of the actual temperature change are very similar but separated by about 12 months. Now, either or both observations could just be random events but it does give us something to focus on going forward.

Lots of Polls . . . .

But no movement. A total of nine state polls and four national tracking polls produced no meaningful net movement in my model meaning that the two candidates remain virtually tied in the popular vote and separated by 10 votes in the Electoral College. Below is the current electoral map according to my model:

To give a bit more flavor of the race's dynamics I have colored those states where the leading candidate has no more than 52% share of the two party vote has been colored pink and light blue for McCain and Obama respectively:

The two closest states are Ohio and Missouri, either of which would tip the election in Obama's favor were they to switch to Obama.

Friday, September 19, 2008

By the Hair of His Chinny, Chin, Chin

Is all that separates John McCain from Barach Obama. An avalanche of 27 state and 6 national polls released yesterday suggest that the race is a virtual dead heat in the popular vote and that McCain is hanging on by a thread in the Electoral College:

The table below shows the state of the race in the eight states where the candidates are separated by less than four percentage points:

I've made a few adjustments to the model over the past week or so. First, I've given the state polls equal weight with the national and correlated state polls in determining the expected distribution of the votes in each state. That is, the state polls, national polls and the correlated state polls each contribute 33.33% to the estimated value in each state. Second, where there is a state poll, I use the results from only one iteration of the model in order to derive the estimated shares in that state. If there are no state polls, I use the results from three iterations of the model in order to derive the estimated percentage shares in that state (if you have further questions, please leave a comment and I will be happy to respond).

Thursday, September 18, 2008

More Uncertainty

The lastest polling confirms the trend in Barack Obama's direction I discussed the other day. The movement is not dramatic, but dramatic is probably not what the doctor would order at this stage in the race anyway:

More importantly, the shift has removed Michigan, Minnesota, Oregon and Pennsylvania from Obama's list of endangered states and has put Virginia back on John McCain's side of the ledger. As you can see below, the pressure is now on McCain to hold on to a number of states, most notably Missouri, New Mexico and New Hampshire.

These three states have a total of 20 electoral votes, enough to put Obama back in the lead. For now, however, McCain remains ahead in the Electoral College by a 288 to 255 vote margin.

Wednesday, September 17, 2008

Watching Grass Grow

As I predicted yesterday, the momentum towards Barack Obama has continued over the past 24 hours. At a grass-growing pace. The three national tracking polls did show a shift towards Obama of about one percentage point. However, this movement was blunted by a George Washington University sponsored poll that showed John McCain with a four percentage point lead over Obama nationally. Further lack of evidence of substantial movement towards Obama came in the shape of a Public Policy Polling poll (try saying that quickly three times) in Ohio that showed McCain with a four percentage point lead in the Buckeye State. On the other hand, two polls from New Jersey more or less confirmed my model's view that Obama's lead there is in the six to eight percentage point range. The result of this mix is a very, very slight movement towards Obama, but not enough to quicken the pulse.

This movement was however enough to take Minnesota off of the list of Endangered States for Barack Obama as his share of the two-party vote there moved one tick up to 52.1%. Otherwise, the swing state list showed little change:

My model shows no change in either the national vote share or in the Electoral College vote totals. One gets the feeling that this race is settling in and we may see little significant movement until after the first debate on September 26th in Oxford, Mississippi.

Tuesday, September 16, 2008

Delaware Speaks!

More precisely, we finally have a poll from the First State (so named because it was the first state to ratify the new Constitution in 1787). The Rasmussen poll of 500 likely voters gives Obama a 55%-43% lead in Delaware、within a point of the lead my model has been showing. I've been adding Delaware to the model (it takes time because I have to go through every state in which Delaware meets the correlation threshold for that state, which is most of them, and add the Delaware data for that state). In the process, I have observed that this new information changes the vote totals in each state where Delaware meets the threshold by at most several thousand votes. If you conclude that this is because Delaware is unimportant, you are probably wrong. In spite of its small size (about 2,000 square miles) and tiny population (about 850,000), Delaware is an important bellwether in US presidential elections.

This is because Delaware has an incredible amount of cultural and economic diversity in spite of its small size. The economy of the north is based on industry (Du Pont is headquartered there) and services (most of America's corporations are legally based there). Philadelphia is but minutes away by train or car and, not surprisingly, the accents and attitudes are northern. But travel to the south of the state and things quickly change. Roughly around the middle of the state the accents change and one enters the world of the southern US, and along with the change in accents is a change in attitude. The landscape is rural and pastoral; chicken farming dominates the local economy in the south of the state. The north and its attitudes seem far, far away.

The result is a state that is a microcosm of mid-size urban and of small town/rural and of that great cleavege of America, the North and the South. For decades the state voted within a percentage point or two of the vote shares at the national level. While Delaware has been pulled more into the orbit of the northeast in recent elections and thus has become more Democratic, the state remains a consistent indicator of the state of the race at the national level.

There was also a poll from New York that showed Obama with a five percentage point lead. While its hard to believe that Obama is only up by five points in New York State, it is our first post-convention poll from there and the results of this poll did have some impact on conditions in other states, most notably Pennsylvania, Michigan and Minnesota. There are now 10 states on the list of swing states (that is, a lead for either candidate of less than four percentage points).

The five states on the list in Obama's column are must-win states for Obama and he simply can't afford to have a single one of these states slip into the Republican column.

At the national level, the polling showed a very slight movement towards Obama which had little impact in our model but which should begin to make a dent in McCain's lead as it continues over the next several days. McCain's lead in the popular vote is 2.2 percentage points and his Electoral College lead remains at 283-255. I expect Obama's momentum to continue for at least the next several days as voters assess the impact of Wall Street's meltdown on their own well being. There is every indication that this is going to remain a close election for at least the next several weeks.

Monday, September 15, 2008

Tamino's Bet Month Eight

A reader kindly noted to me last month that Tamino made a slight correction to his "bet" , sometime in July. Tamino notes in his blog that "In this post I made a simple error in the statement of the bet: rather than give the equation of the lower dashed red line, I gave the equation for the trend line itself. The constant was given as 0.277455, when it should have been 0.085655. This has been corrected. The “critical value” is therefore 0.085655 + 0.018173(t-1991) , where t is the year, and for 2008 this turns out to be 0.3946."

I had noticed at the time, and noted in a response to his post, that the original equation he posted appeared to be incorrect. It appeared that instead of the equation for the upper limit for the "cooling" position Tamino had given the equation for the mid-point trend line. As a result, I calculated the lower point of the trend line and used the year mid-point in order to calculate the average value for a given year during the bet (assuming that, say, 2008 actually means the beginning of 2008 rather than the average for the year). As a result, I derived a rounded figure of .4035 degrees and have used this figure in order to monitor the status of Tamino's "bet". Because I used the year's mid-point, my figure differs slightly from Tamino's .3946 degree figure. Since it is Tamino's bet, it seems reasonable to use his definition.

With that in mind, we move to the August figures from Nasa's GISS. GISS reported a global temperature anomaly of 0.39 degrees in August. It was the coolest August of the decade; it was the fourth time this year that the global anomaly in that month represented the coolest month of the decade. For the year through August the average global temperature anomaly is now .3725 degrees or .0221 degrees below Tamino's threshold. The minimum average temperature would need to be .4388 degrees for the remainder of the year in order to meet Tamino's threshold for neither side to receive a point; below that average figure the "cooling" side would win one point.

While the issue as painted by Tamino is whether or not the earth is continuing to warm, my own opinion is that the issue is much more murky than this. I think that both are likely the case, that is, the earth is warming through human activities and the earth is presently cooling for purely natural reasons. It would be interesting to see the "cooling" side get a point this year but I suspect that the world's media would little note such an event. Why do so when one can focus on minutia like polar bears rather than the big picture?

Aye, Iowa

Yesterday I noted that there had been no post-convention poll in Iowa and that, in its absence, my model was showing McCain ahead in the state. Well, the pollsters heard of my dilemna and rustled up real quick an Iowa poll to fill that gaping gap. According to the Des Moines Register's poll, Obama is ahead in Iowa by a 52% to 40% margin. As this margin is similar to that shown in at least one pre-convention poll in the state, it is highly likely that Obama is leading in Iowa and quite likely that his margin is fairly substantial. To compensate for what appears to be reality, I have tacked five percentage points on Obama's share for Iowa (and correspondingly have taken five percentage points off of McCain's share).

The map below shows how unusual Obama's performance in Iowa is:

The figures in each state are Obama's estimated vote percentage shares. Obama appears to be struggling to stay on top in Minnesota and Wisconsin two upper mid-western states once reliably Democratic but which have now become swing states. Note the overwhelming Republican support in the tier of agricultural states running north to south from North Dakota to Kansas. Obama's performance in Iowa is just a shade behind that of his home state of Illinois, a state dominated by the Chicago metropolitan region. In contrast, Iowa's economy revolves around agriculture and Des Moines is the only significant city of any size in the state.

So why is Obama's support in Iowa so strong? Corn might well be the reason. Corn is the state's most important crop and virtually the entire state is located within America's prime corn growing area. There's this little thing called ethanol, the production of which is heavily subsidized by the US Government and which uses corn as its primary input. Iowa has benefited handsomely from this generous US Government program. So has Illinois, which is also a major corn producer. Now it just so happens that one Barack Obama, Senator from Illinois, is a strong supporter of corn ethanol. Looks like the good folks of Iowa just may be showing their gratitude for good old fashioned pork.

I mentioned yesterday that I would discuss a bit more the fivethirtyeight.com election forecasting model. My model has been tracking quite closely the performance of their model both at the national and state levels. It appears we are using somewhat the same approach, that is, we are using results from the national polls, other states polls and the in-state polls in order to estimate the current condition in a given state. When I compared the results of my model yesterday with the 538.com model on a state by state basis, I noticed a very curious pattern. In spite of the fact that the overall percentage shares at the national level were virtually the same, Obama's percentage shares in the 538.com model were higher than those in my model in 31 of 50 states. When I checked these 31 states where Obama's percentage shares were higher than in my model, my model was showing McCain ahead in 23 of these 31 states (McCain was ahead in 21 of 31 states according to the 538.com model). In constrast, for those 19 states in which Obama's percentage shares were lower than those in my model, Obama was ahead in 11 these states while McCain was leading in 8 in both of our models.

In other words, the 538.com model appears (compared with my model) much more likely to estimate higher percentage shares for Obama when McCain is ahead in a given state and is conversely more likely to estimate lower percentage shares for Obama when Obama is ahead in a given state. I've tested this relationship using a simple chi-squared test and found that these differences were significant at about the 2% level. That is, it is pretty unlikely that the observed relationship has occured by chance. Instead, either I am biasing the results in favor of McCain by tilting swing states in McCain's direction or 538.com is biasing the results in favor of Obama by tilting swing states in Obama's direction. Since I am without bias or prejudice, I can only point the finger (tongue-in-cheek) in their direction since the 538.com folks appear to be quite highly partisan in favor of Obama. Jokes aside, this does point out that the human factor plays an important role in these models. Studies suggest that the more mechanical the model (that is, the less reliant on human judgement) the better the model's performance.

Other than the Iowa poll, two Minnesota polls came out today suggesting that the race there is very close wnile a New Jersey poll reinforces the belief that Obama has a modest, but real, lead in that state. Oh, yes, there was another poll from Utah shows McCain far, far ahead. If someone hadn't gone through the trouble to take a poll in that state, how could we possibly know who was ahead? The updated tracking polls show a breath of movement towards Obama, now trailing McCain by 2.0 percentage points, and, after the above-noted Iowa adjustment, McCain leads in the Electoral College by 283 to 255 votes.

Sunday, September 14, 2008

Table's Turned

My computer screen unexpectedly died the other day, requiring a trip by the computer to the computer hospital for an operation to replace the screen. The operation was a success and the computer was returned to my office this morning. As my model was on this computer, I was forced to wait patiently for the computer to come home so that I could update the model.

While the computer was ill, the character of the presidential race naturally changed. The Republican post-convention bounce has turned out to be real and the swing of the pendulum has put McCain out in front by 2.2 percentage points:

More importantly, McCain's modest margin translates into a near-comfortable margin in the Electoral College:

The updated electoral map shows that the south and interior west are now a solid bloc for McCain:

My model is showing McCain in the lead in Iowa and New Hampshire. This is likely to be somewhat controversial since Obama is widely thought to be ahead in both states by a modest, but reasonable, margin. However, no post-convention poll has been taken in Iowa so we have no direct measurement of any kind yet from that state. A poll showing Obama with a six point lead in New Hampshire was released after the conventions, but that is just a single poll and the error in one poll could easily be greater than the six point margin. My model's estimations for these two states and for the other seven states where the margin is four percentage points or less is shown below:

McCain's Iowa lead is just a sliver and is hardly comforting in Colorado, Missouri, New Mexico and Ohio. Ominously for Obama, Oregon and Pennsylvania have crept into this list of swing states and Wisconsin is simply too close for comfort. If any of these states slip from Obama's grasp, the election is effectively over.

I've compared the results of my model with the model on the website fivethirtyeight.com. Although this model shows McCain with a 2.4 percentage point lead (0.2 percentage points greater than my model), their state-by-state figures show McCain with a slender 274-264 vote lead in the Electoral College.

I'll have a bit more to comment on their model and this map tomorrow.

Saturday, September 6, 2008

Here Come the Elephants

The tracking polls are clearly suggesting that the pendulum has now begun to swing back towards the Republicans. My own model is showing that the race is very tight with Obama clinging to a 0.6 percentage point lead (50.3%-49.7%). However, since most of the state level polls were taken before the conventions and since the state level polls were a bit weighted towards McCain by the time the conventions began, I suspect that Obama's actual lead is closer to 2 percentage points. My model is also showing that Obama's Electoral College vote lead is a slender 275-263. However, the model has McCain up by less than 1,000 votes in Ohio (out of more than 5.6 million) but I suspect that in reality Obama remains very slightly ahead in Ohio. This in turn means that the electoral vote is more likely remains at 295-243 in favor of Obama.

The next 48 hours are probably critical. Should the Republican bounce be limited, the election is likely to stay close until at least the debates, if not just before the election. However, should the bounce be significant, then McCain would then be in the driver's seat and Obama would be forced to take chances.

Friday, September 5, 2008

Waiting Patiently

For the returns, er, polls to come in. The size of the Republican post-convention bounce should become apparent over the next several days after which the shape of the election heading into the stretch run should become clearer. The movement towards McCain over the past several days seems fairly slight as my model suggests that Obama continues to hold on to a slender, but definitive, lead:

The Electoral College vote totals remain unchanged as before at 304-234 in favor of Obama. Most of the razor-tight swing states narrowed slightly over the past 24 hours:

Expect some significant changes in the above figures and the electoral map over the next five to seven days. Except for the final days of the campaign, the next period will likely be the most important of the campaign.

Thursday, September 4, 2008

The Perils of Palin

You have to admit, Sarah Palin's name lends itself naturally to interesting wordplay. She wowed the faithful in her acceptance speech at the Republican convention on Wednesday evening. If you want a gauge of the impression she made, go now further than this Mother Jones Seal of Approval. Mother Jones is certified genuine left-wing article and so you know that David Corn is trying to wake up the largely somnolent Democrats who have assumed that Palin's pick was a disaster for the elephants. Compare that story to this nonsense in the NYT from Timothy Egan. This is basically pablum for the "anointed". Egan is the quintessential Obama follower - an arrogant, upper-middle class elitist who represents that new, upcoming brand of Democrat whose aim is to make sure the world is run by and for the deserving, namely those with the combination of advanced degrees and the "right" values.

Anyway, there were a few state polls, one or two new national polls, plus the two tracking polls to deal with today. They suggested a slight narrowing of Obama's lead at the national level to two percentage points; the Electoral College figures remain unchanged at 304-234 in Obama's favor. However, I just wouldn't put very much into this shift today as a number of state level polls have fallen out of the model over the past several days. The first real impact of Palin's speech should make itself apparent over the next 48 hours and, with McCain's speech on Thursday evening, the amount of bounce for the Republicans will become clear by the weekend or early next week.

Palin Second Thoughts

Some evidence of a Palin anti-bounce emerged in the polling yesterday as Barack Obama's lead jumped to 2.6 percentage points, according to my model. I'm calling it an anti-Palin bounce because the Demo's post-convention bounced seems to have been completely cut off, meaning that the public are now responding to event surrounding the Republican convention including the news that Palin's 17-year old daughter decided, apparently on her own, to make grey-less haired Mom a grannie. Such qualifies as a distraction.

There was no change in the estimated electoral vote with Obama continuing to lead by a 304-243 margin. Although the electoral vote did not change, there was a significant movement towards Obama in all of the toss-up states currently on our list especially in Colorado where Obama opened up a modest lead:

Tuesday, September 2, 2008

Reasonably Steady As She Goes

Polling showed very little movement in the presidential race over the past 24 hours. Barack Obama gained 0.2 percentage points on John McCain according to my model, but this may be little more than noise in the data. Obama's leads by 50.7% to 49.3% in the popular vote and leads 304-234 in the Electoral College. The table below shows the race in the closest states:

Colorado remains on the knife edge with the two candidates separated by little more than vapor.

Monday, September 1, 2008

No More Bounce

The Democrats convention and post-convention bounce, such as it was, is over. Sarah Palin's selection as the Republican vice presidential candidate appears to have already created a minor bounce of its own, and this is before the Republican convention gets underway in earnest.

Perhaps I should say in gustav in deference to the Norwegian-inspired tropical hurricane now rampaging through Louisiana. The hurricane will keep the Republicans off of the TV screens for another 24 hours or so and in the end the whole show may be limited to the nominees getting annointed and giving their acceptance speeches. The good news for the Republicans is that the President and his Vice side-kick had to run off to the Gulf rather than be made to feel uncomfortable in St. Paul. This turn of events will eliminate the need by the Republicans to try to defend the Bush presidency in any way during the convention and it may be that George Bush will be asked to sit quietly in the White House over the next 60 days or so.

Oh, yes, getting back to the post-bounce. My model suggests that Barack Obama's lead shrank by 0.3 percentage points over the past 24 hours:

Obama's lead in the Electoral College has dropped back to 52 electoral votes:

The following table shows, according to my model, the state of the election in eight states where the margin held by either candidate is less than 4 percentage points:

Obama's lead appears to be based on extremely narrow margins in three states.