PPP: Obama leads by 7 in OH

Ohio is another crucial swing state, worth 18 electoral votes (EV) out of the required 270 EV. Polling this year has shown Obama with varying leads, with the exception of a February Fox News poll showing Romney leading by six points. Even Rasmussen gave Obama a four-point lead in April (46/42). HuffPo averages out the polling to show Obama with a five-point lead (47/42), making it a State that "leans Obama". The President won Ohio 51/47 in 2008.

The latest poll from PPP for OH (875 RV, ±3.3% MoE, 5/3 - 5/6) gives Obama a seven-point lead (50/43).
Let's look at the ideological breakdown of the sample to dispel notions of bias:

Q13 Would you describe yourself as very liberal,
somewhat liberal, moderate, somewhat
conservative, or very conservative?
Very liberal ...................................................... 9%
Somewhat liberal ............................................ 18%
Moderate......................................................... 31%
Somewhat conservative.................................. 22%
Very conservative ........................................... 19%

So, that's 41% conservative, 27% liberal.

PPP's Tom Jensen sums up the phenomenon that has allowed Obama  to assume this lead:

Ohio voters don't love Obama. They're evenly split with 48% approving and 48% disapproving of him, although that is an improvement from the negative numbers he's posted during most of his time in office. Obama's lead in the state may be driven more by the fact that Ohioans just don't much care for Mitt Romney. 37% have a favorable opinion of him to 53% with a negative one. That includes a 33/59 spread with independents. Romney's performance in Ohio in the primary wasn't terribly impressive and his issues seem to be extending to the general election.

We see this frequently in state polling. It's not so much that Obama is immensely popular, but rather that Romney is just not well-liked at all. This also means that significant margins of key demographics for Romney show soft support -- a greater percentage of a particular demographic saying that they will vote for him than the percentage that has a favourable opinion of him. Voters exhibiting soft support are less likely to turn out, volunteer, donate, or "talk up" their candidate, though they don't usually flip to the opposing candidate.

Even worse than Romney's "favourables" are those for Ron Paul:

Q2 Do you have a favorable or unfavorable opinion
of Ron Paul?
Favorable........................................................ 26%
Unfavorable .................................................... 51%
Not sure .......................................................... 23%

Paul loses to Obama in a match-up 48/40. He loses women to Obama in the match-up by 24 points, splits white voters (45/43 O+2), loses voters 45 and under by at least 15 points, and loses moderates by nearly 30 points. He has all of Romney's weaknesses with key demographics, but he holds a higher percentage of undecided voters in most of them than Romney does.

We are now well past "Rosen-gate", where Ann Romney was supposed to have won over female voters and erased the gender gap by playing the victim card. It hasn't worked. Jensen notes:

The GOP's struggles with women and young voters really show themselves in Ohio. Obama's up 55-36 with women and has a 62-30 advantage with those under 30. If you extend the definition of 'young' voters to those under 45 Obama still holds a massive advantage at 56-35. Romney's winning seniors 49-45 but he needs a much bigger lead than that to make up for his weakness with young people.

 Seniors are a reliable Republican demographic, and they are only breaking for Romney by four points. Men break for Romney by six points (50/44, 7% undecided), which isn't strong. It hardly begins to counter the nineteen-point lead  Obama has with women. Moderates, as we've seen in so many States, break strongly for Obama. Apparently they don't believe that Obama is some kind of radical Marxist. In Ohio, Obama wins moderates 56/31 (13% undecided) in a match-up against Romney. The moderates that do support Romney show five points of soft support.

Conservatives, as one would expect, are Romney's biggest supporters. Those who describe themselves as "somewhat conservative" break overwhelmingly for Romney in a match-up 67/25 (9% undecided). Unfortunately for Romney, 67% of the 22% of the electorate that describe themselves as "somewhat conservative" is only 14.7%. He gains less than 15% from this overwhelming margin toward his 43% support. And only 59% of this demographic has a favourable opinion of Romney, which leaves him with eight points of soft support among those "somewhat conservative" voters.

For those describing themselves as "very conservative", the situation is much the same. Romney wins them by an even larger margin (79/15, 5% undecided). However, this only nets him another 15% of the electorate. Out of his 43 points of support, nearly 30 points (69%) of his support is from conservatives. The price of this is losing moderates steeply to Obama. And only 65% of those "very conservative" voters have a favourable opinion of Romney. That leaves a considerable 14% of soft support with this demographic.

Let's look at Obama's soft support. Among moderates, 58% have a favourable opinion of him -- two points more than break for him. Actually, with these subgroups, these small margins are meaningless. Among women, 53% are favourable (negligible two points soft support). Among white voters, 44% have a favourable opinion, leaving three points of soft support. Obama's senior supporters are solid at 47% favourable, compared with his 45% share of the demographic. Independent voters break for him at  43%, while 41% approve. Even the 25% of "somewhat conservative" voters breaking for Obama show 23% favourable opinions of him. He shows five points of soft support among "somewhat liberal" voters, but he's solid with "very liberal" voters (for what their 9% of the electorate is worth).

With Obama's support being so firm, it's unlikely that negative attacks will be effective against him. Romney, however, is vulnerable to depressed enthusiasm among his base -- especially among the "very conservative" voters showing considerable soft support for him.

Even worse, Romney's team has shown no ability to make Romney himself look better. Instead, they have concentrated on trying to make Obama look bad.

Can a "rally" save Romney in OH?

The current delusion from conservatives is that, after the nomination, Republicans will "rally" to their nominee and provide the edge needed for a Romney victory. 

I'm going to examine the recent polling of one swing State and determine how realistic this "rally" scenario really is. 

Quinnipiac polled OH (1246 RV, ±2.8% MoE, 3/20 - 3/26) and found Obama leading 47/41. The survey's demographics are here. The cross-tabs are here. 

The weighted numbers from the sample show 29% self-identify as Republicans. These Republicans break 83/7 for Romney when matched up against Obama. Beyond that, 4% "don't know", 2% "won't vote", and 3% will vote for "someone else". 

Let's assume that the 7% of Republicans saying they will vote for Obama flip to Romney, and that all 4% undecided Republicans also rally to Romney. This is 11% of the Republicans. 11% of 29% is a whopping 3.2% of the general electorate. Obama leads by six. 

This, then, is the conclusion: if every Republican in Ohio flips to Romney, and actually turns out to vote, Obama would still win the state by about three points. There is about 5% margin of error (MoE) in the Republican-only sample, which could go either way. The best-case scenario is that a Republican "rally" could put Romney ahead by two points, and that's stretching at the limits of reality here. 

Whatever it would take to rally every Republican would probably alienate Independent voters (33% of the sample). Even a five-point gain for Obama among Independents would make the race a virtual toss-up, and that would occur if half of the undecided Independents voted for Obama. If Obama won over the 5% of undecided Democrats (33% of the sample), that would also make the race a virtual toss-up. Again, this is taking into account the unrealistic assumption that all of the poll's margin of error falls to Romney's advantage. If only half of the MoE falls on Romney's side, then Obama narrowly wins.

Romney fails to win a majority of white voters (45/41), with 7% undecided. With the white vote as 83% of the sample, Romney would gain 5.8% of the general electorate if he could pick up every undecided white voter. This would make the race a toss-up. He would need the MoE of 2.98% to actually pull ahead, but it would be a narrow victory.

A Republican "rally" could make OH a "toss-up" at best, and that's a coin flip that Romney cannot afford to lose. Obama, on the other hand, can afford to lose it and pick up those electoral votes in other States.

Quinnipiac: Obama leads in FL, OH, PA

In a new poll from Quinnipiac, Obama leads in three crucial swing states: FL, OH, and PA. 

The GOP cannot afford to lose two of these States, and Obama basically has the election if he wins Florida. With VA looking bad for the Republicans, and CO, NV, and NH looking close, the GOP needs to win all three of these just to stay in contention. Let's look at the results:

Florida: (1228 RV, ±2.8% MoE, 3/20 - 3/26) 49/42 (O +7)

Florida voters give Obama a split 47 - 49 percent job approval rating, and say 50 - 47 percent he deserves to be reelected.

The economy is in a recession, 68 percent of voters say, but 57 percent say it is beginning to recover. Romney would do a better job on the economy, 48 percent of voters say, while 45 percent pick the president, but Obama tops Santorum 50 - 39 percent on this issue.

Oil companies are most to blame for gas prices, 32 percent of voters say, while 23 percent blame oil-producing countries most; 18 percent blame Obama and 16 percent blame supply and demand.

Ohio:  (1246 RV, ±2.8% MoE, 3/20 - 3/26) 47/41 (O +6)

Ohio voters also give Obama a split 47 - 49 percent job approval rating, and split 48 - 48 percent on whether he deserves to be reelected.

The economy is in a recession, 68 percent of voters say, but 58 percent say it is beginning to recover. Voters split 45 - 45 percent on whether Obama or Romney would do a better job on the economy, but Obama tops Santorum 48 - 41 percent on this issue.

Oil companies are most to blame for gas prices, 39 percent of voters say, while 19 percent blame oil-producing countries most; 18 percent blame Obama and 14 percent blame supply and demand.

Pennsylvania:  (1232 RV, ±2.8% MoE, 3/20 - 3/26) 45/42 (O +3)

Pennsylvania voters disapprove 50 - 45 percent of the job Obama is doing, still negative but his best score in recent surveys, and say 50 - 46 percent he does not deserve to be reelected.

The economy is in a recession, 65 percent of voters say, but 57 percent say it is beginning to recover. Voters say 48 - 42 percent that Romney would do a better job on the economy than Obama, but Obama tops Santorum 49 - 41 percent on this issue.

Oil companies are most to blame for gas prices, 34 percent of voters say, while 25 percent blame oil-producing countries most; 17 percent blame Obama and 15 percent blame supply and demand.

Romney's favourables are, as in other States, worse than Obama's.

FL: 41/36 (Obama 47/49)

OH: 36/47 (Obama 47/49)

PA:  37/38 (Obama 45/50)

Previous polling for FL:

• Quinnipiac (1518 RV, ±2.5% MoE, 1/19 - 1/23) 45/45 (tie)
• PPP (700 RV, ±3.7% MoE, 11/28 -12-1) 45/44 (O +1)

Previous polling for PA:

• PPP (500 RV, ±4.4% MoE, 11/17 - 11/20) 45/45 (tie)
• Susquehanna (800 RV, ±3.5% MoE, 2/2 - 2/6) 45/43 (R +2)
• Muhlenberg (625 RV, ±4% MoE, 2/15 - 2/21) 48/37 (O +13)
• Franklin & Marshall (592 RV, ±4.0% MoE, 2/14 - 2/20) 41/33 (O +8)
• Rasmussen (438 LV, ±4.5% MoE, 2/8 -2/23) 45/44 (O +1) 
• PPP (689 RV, ±3.7% MoE, 3/8/ - 3/11) 49/42 (O +7)

Previous polling for OH:

• PPP (1022 RV, ±3.1% MoE, 11/4 - 11/6) 50/41 (O +9)
• PPP (820 RV, ±3.4% MoE, 1/28 - 1/29) 49/42 (O +7)
• Marist (3079 RV, ±1.8% MoE, 2/29 - 3/2) 50/38 (O +12) 

Additionally, the PurplePoll (1424 LV, ±2.6% MoE, 3/16 - 3/19) polled a composite of NH, OH, and PA. Obama led Romney 48/43 (O +5). 

Marist: Obama strong in VA

In tandem with the last post involving OH, there is another Marist poll (2518 RV, ±2.0% MoE, 2/29 - 3/2) that shows Obama winning by double digits in the crucial swing state of Virginia. 

In a match-up with Romney, Obama prevails 52/35 (O +17). A slightly wider margin is shown in a match-up with Santorum (54/32). Likewise with Ron Paul (53/32).

Also, the overwhelming majority of VA Republican primary voters think Romney will be the eventual nominee:

Regardless of whom you support, who do you think will be the Republican nominee to run against President Obama?

• Mitt Romney 78%
• Rick Santorum 9%
• Newt Gingrich 5%
• Ron Paul 3%
• None of the above 1%
• Undecided 5%

 Also instructive, given the polling results, is this question:

Which one of the following candidates do you think has the best chance to beat President Obama?

• Mitt Romney 65%
• Rick Santorum 14%
• Newt Gingrich 9%
• Ron Paul 4%
• None 3%
• Undecided 5%

I examined the electoral math here and here. 

Again, assuming CO and VA are won by Obama, he only needs one of four States in his column to reach 270 electoral votes: NC, PA, FL, or OH. 

Colorado is a toss-up. 

Previously, VA has been shown tied in one poll, and Obama leading in three others:

• PPP  (48/42, O +6)
• Quinnipiac (47/43, O +4)
• Rasmussen  (49/43, O +6)

Seeing a fourth poll with the strongest lead for Obama yet is very encouraging.

In spite of three polls showing a lead, the National Journal (highly conservative) calls this new Marist poll an "outlier":

The new poll appears to be an outlier -- a conclusion that is supported by other results from the poll showing President Obama with a massive, 17-point lead over former Massachusetts Gov. Mitt Romney. The Roanoke College poll, on the other hand, showed Romney leading Obama by one point.

I'm not sure how four polls showing Obama ahead can mean that he's not ahead.

With VA looking more solid than ever, regardless of what the lead really is, the importance of other polling showing Obama ahead strongly in OH and PA is huge. 

FL has been polled as a tie by Quinnipiac and PPP. It's basically a coin toss, and if the Republicans lose that coin flip, it's all over for them.

NC (15 EV) is also a tie, per PPP. Those 15 EV would take Obama over the finish line if CO and VA are going along with Obama. The Republicans have a very difficult path to 270 if they fail to win NC. It's a coin toss that the GOP really must win.

Obama's solid base of 239 EV, plus VA with 13 EV, leaves us at 252 EV. 

From there, Obama only needs 18 EV to win. 

NC + NH = 19 EV.

FL = 27 EV 

OH = 20 EV 

PA = 21 EV

CO + NV + NH = 19 EV 

CO + MO = 18 EV 

CO + AZ = 18 EV 

MO, AZ, NH, and NV are all currently showing as toss-ups, though the Republican candidate is likely to have an advantage in all of them. Still, dropping two of them would be deadly. 

A possible obstacle to these plans is Iowa, and its 7 EV. 

With Selzer's outlier poll showing Obama losing IA by two points to Romney, it may be the only one of Obama's 239 EV base that is not so solid. 

This makes CO (9 EV) even more important, in order to cover the possible loss of IA. 

Still, with PA and OH looking more solid, that path alone would allow Obama to drop CO, IA, NC, NV, NH, MO, FL, and AZ and still win. 

Likewise, winning FL alone would allow Obama to drop IA, CO, and all of the other close States - and still win.

NC and either OH or PA would put Obama over 270 without needing IA or CO.

Of course, this leaves a lot of ground for the Republicans to cover. Unlimited PAC money only gets you so far, and there are Senate races, too. Polling isn't showing a lot of good opportunities for Republicans at this point, and Maine is now a new question mark. It's possible that the GOP could end up with a net loss of Senate seats. They will certainly lose on net House seats.

 -------------------------
Pollster ratings from fivethirtyeight.com, based on 2010 elections:

Marist: Obama strong in OH

A new poll of Ohio, a crucial swing state in the 2012 election, shows Obama with a strong lead.

Marist released a poll (3079 RV, ±1.8% MoE, 2/29 - 3/2) showing Obama leading Romney 50/38 in OH. 

From the release:

Obama Leads Possible GOP Rivals
How does the field of Republican candidates fare in Ohio against President Obama?

• Obama leads Paul, 48% to 38%.  13% are undecided.
• 50% support Obama compared with 38% for Romney.  12% are undecided.
• 50% of registered voters back Obama compared with 36% for Santorum.  14% are undecided.
• Obama — 51% — leads Gingrich – 36%.  13% are undecided.

Obama manages 50% in the match-up, in spite of an even break in approval:

45% of registered voters in Ohio approve of how President Obama is doing in office.  45%, however, disapprove, and 9% are unsure.

So, Romney is unable to pick up 7% of OH registered voters who are unhappy with Obama. Does that sound like a strong candidate to you?

It's even worse when you look at the party ID breakdown:

Ohio voters are more likely to identify as independents than they were in a similar poll conducted in October 2008.  35% of registered voters now consider themselves to be independent compared with 31% at that time.  Republican identification has decreased from 31% to 26%, and Democrats are little changed.  They went from 37% in 2008 to 38% now.

“President Obama’s advantage over his potential challengers in Ohio is underscored by the fact that more voters support him over the GOP field than approve of his job performance,” says Dr. Lee M. Miringoff, Director of The Marist College Institute for Public Opinion.

So Romney is drawing 12% above his base partisan support. And Obama is drawing 12% above his partisan base support. That would be okay if Romney's partisan base of support hadn't dropped 5% since the last election - which the Republicans lost. 

Obama now has a twelve-point lead over Romney in a State that Republicans cannot afford to lose. 

Gov. Kasich is slightly underwater in his "favourables": 40/43, with 17% unsure.

And this shows the real "value" of those right track/wrong track poll metrics:

Majority Says Nation on the Wrong Track

55% of Ohio registered voters think things in the nation are off on the wrong track while 40% believe they are moving in the right direction. Five percent are unsure.

If a majority of OH registered voters believe that we are on the "wrong track", how can Obama get 50% in a match-up? Maybe this is a stupid metric to judge the President's electoral chances by. Maybe it reflects dissatisfaction with Republicans as much as it reflects a liability for Obama. 

Also, the poll shows a strong lead for the Democratic candidate in the Senate race:

If Democratic incumbent Senator Sherrod Brown were to face off against Republican challenger Josh Mandel, what are Mandel’s chances of victory? As the contest stands, Brown leads Mandel, 47% to 37%. 16% are undecided.

Ohio has 20 EV, and is a "must win" state for the Republicans. As was discussed previously,  if Obama wins OH, he is virtually guaranteed to win the needed 270 electoral votes (EV) for victory - as things stand now.

Electoral math:

Republicans must also mind their defences in MO, NV and AZ, as they are only narrowly going to the Republicans currently. They certainly can't afford to see these base States slip away, especially since Texas is only in their favour by 7 points. Texas is the "crown jewel" of the Republican election strategy. Without TX and its 34 EV, there is no hope for a Republican to win the White House. It's still solid for them now, but it might be a swing state in 2016.

Assuming Obama can win CO and VA, he then only needs to win one of these States: PA, OH, NC, or FL. We already see him ahead in OH.

Colorado looks close, and a recent poll set VA as a virtual tie. Both add up to 23 EV. Winning OH or PA would mostly offset losing the coin flip twice in CO and VA. 

Winning both OH and PA would mean that Obama would probably win the national election. 

It's a good thing that Obama is leading in both of those States, then.

Rasmussen gave leads to Obama in VA (49/43 O +6).

Republicans are in bad position at the moment. They have to basically win three coin flips in a row: CO, FL, NC. If they lose one of those, Obama wins. 

NH is another coin flip, but only has 4 EV. It might come into play if the Republicans win CO and VA, but lose PA and OH. 

This is why any polling of CO, FL, or NC will be closely watched.

Obama has a solid base of 239 EV right here:

CA (55), OR (7), WA (11), NM (5), MN (10), MI (17), IL (21), NY (31), MA (12), VT (3), ME (4), RI (4), DC (3), MD (10), DE (3), NJ (15), IA (7), WI (10), CT (7), HI (4)

Previous polling for OH:

PPP's November 2011 polling shows Obama with a 9-point lead in OH.
In 2/12 polling by PPP,  Obama leads by 6 points.
Rasmussen, for what it's worth, shows Obama with a 4 point lead over Romney and tying Santorum.

------------------------------
 Pollster ratings from fivethirtyeight.com, based on 2010 elections: