August 31st: State of the Race

Welcome to my 2nd state of the race report. I’m doing this report today instead of on the 1st merely because tomorrow is Saturday. My next report will likely come in two weeks now that we’re into convention season, on September 14th.

As before, I’ll evaluate the race using five different metrics:

  • Current state of the race with TCtC states
  • Current state of the race with leaners
  • Obama EV total in states where he has 50% in the average
  • Obama EV total in states where he has 47% in the average
  • Electoral Vote total assuming a 2:1 undecided swing to Romney

State of the Race: Electoral Count including Too Close to Call

Obama: 221 (-50)
Romney: 191 (NC)
TCtC: 126 (+50)

This number reflects the general narrowing of the race, especially in the midwest. Obama has lost Ohio, Michigan, Wisconsin, and Nevada to the TCtC category, accounting for the 50 electoral votes. Three of those four states are either Wisconsin, where Ryan is from, or are in the midwest region. Missouri has switched between “weak Romney” and “too close to call” a couple times but for now rests in “weak Romney” as it did a month ago.

Obama’s 221 electoral votes is the lowest “strong” and “weak” total he has had during the entire campaign. Romney’s 191 is more or less what he’s been in April. His entire electoral vote range thus far has been between 171 and 191.

State of the Race: Electoral Count including leaners

Obama: 332 (NC)
Romney: 206 (NC)

While some of the races in the states have gotten closer, none have actually seen a change in who leads in them, thus the electoral count including all leaners regardless of the size of the lead remains the same as it did a month ago.

State of the Race: Electoral Count assuming Obama needs 50% in the average

Romney: 358 (-11)
Obama: 180 (+11)

This metric is and will always be the most pessimistic for Obama. It essentially assumes that if Obama isn’t at 50% in a state, he will lose that state. Historical patterns show that this just isn’t very realistic, but it is a metric nonetheless. Despite some narrowing in some polls, Obama actually gains 11 electoral votes in states where he sits at 50% or better in the average.

State of the Race: Electoral Count assuming Obama needs 47% in the average and leads

Obama: 316 (+16)
Romney: 222 (-16)

This is, I think, a better although not perfect metric than the previous one. Here, Obama still leads, and actually has a slightly better margin than he did a month ago. This suggests a few things: the margin may be narrowing and there may be less undecideds, but some of those undecides are going into the Obama column. Noteably, one of the states not counted for Obama here is Minnesota, where his average support is 46.5%, though his lead over Romney is 6.6% in the state. Iowa is currently the only other state where Obama holds a lead but isn’t at 47% or better in the average.

State of the Race: Electoral Count assuming undecideds break 2-to-1 toward Romney

Obama: 284 (-26)
Romney: 254 (+26)

I still feel this scenario may be the most realistic, though we’ll see about that on election day. Out of the 5 metrics, this is the only one other than the first one where he performs worse than a month ago. He falls from 310 EVs to 284 EVs if one allocates 2/3 of undecided votes to Romney, though Obama would still win under this scenario.

Under this scenario, Obama wins Wisconsin, Ohio, Colorado, and Michigan (Michigan by only 0.2%) and loses Florida (by 0.2%), Virginia, Iowa, and North Carolina.

Of course, this assumes that every last person who doesn’t respond as supporting for Obama or Romney will 1) vote and 2) vote for one of those two candidates. Of course, some of these people may stay home, and some of those people may vote 3rd party. Perhaps most significant of these may be votes that Virgil Goode could possibly draw away from Romney in Virginia. In this scenario, Romney wins Virginia by only 0.6%. Goode could easily draw that away from Romney in Virginia.

Two-Week Average for August 14-27: Obama 47.1%, Romney 45.6%

Obama’s streak of leading Romney big in the two-week national average is now over, with his lead being reduced to 1.5%, the second closest margin thus far in the race. However, there are several causes for margin getting closer. But first, the average:

Obama: 47.1% (-1.7%)
Romney: 45.6% (+3%)

Margin: O+1.5% (-4.7%)

This average included 12 national polls:

  • Monmouth/SurveyUSA/Braun from August 15-19 (Obama 46, Romney 45)
  • LA Times/USC from August 13-19 (Obama 48, Romney 46)
  • NBC News/Wall Street Journal from August 16-20 (Obama 48, Romney 44)
  • Associated Press/GfK from August 16-20 (Obama 47, Romney 46)
  • Fox News from August 19-21 (Romney 45, Obama 44)
  • Resurgent Republic (R) from August 16-22 (Obama 46, Romney 45)
  • CNN/Opinion Research from August 22-23 (Obama 49, Romney 47)
  • ABC News/Washington Post from August 22-25 (Romney 47, Obama 46)
  • Washington Times/JZ Analytics from August 23-25 (Obama 46, Romney 46)*
  • CBS News from August 22-26 (Obama 46, Romney 45)
  • PPP/Daily Kos (D) from August 23-26 (Obama 50, Romney 44)
  • Democracy Corps (D) from August 23-27 (Obama 49, Romney 47)

*Note: The Washington time poll also includes Gary Johnson in their polling.
Continue reading

Two-Week Average for July 31 – August 13: Obama 48.8%, Romney 42.6%

There has been little change between the national poll average for mid- to late-July and the average for the first two weeks of August:

Obama: 48.8% (NC)
Romney: 42.6% (-0.2%)

Margin: O+6.2% (+0.2%)

This average included five national polls:

  • Reuters/Ipsos from August 2-6 (Obama 49, Romney 42)
  • Fox News from August 5-7 (Obama 49, Romney 40)
  • CNN/Opinion Research from August 7-8 (Obama 52, Romney 45)
  • Politico/GWU/Battleground from August 5-9 (Obama 48, Romney 47)
  • IBD/CSM/TIPP from August 3-10 (Obama 46, Romney 39).

Continue reading

Two-Week Average for July 17 – 30: Obama 48.8%, Romney 42.8%

I posted the two-week national average for July 17 – July 30th today. I meant to do it at the end of last week, but several things came together that delayed me. I’ll also try to do a post for these 2-week averages from now until the election. In any case, the average now stands at:

Obama: 48.8% (+1.9%)
Romney: 42.8% (-1.8%)

Margin: O+6% (+3.7%)

How it works is this: I take non-tracking national polls (usually grabbed from RCP) whose FINAL survey date is within the range. If, for some reason, there are not at least 3 polls, then I’ll throw in the Gallup and Rasmussen tracking poll for the final date in the range to fill it in, but otherwise I won’t include them.

Up until now, the average has been pretty constant. With the exception of May 8 – 21, Obama has ranged from 46.9% to 47.4% – a whole 1/2 of one percent. Romney, meanwhile, has ranged from 43.6% to 45% – a mere 1.4%. So in short: the race has been pretty stable at about 47-44 since I started doing the 2-week averages in April.

The span for July 17 – 30 shows a different story. This span had four polls, which is low: Fox News showing a 45-41 Obama lead, NBC News/Wall Street Journal showing a 49-43 Obama lead, Democracy Corps showing a 50-46 Obama lead, and the somewhat infamous Pew poll showing a 51-41 Obama lead.

These four polls give an average of 48.8% for Obama and 42.8% for Romney. This is both a 2-week average high for Obama, a 2-week average low for Romney and, not surprisingly, the largest margin between Obama and Romney in a 2-week average, at 6% (the previous largest margin was 3.7%).

Now to address two points I’m sure will come up: First, I’m sure there will be some objections to the Pew poll, as it is pretty clearly an outlier, and given there are only four polls, it’s going to have a rather major impact. However, even without the Pew poll, this is the best 2-week span for Obama so far.

If one takes the Pew poll out, Obama still leads 48% (his best number so far) to 43.3% (Romney’s worst number so far) with a margin of 4.7%. So the Pew poll padded the margin by 1.3%, but it’s still Obama’s best 2-week span and Romney’s worst 2-week span.

Now to address the other point: it is way too early to celebrate if one is an Obama fan. This could be a short bump in Obama’s numbers for whatever reason. In the span of May 8 – 21, Obama’s average dropped to 45.1% and had a lead over Romney of less than 1%. But by the next 2-week span, Obama’s number went back to his typical 47%. If we see a margin of 4% or more in the next 2 week span, then Romney’s people have real cause for concern.

Also, people have led late and lost before. Gore led by 3% over Bush as late as mid- to late-September in 2000 and lost, although he did win the popular vote. Kerry led over Bush as late as mid- to late-August in 2004 and ultimately lost. While McCain never led Obama in a 2-week average in 2008, they were tied after the conventions in early September.

Also, it’s always too early to start celebrating if the conventions haven’t occurred yet. Bush led Gore by 5-7% through most of the summer until the conventions, when the race tightened up significantly. Meanwhile, Kerry led Bush through July and the first half of August until the conventions in 2004, when Bush took the lead for good.

But to Obama’s credit, he does well in two credits in which I am most interested: How does he compared against Obama ’08, and how does he compare against Bush ’04.

In the first metric, out of the six data points which can be compared, Obama ’08 has had clearly better poll numbers than Obama ’12 only once, and that was the anomalous week where Obama came in at 45.1% in May. Obama ’12 has beaten Obama ’08 in polling in four of the six two-week spans up until this point, and has for the most part seen the same level of support in ’12 and he did in ’08 to this point in the campaign.

In the second metric, Obama ’12 has outperformed Bush ’04 in all eight two-week spans so far, with Obama out-performing him anywhere from 0.5% to (before this week) 4%. Obama has outperformed Bush ’04 by an average of 2.6% so far over those eight periods.

And if you’re wondering if Romney and Kerry have mirrored each other – yes, they have, sort of. Kerry more or less hung around the same 44% mark Romney has up until the start of July. However, at that point, Kerry announced Edwards as his running mate, which started a 6-week span where Kerry polled around 46-47% instead. Romney hasn’t announced his VP yet and hasn’t had such a bounce, so he is currently trailing Kerry’s numbers at this point by about 3% or so.

The real race will begin when the parties have their major conventions, and the state of the race will become much more clear starting in mid-September.

August 1st: State of the Race

This is my first in a series on the state of the race. I’ll try to post these at the first of the month (or first weekday of the month) with a review of what the race looks like, and how the race has changed since the last state of the race. I’ll give the state of the race giving a variety of metrics including:

  • Current state of the race with TCtC states
  • Current state of the race with leaners
  • Obama EV total in states where he has 50% in the average
  • Obama EV total in states where he has 47% in the average
  • Electoral Vote total assuming a 2:1 undecided swing to Romney

I hope these will provide a variety of metrics to give an idea of where the race stands. Of course, the more metrics that Obama or Romney lead, the better the race is going for them.

Continue reading