Showing posts with label monthly change. Show all posts
Showing posts with label monthly change. Show all posts

Monday, August 25, 2008

The Pre-Convention Swoon Revisited

I don't necessarily want to reprise the post and the resulting discussion here, but I thought it was important to update Saturday's examination of the August poll movement to reflect all the data we have through August. I'm treating the Biden announcement as part of the convention. Therefore, any polls conducted prior to August 23 are included in the analysis below. In other words, all the polls covered in yesterday's electoral college projection post plus the Epic/MRA Michigan and Columbus Dispatch Ohio polls are rolled into this update. The Suffolk poll from Colorado is omitted because the survey period extended through yesterday. No, the Biden announcement was not part of the convention, but with its proximity to the Democractic convention, it will be difficult to parse out those differences in the polls, ex post facto. As such, we will treat the two -- in relation to state level polling -- as one in the same.
[Click Map to Enlarge]

The map above provides a much the same picture we saw just two days ago. In fact, none of the states where polling had been conducted prior to Saturday, saw any change to the intensity of their shifts thus far during August. However, both Utah and Wyoming were shaded in based on the large margins the Mason-Dixon polls in each gave John McCain. Neither really provides any substantial break from the conventional wisdom though. If anything the relative "tightness" of both to this point in the race was notable.

There are three other states that I kind of gave short shrift to the other day that I'd like to take the opportunity to address here in the absence of an in-depth analysis of the toss up states. Iowa, Minnesota and New Mexico are trending toward Johm McCain since the beginning of August. We've mentioned Minnesota's recent trends before, but they're worth noting again. The Land of 10,000 Lakes has seen a decided tightening in two of the three August polls. Both stand out as anomalous in the overall progression of polling in Minnesota. The slim two point lead in the Survey USA poll, as Scott mentioned, may have something to do with the way the firm is screening its sample. Indeed, the most recent Survey USA poll of Minnesota was in June during Obama's jump in the the polls in most states. While most polling firms showed an Obama lead in the mid- to upper teens, Survey USA had the race in a dead heat with Obama up by a scant one point margin. Minnesota and Survey USA may be like oil and water then. Still that only "explains away" one of those recent narrow leads for Obama in Minnesota. The other was the four point margin Rasmussen found in their latest sample of the North Star state. That one I won't dwell on, but I will say that when you take into account the high numbers of the "bounce" period -- and then peppered throughout the summer -- with the uncharacteristically low numbers we've seen recently, you get an average very similar to the one FHQ has overall now; one in the upper single digits.

In Iowa the story is a bit different. Other than a 17 point margin in a February Selzer poll, the range of polling has been between 2 and 10 points within no discernible pattern emerging. This one has settled in and remains fairly static within the area of a 6 to 7 point edge for Obama. New Mexico is essentially a fringe toss up state, propped up by two favorable Zogby Interactive polls. When those are removed New Mexico does move into the toss up region but it is on par with a state like Missouri on the flip side of the Electoral College Spectrum. It is possible that McCain moves in and picks off New Mexico, but at this point that is the equivalent to Obama pulling Missouri into his column. The probability of those two things happening simultaneously in November is quite small, but one candidate winning both would be a clear signal that the momentum down the stretch has swung decidedly in that candidate's direction.


Recent Posts:
The Electoral Map (8/24/08)

Swoon? What Swoon? A Look at the Changes During Pre-Convention August

Obama-Biden

Saturday, August 23, 2008

Swoon? What Swoon? A Look at the Changes During Pre-Convention August

Well, swoon may be a bit strong, but the perception is that the momentum is headed in McCain's direction. Is that the case, though? Sure it is. "They" told us so. However, just for fun, let's look at what has happened in August so far and see. The plus here is that this will give us a baseline in which to begin assessing VP selection and convention effects. Due to the compression of all four events (Democratic and Republican selections and conventions), it is difficult to definitively say, but we'll have these numbers as a jumping off point when that time comes, though.
[Click Map to Enlarge]

Thus far there has been polling conducted in 34 states during August, including 12 of FHQ's 14 toss up states. And that is where we'll keep the focus for this examination, on those toss up states. Coming into the month, McCain held advantages in 8 toss up states totaling 86 electoral votes and Obama was ahead in 6 states with 76 electoral votes. In that time only Ohio's 20 electoral votes shifted (from Obama to McCain), though -- spoiler alert! -- that will change tomorrow. Regardless, that was the state of play heading in to the heat of August.

Let's look at those states:
McCain -- AK, FL, IN, MO, MT, NC, ND, VA

Obama -- CO, MI, NV, NH, OH, PA

Immediately, we can take Montana and North Dakota out. Neither has been polled this month, but among the rest there was some movement of note. Among the six Obama states, Colorado, Nevada, New Hampshire and Ohio shifted away from the Illinois senator while Michigan and Pennsylvania moved toward him. As we've said over the last couple of weeks, Colorado, Nevada and Ohio are among the closest of states now. Having all three move away Obama is not a welcome sight for those within the campaign. While that is certainly negative, the movement among the McCain states is also noteworthy Of those six McCain states where surveys took place, only two moved in the Arizona senator's direction while the remaining four trended in Obama's direction. So while McCain is making strides in the closest of states, Obama is actually pulling some of those McCain toss ups further into play, going against the prevailing perception of the moment. Alaska made a strong move toward Obama and Florida, North Carolina and Virginia edged ever so slightly to the left.

Now, I should note that this data includes the new release of Zogby Interactive polling in a series of battleground states. I'll have more on this in tomorrow's electoral college post, but in this context I should make some comments about how those numbers affect our weighted averages in those ten states. Despite the sample being self-selected, most of these numbers jibe well with recent polling in those states.

Battleground States

Obama

McCain

Barr

Nader

Not Sure/Other

Colorado

44%

38%

8%

2%

8%

Florida

40%

43%

5%

1%

12%

Michigan

46%

37%

5%

1%

12%

Nevada

39%

38%

10%

3%

10%

New Hampshire

38%

42%

11%

1%

9%

New Mexico

46%

37%

5%

1%

11%

North Carolina

47%

39%

3%

2%

9%

Ohio

41%

36%

8%

1%

13%

Pennsylvania

46%

37%

5%

3%

8%

Virginia

43%

41%

5%

1%

10%


The exceptions are Colorado, New Hampshire, North Carolina and Ohio. And while the Ohio result shifts -- as you'll see tomorrow -- the Buckeye state back to Obama, it is still a very small change overall (and it only takes a little bit of one). Of the rest, only North Carolina changes the direction of the trend for August up to this point. And even then, the move toward McCain was minimal before the Zogby numbers were incorporated.

In the end, while it looks like an even distribution as far as the number of toss up states (and all states for that matter -- 19 - 14 in McCain's favor) heading in one candidate's direction or the other, McCain has to be given a slight edge because he has moved the three closest states (on the Electoral College Spectrum) in his direction. We should also note that Minnesota has crept toward McCain for the second consecutive month (or partial month in this case), yet remains on the upper end of Obama lean states. The situation is similar in Iowa as well. Though the Hawkeye state hasn't moved toward McCain in consecutive months, it has inched closer to McCain during August. It too remains solidly within the lean category for Obama, but if the current trend continues that could change.


Recent Posts:
Obama-Biden

More on the Effort to Curb 2012 Frontloading

The Links (8/21/08): National Party Convention Bounces

Monday, August 4, 2008

Did Obama or McCain Win July?

***I said I'd have this up yesterday, but Blogger wasn't for some strange reason allowing the uploading of images during the latter half of the day and I didn't want to post this without the maps. It gave me a chance to incorporate the last of the July polls that were released yesterday anyway. There will likely be some other July stragglers that emerge well into August, but I'll add them to a revised edition in the post for August. The June revisions are at the conclusion of this post. Thanks, and sorry for the delay.

June, it seems, was an Obama month. The Illinois senator both wrapped up the nomination and showed improvement in the polls, with few exceptions, across the map. The dynamic changed in July, however. Whereas McCain only had a handful of states trend in his direction in June, July was with a yellow (pro-McCain) tint. 24 of the 39 states in which polling was conducted in July pushed FHQ's weighted averages toward the Arizona senator.
[Click Map to Enlarge]
By contrast Obama's June dominance gave way to a July map where his trendlines decreased in 21 of the 39 states where he gained a month earlier. Certainly the inclusion of the Zogby numbers in June inflates Obama's end of June averages, and in many cases, the Zogby polls were the last polls conducted in those states and were thus given the most weight. Any subsequent move toward McCain in July could then really have an effect on the the changes to the average, making what may have been a small change to McCain a big change. Even when the Zogby numbers are completely removed from the equation, McCain still countered Obama's June increases by pulling closer or extending his lead in 20 states (as opposed to 24). The omission of the Zogby data translated into trendlines reversing in Obama's direction in six states (AR, MI, MO, NM, OR and VA) and toward McCain in two others (NH and NJ). In other words, in nearly a quarter of the states where Zogby conducted an interactive poll in June (8 of 34), those polls affected the direction in which FHQ's weighted average was going.

The Illinois senator did manage to increase his standing in 15 states (or 19 if the Zogby data is excluded -- a number much closer to the number of states in which McCain gained in the absence of those polls) overall in July. Let's take a moment to look at the states deemed toss ups as of Sunday in FHQ's weighted average. Of those 13 states, 12 had at least one poll conducted in July. Of those twelve, half moved in Obama's direction (FL, MT, NV, NC, ND and PA) and two shifted toward McCain (CO and OH). The remaining four (MI, MO, NH and VA) were among the group of states that shifted directions based on whether the Zogby data was used. Michigan, Missouri and Virginia trended toward McCain in July with the Zogby data included, but reversed course in the absence of that data, shifting in Obama's direction. New Hampshire had the exact opposite effect: favoring Obama with the Zogby data, but away from him in their absence.

Now, the Zogby data is actually included in FHQ's average, so we'll discuss these toss ups in that light. [I have to add the distinction, though, in the interest of transparency.] As such, these swing states, where these trends are of the most consequence, are evenly distributed between the two candidates (6 to 6). However, three of Obama's seven gains and four of McCain's are states that are currently favoring the other candidate in the overall average, albeit slightly. Sure that's the nature of a toss up state, but still, that's an interesting bit of information to take away from this. Obama leads in Colorado, Michigan and Ohio, but during July those three states shifted toward McCain. Likewise, McCain holds an advantage over Obama in Florida, Montana, North Carolina and North Dakota, but is witnessing the Illinois senator gaining on him in those four states. If those seven states completed the switch to the other candidate Obama would net a gain of 2 electoral votes in the electoral college, stretching his advantage to 300-238 (McCain states: FL+MT+NC+ND=48 EVs, Obama states: CO+MI+OH=46 EVs).

State Shift Rankings (July)*
MT-3
WA-11
MO-11
OR-7
AR-6
SD-3
NV-15
MI-17
AK-3
ID-4
FL-27
GA-15
VA-13
NM-5
KS-6
ND-3
NJ-15
CO-9
AZ-10
ME-4
IA-7
NC-15
MN-10
LA-9
SC-8
WI-10
NH-4
OH-20
NE-5
IL-21
PA-21
TX-34
AL-9
CA-55
OK-7
KY-8
CT-7
NY-31
MS-6

* The states are ranked from biggest shift toward Obama to the biggest shift toward McCain. The darker the shade (of yellow or green) the bigger the shift during July.


Well, that's all well and good, but there is one additional caveat we need to make, isn't there? What about those Rasmussen "leaners"? Did using the "with leaners" or "without leaners" data make any difference in how the July map looks above? Below are the July polls from Rasmussen since the "with leaners" distinction was added to the firm's polling press releases (up to and including the polls that came out just yesterday -- AL, AZ and CT):

Rasmussen Polls Since w/Leaners Distinction was Added (7/9/08)*
StateDatew/o Leaners
w/Leaners
Change
Undecideds Drop
Missouri
7/7+5
+50
-10
New Jersey
7/7+5
+3+2
-4
Illinois
7/8+13
+11+2
-6**
North Dakota
7/80
+1+1
-7**
Wisconsin
7/8+13
+10+3-6
Louisiana
7/9+20
+19+1
-2
South Dakota
7/9+4
+40
-4
Washington7/9+9
+8+1-6
Iowa
7/10+10
+100
-9
Michigan
7/10+8
+80
-5
Minnesota
7/10+18
+17+1
0
Kansas
7/14+20
+23+3
-9
North Carolina
7/15+3
+30
-5
Oregon
7/15+9
+90
-5
Nevada
7/16+2
+20
-5
Virginia
7/160
+1
+1
-6
Alaska7/17+5+5
0
-7
Arkansas7/17+10
+13+3-4
Georgia7/17+9
+11+2
+1
Maine
7/17+10
+8+2
-2
Colorado
7/21+7
+3+4-13
Ohio
7/21+6
+10+4
-7
Florida
7/22+1
+2+1-7
Minnesota
7/22+12
+13+1--***
New Hampshire
7/23+6
+4+2-5
Pennsylvania
7/23+5
+6+1-8
California
7/24+12
+10+2-6
New Mexico
7/24+5
+6+1-6
Nebraska
7/28+18
+19+1-3
Kentucky
7/29+10
+9+1-9
Mississippi
7/29+11
+12+1-2
Montana
7/29+1
0
+1-3
Alaska
7/30+5
+6+1--***
Arizona
7/30+16
+19+3-6
Texas
7/30+9
+8+1-9
Alabama
7/31+18
+20+2-9
Connecticut
7/31+15
+13+2-6
Avg. Change+0.95-5.7
*The "with leaners" distinction was added to reports that were released beginning on 7/9/08. The date on which these polls were conducted (The ones that these releases were based on) stretches back to 7/7/08.
**Rasmussen has only conducted one poll in these states. Therefore, the difference was taken from between the with and without leaner numbers within the same poll in these cases.

***Previous poll had been taken after "with leaners" change had been made. We expect no out of the ordinary drop in the number of undecideds when comparing two "with leaners" interpretations.

The map above uses the "with leaners" data. If, however, we shift and inpute the "without leaners" data, the trends remain the same in 33 of the 35 states in which Rasmussen conducted July polls (post-July 7). Only Colorado and Texas would have taken different courses in the event the "leaners" were withheld. Colorado would have trended toward Obama while Texas would have favored McCain. In Texas that's probably not that big a deal. The Lone Star state is comfortably red currently. In Colorado, though, this is of note simply because the state is a toss up by FHQ's estimation. And while the trend would have changed, the overall average favors Obama regardless.

So who won July? Well, the map looks awfully yellow (pro-McCain), but the Arizona senator's gains are in states where he is already way out in front or too far behind to make much of a difference. The swing states are even allocated in terms of how many are trending toward each of the candidates. And when that sort of analysis is stretched to the lean states -- 11 states -- on both sides, McCain has the averages trending in his direction in only four (AK, NM, OR and SC). Of the 24 states where July polling was favorable to McCain, then, 14 of them were already solidly red. So, while the map is decidedly more yellow than it was a month ago, the Illinois senator is still moving in a positive direction in 12 of the 22 toss up or lean states where polling was done in July.

----
Appendix: June Revisions

One thing you'll notice immediately is that this map is even greener (pro-Obama) than the original map was. This revised version incorporates several polls that were released after I posted the first look at the changes during June. Granted, I posted that on July 1 and Rasmussen released a series of polls during the latter half of that holiday week. Additionally, all those Zogby polls are factored as well, making for an inclusion of around 40 new polls.
[Click Map to Enlarge]

Those Zogby polls were seen as favoring Obama, so they tend to have biased the map in the Illinois senator's direction. McCain, for example, had six states that trending toward him in June with the late releases excluded. However, when those are added in, the McCain total drops to four. Missouri, Oregon and South Carolina all flip to Obama, while the Zogby poll in Illinois pulled the average for the Land of Lincoln down. How is it then that the Zogby polls favor Obama, yet his home state is trending against him? Well, Illinois, prior to June, had had only one poll conducted in the state -- a February Survey USA poll that gave Obama a 29 point edge. In other words, Obama pretty much had nowhere to go but down there. And despite the shift, Obama still maintains a strong advantage over McCain in the state. With the inclusion of the additional (late-breaking) polls, Obama's bounce throughout June looks even bigger.


Recent Posts:
About Those Rules: What Obama's New Florida/Michigan Stance Means for 2012 and Beyond

The Electoral College Map (8/3/08)

So, Who's Going to Win This Race? The Forecasts are Starting to Come In

Tuesday, July 1, 2008

Did Obama Bounce Everywhere in June?

There has been a lot of chatter since Barack Obama effectively won the Democratic nomination on June 3 about an Obama (or unity) bounce in the polls. I was--and still am--skeptical about the bounce based on semantics. As I've argued, a bounce--like those after a convention--implies that there is a sudden uptick in polling following a certain event that will eventually regress to the mean (the one established prior to the event occurring). However, has there been a bounce or just an increase in the margin between McCain and Obama that reflects the conditions on the ground (unpopular president and war, poor economy, etc.)? Let's look at that question with the data we now have from June.

[Click Map to Enlarge]

The map above shows how much FHQ's weighted averages have changed as a result of the polls that emerged during June. Those states in green show varying degrees of movement toward Obama while the yellow states reflect a shift toward McCain (White states are states where no polls were conducted in June. There was no case where there were new polls, but no change in the average.). From the outset, it is apparent that this map is largely green. There were 37 states where polling was done in June and of those, 31 shifted in Obama's direction. The states that moved toward Obama the most, as a result of the Illinois senator securing the Democratic nomination, are, for the most part, the states where he trailed by the largest margins prior to that point. We see big jumps in the Appalachian states where Obama lost primaries to Hillary Clinton: West Virginia, Kentucky and Tennessee. That grouping of states stretches to encompass Arkansas and Oklahoma as well. Beyond that, there are a handful of typically blue states in that have shown significant shifts toward Obama (A significant shift is defined as anything greater than a two percentage point jump in the weighted average during June. 2 points! Yeah, keep in mind that we are talking about changes in the average here. In other words, some states are more prone than others to significant shifts. On the one hand, large outliers--at least by comparison to the pre-existent data--will pull the average in their direction. But states that have had comparatively little polling are much more vulnerable to bigger shifts.): Massachusetts, Washington and Maine.

And what of McCain? Well, it isn't all bad for the Arizona senator. The momentum is against him, but he is making gains. And the places where he has made gains during the month are swing states: Nevada, Colorado, Missouri and--depending on who you talk to--Oregon and Iowa.
With the exception of Missouri, those are all Obama states in the current electoral college breakdown and comprise 28 electoral votes. That would be enough to pull McCain within 2 electoral votes of victory, but would still put him behind Obama. While Oregon and Iowa may be moving in his direction, some of the other states would potentially be easier pick ups for McCain (Think light blue states on the electoral college map.). States that are both light blue on the electoral college map and light green on the map above (So, Obama toss ups and slight Obama gains during June) are probably more likely targets for McCain than Iowa and Oregon. New Mexico and Pennsylvania fit that bill.

Of the states in white, there are only a couple that seem like they could be competitive in the fall, yet did not have any polls conducted in June. Connecticut and Montana are both leans toward Obama and McCain, respectively, but have shown signs of being in play. They are both less intensely red or blue than they have been during recent cycles. And both are cases where more polling will help to clear the picture. Also, because each has had only a few polls during all of campaign '08 (primary season, too), both are more likely than others to move significantly in one direction or the other.

Was June a bounce month for Obama, though? There's definitely an uptick in the poll numbers since he wrapped up the Democratic nomination, but as of now, there hasn't been any detectable reversion to the pre-June mean.


Recent Posts:
The Electoral College Map (6/29/08)

The National Popular Vote Plan...and Other Ways of Reforming the Electoral College

The Electoral College Map (6/25/08)