Primarily a Bad Sign

We can start with yesterday's AP piece trying to divine broader meaning from voter participation in Tuesday's primary. Here's the nub:

Final, unofficial totals show more votes were cast for Republicans in every statewide race except U.S. Senate, the Democrats' most high-profile primary.

Experts say the participation rates show Republicans are energized — perhaps to beat Democrats, but perhaps to either support or defy the nascent tea party movement. Lessons for fall are still being determined.

I'll mostly go along with that, especially since the participation rate parallels the polling data showing that Dems are demoralized and Repubs are energized. But it's worth noting a few factors that may exaggerate Tuesday's numbers:

• The Roll-off Difference. Studies show that Democratic voters are more likely to roll-off, that is vote only the top of the ticket as opposed to filling out the entire ballot. This is, for example, why Republicans continued to win judicial races even in the big Dem cycles. Setting aside arguments about what this says about the respective parties, its certainly possible that Dems are less likely than Republicans to vote in all the uncontested races.
• New Ballots. The new optical scan ballots make it more of a pain in the butt to vote. Filling in an oval is that much more tedious and annoying than punching a button or tapping a screen. Not that anyone would not vote in a real race, but when there is a real (if minimal) cost to doing something that matters not at all, fewer people will do it. This may have heightened the roll-off effect.
• The Tea Party. While the Senate campaign was big for the true political junkies on the left, the Republicans had a near civil war over the Auditor's race in which the Tea Partiers had there guy against the establishment Party pick. This arguably injected an energy into turnout on the Republican side of the ballot.
• Things Change. This an events-driven election cycle. The prime mover in this election is not big ideas about the role and size of government, it's that people want something/anything to happen so they start hurting. It's unlikely the economy -- and in particular the employment numbers -- will pick up appreciably, but if they do, the picture changes.

Make no mistake, things are not good for Democrats this cycle. But it is still good to see the entire picture.

UPDATE: Apparently the GOP's "voter enthusiasm" advantage is dampening. h/t Progress Ohio's Twitter feed.

The Good and the Bad in Dispatch Election Analysis

Like it or not, the Columbus Dispatch is the paper of record when it comes to statewide politics. Between some political reporters and the natural proximity to politics central, the paper will run more stories and have more statewide information than anyone else.

The Good.

The best Dispatch post-mortem so far is a story analyzing Obama's 88-county strategy. The story doesn't stop at the obvious (scoreboard; it worked.) Instead the writers dig through the numbers to show that the strategy paid off. Obama improved Kerry's performance in most regions throughout the state, winning some new counties in the northwest and blunting the Republican advantage on its home turf. Moreover, the paper notes that Obama actually underperformed Kerry in Appalachian counties and in the Mahoning valley.

Click. Read. The whole article is worth the time.

The Bad.

The headline alone tells you we're heading off the tracks: "As Usual, Dispatch Poll Was Accurate." As usual? Do you guys remember any of your earlier polls? Or are we engaging in a bit of revisionist history.

Let's review. As accurately noted, the Dispatch poll showed a dead heat in 2004 just before the election which was in the MoE of Bush's two percent win. But a month before, the Dispatch poll was an outlier, showing a seven percent Bush lead when the average of Ohio polls showed two percent.

The next year, the Dispatch suffered a humiliating stumble, predicting victory for two of the four Reform Ohio Now amendments before all four were crushed at the polls.

In '06 the poll showed Strickland ahead by six and a half points more than his ultimate margin of victory and was, again, the only poll to do so. For a poll that claims a 2% MoE, that's ugly. By the way, the RON poll and the '06 gov poll were so far of that they gave Fitrakis fodder for his usual moonbattery.

Happily, no significant statewide races last year offered the paper a chance to issue a meaningful poll.

"As usual" it was accurate? A better headline would have been "Dispatch Poll Fails to Embarass the Paper This Time."

Obama Leading Big Among Ohio Early Voters

FiveThirtyEight.com highlights SurveyUSA's breakout of early voters in five key states including Ohio. The result is an overall average 23 point lead for Obama within those states. In Ohio Obama leads 57-39% among people who have actually voted, versus 49-45% among likely future voters and 50-45% overall. In SUSA's Ohio poll, 12% reported having already voted, versus 88% of the sample identified as likelies.

A few notes.

• The big variables among polls are sampling and methodology for identifying likely voters. Obviously this result takes care of the second. As to the first, it should be noted that while the overall poll includes a statistically significant sample, we can be less sure that the sample of early voters is so.
• To the extent this means anything, it offers evidence of 1) the enthusiasm gap and 2) better organization in the Obama ground op.
• Since the early voting question takes care of likely voter identification, it's possible that the gap to some extent reflects a long-suspected theory that early voter panels are washing out newly mobilized (and registered) Obama supporters. Without knowing how SUSA does likely voter panels, it's impossible to take this further.
  
• One reason Obama supporters may be going early in disproportionate numbers is that folks with election day time issues -- especially students and hourly wage-earners -- have a strong incentive to take advantage. Meaning that (as suspected) early voting over time offers an advantage to Dems turning out their constituencies.
  

Final PA Predictions

Over at Pollster.com we get this assessment of the polling trends:

Clinton has increased her lead in the trend estimates over the course of the last polls to 6.6 points using the standard estimator, and to 8.4 points using the sensitive estimate. Last minute polls have given her bigger margins.

Now the key question is whether undecideds push her over a 10 point win, or whether increases in turnout by new "unlikely" voters raises Obama's total.

My guess is the latter. First, the weather is nice. More importantly, Obama's money and grassroots volunteer strength hopefully have translated to a superior GOTV organization.

While Team Hillary has run out the same "If he doesn't win he's a loser" argument they threw against the wall before Ohio, I hope the media maintains a realistic BTE/WTE. Right now the line is a 7-8% Hillary advantage. Six or less is BTE for Obama; I'll put my final prediction at five. Anything over nine is WTE and makes closing that much harder.

Once again Obama had a chance to put Hillary away and once again he has let her up off the mat, this time with his ill-advised speculation about bitter Pennsylvanians. Whatever spin Ohio bloggers may want to put on it, he gaffed in a bad way and will have to make his case in Denver as a result.

New SUSA Poll: Hillary Continues to Slip

A new Survey USA poll commissioned by WKYC TV shows Hillary Clinton losing another three points to Barack Obama. The new poll shows Clinton at 50% and Obama at 44%, compared to 54-43 a week ago. This is also the third week in which Clinton's support has dropped in the SUSA poll. (Comparing results by one polling company is somewhat more reliable than comparing among different polls.)

One factor that may be adding to Obama's numbers -- the percentage of likely voters keeps going up. In the Feb. 10 poll, the SUSA voter screen found 39.7% of registered voters to be likely voters; in the Feb. 17 it was 41% and in the latest poll, 44%. The very small percentage of undecideds in the poll (smallest among the polls testing Ohio) suggests that SUSA's screen is particularly tight.

Pollster.com on Ohio vs Wisconsin

A post on Pollster considers the following puzzle: Ohio and Wisconsin are demographically similar and the differences would tend to militate in favor of Obama (more urban, larger Black population), so why is Clinton doing so much better in Ohio polls than she did in Wisconsin leading up to the primary?

The author, who declares himself an Ohio native, breaks down the demographics and comes up with the provisional half-answer that:

The answer, for the moment, appears to stem mostly from her continuing strength among Ohio's downscale white Democrats. In Wisconsin, as noted above, Obama ran slightly ahead of Clinton among less-educated white voters. However, in both the Quinnipiac poll conducted two weeks ago and the ABC/Washington Post poll done earlier this week, Clinton continues to hold an enormous lead among less-educated white voters.

Which leaves the obvious follow up: Why is Clinton doing better in that cohort in Ohio than in Wisconsin. My hypothesis -- trotted out in this brief exchange with Redhorse -- is the influence of Appalachian culture in the Ohio electorate. From speaking with people who study Appalachia, verified by my personal observations living around here, is that as a whole¹ Appalachians tend to be suspicious of institutions, outsiders and people who are different. Appalachians tend to respond well to populist appeals -- such as those Hillary has been making -- and they have a tendency toward less-than-progressive attitudes about race.

The regional cross-tabs in the Survey USA poll offer equivocal support for my hypothesis. On the one hand, Obama's support is by far the lowest (24%) in SE Ohio -- the part of the state that is physically within Appalachia. On the other hand, Hillary has the strongest support in the Columbus and Toledo areas. I know Columbus has a fairly significant population of Appalachian migrants -- don't know about Toledo.

There's a number of interesting studies one could do with this. I don't have answers, just a couple of suggestions.


¹Yes, I am generalizing here. That's what you do when you try to predict the behavior of a group. Everything I've seen says the generalization is valid. The danger comes when one tries to use a generalization to predict the behavior of an individual. When I meet someone whose parent moved up from the hills of Kentucky I try not to assume that he/she harbors anti-Black attitudes. This is dicey business given that a) the former is valid social science, but the latter is bigotry and b) that sort of bigotry can influence the generalized assumptions in the first place, throwing the whole enterprise into a soup of elitist snobbery. Nonetheless, it doesn't do to pretend the issue doesn't exist in the presence of evidence that it does.

Obama's S-USA Numbers Aren't Pretty

The new Survey USA Ohio poll shows Hillary Clinton leading Barack Obama 56%-39% (h/t Taegen Goddard). Morning giddiness over the prospect of Ohio delivering Obama the nomination has disappeared in a bracing reality slap. This was the first poll taken since Edwards dropped out, and the first since Jan. 31.

S-USA notes that "Clinton's lead comes entirely from women, where she leads by 29 points." Fine, but to me the most salient -- and disheartening -- statistic is: Among the 69% of voters who feel they have firmly made up their minds, Clinton's lead beats the overall result -- 59%-39%. Furthermore, only 2% overall are currently undecided.

Obamaniacs were hoping to see more Edwards voters swing to Obama, or at least see softer support for Hillary.

Jerid offers the whistling-past-the-graveyard take that Obama gained 20% since Quinnipianc's December poll. Well, that's a happy thought. But Hillary gained 11%. More to the point, she expanded her lead from the January Dispatch poll. Much much more to the point, she maintains a double-digit lead, is six points in the black and has respondents voicing solid support for her. None of this is good news. (And, oh by the way, you are always on shaky ground trying to divine a trend by comparing different polls using different methodologies.)

Any good news? Well, Obama seems to have done better as they split Edwards support and undecideds -- just not well enough. And the poll only netted 11% independent likely voters whom Obama is winning 48-42%. So there is some room for persuading I's to get out and vote.

The poll would be great news if Obama was only trying to mitigate damage. But if Hillary sweeps Ohio and Texas, it seems unlikely Obama will persuade enough superdelegates to take a chance on him. He needs to turn some folks around, ramp up his GOTV, visit often and pray a whole lot. He may even need to choose either Ohio or Texas and set down some roots there. And yes, hopefully that will be here.

Hillary's Electiblity Argument

Chris Cillizza just went up with a post based on a conference call with Hillary's pollster/strategist Mark Penn on the issue of Hillary's superior "electiblity." Here's the nub of Penn's argument:

"The Republican attack machine redefines the Democratic candidate," said Penn, pointing out that Vice President Al Gore and Sen. John Kerry (Mass.) both felt the sting of the GOP efforts during their respective national bids. Penn added that while Clinton is well known in national circles, Obama is less so -- a lack of name recognition that leaves the Illinois senator open to being defined by the Republican nominee. "Hillary has withstood this process and this will make a tremendous difference if she is the nominee," he said.

Also, she is allegedly superior on the issue of national security.

Nominating Clinton would "block [Republicans] from playing the national security card," argued Penn, adding that Republicans have already begun to attack Sen. Barack Obama (Ill.) on his foreign policy credentials. (In an interview yesterday with Fox New Channel's Chris Wallace, President George W. Bush had this to say of Obama's foreign policy bona fides: "I certainly don't know what he believes in.")

Cillizza summarizes the argument thusly:

The central difference in the electability appeals by the two campaigns is temporal.

The Obama campaign argues that the way to best understand who is the more electable is to look at current polling and past results to see who leads the likely Republican nominee and who is better able to lure crucial independents to the Democratic cause. The present is what matters, says Obama.

For Clinton, it's the future that's the issue. Sure, they argue, Obama may be ahead right now, but Republicans have only begun to define him, a process that would strip away much of his independent support and leave him on the losing end of a race against McCain.

Really, I must put the blog down and get back to grading. But a few points first.

First, national security. Hillary's argument for neutralizing McCain on national security is that she was for the Iraq Use of Force resolution before she was against it. Seems to me that Dems have tried that argument before. Let me look up how that went and get back to you.

As far as Hillary "neutralizing" the attack machine, by what measure? A year ago Harris Interactive found that fully 50% of Americans would not vote for Hillary under any circumstances. This despite pluralities that approve of her track record. Why? Because of the neutralized Republican Attack Machine. Hillary hasn't neutralized it; she's merely found two venues (New York State and the Democratic Primary) where it doesn't have much sway.

Fact is, there are two Hillary Clinton's in the race. There is the cautious consensus-building centrist she has been in the Senate, and there is the shrill, child-emancipatin', Tammy-Wynett-bashin', Bubba-enablin', hyper-ambitious caricature drawn by the attack machine. The fact that the second Hillary remains in this race is big trouble for the first Hillary. And it's a problem generated entirely by the attack machine she has supposedly neutralized.

The argument really shouldn't be temporal, but geographic. Because of her negatives, Hillary constricts the electoral map. She will compete in all the swing states, but that's about it. Obama can compete in a broader array of states, particularly in the South. In a year when Dems will probably maintain an advantage in fundraising, a candidate who can force the Republican opponent to spend time in money in more states has a significant advantage.

Watching New Hampshire

It's 9:50, the results are half in and Hillary is bouncing between a 2 and 4 percent lead. With that, CNN is not calling the race. I'm guessing the network's exit poll shows Obama winning. Given the sea change, though, it's hard to put much hope in polls.

Changing the media's favored narrative, but here's a try. Hillary as the Comeback Kid is ridiculous. She was SUPPOSED to win New Hampshire. Bill Clinton is wildly popular among New Hampshire Dems and has been camped there for months. Two weeks ago, and almost without exception before that, her lead was double digits. That lead evaporated. She wasn't down, then prevailed, she blew a lead and (to some extent) recovered.

Jeffrey Toobin on CNN raised a good point -- Dem turnout was huge, Repub turnout was down. Dems are in good shape for the general, though I still question whether Hillary is electable in the general.

10:45 CNN is reporting that AP is calling it for Clinton, but that they are "unprepared" to do so. I can't wait to see their exit poll results.

I never considered Arkansas part of the "deep south," but Huckabee referred to it as such.

10:34. NBC just declared for Clinton. And Chris Matthews calls it a "stunning upset." If UNC falls behind North Carolina A&T, then ekes out a win, it's not a "stunning upset." It's a close call for a heavy favorite.

All that said and notwithstanding the asinine media narrative, tonight certainly let Hillary off the matt. It would be nice if someone could dig into how it happened. Did the Clinton campaign turn out the vote better? Did the "likely voter" panels wash out people who ended up showing up? Unfortunately, it's now an accepted fact that Hillary won by tearing up at a diner yesterday. Move on to South Carolina where the winner will be the candidate who says the nicest things about low country shrimp or some such.

Will Richardson bow out now? Edwards says he's in it to the end -- translation: he's in it until he gets blown out in So. Carolina.

10:46. CNN finally calls it.

Internet Usage up to 80%

According to a new Harris poll:

The survey, which polled 2,062 adults in July and October, found that 79 percent of adults -- about 178 million -- go online, spending an average 11 hours a week on the Internet.

"We're up to almost 80 of adults who now are online, or are somehow gaining access to the Internet. That's a pretty impressive figure," said Regina Corso, director of the Harris Poll.

The results reflect a steady rise since 2000, when 57 percent of adults polled said they went online. In 2006, the number was 77 percent.

The page on Harris Interactive includes plenty of additional data, including this tidbit -- 97 percent of those who use computers are now internet users as well, also an all-time high.

Quinny Numbers: Strickland Soaring and Trouble for George

The latest Quinnipiac U poll in Ohio is mostly good news for Governor Strickland. He has a 61% approval rating and a microscopic 15% disapprove of his job performance. One unreported stat: 15% disapprove of Strickland and all those people write conservative blogs (grin.)

Seriously, the numbers reflect a few things. First off is, the Governor's brilliant budget strategy. I've hashed this before but it's worth repeating: The Governor proposed a modest budget that moved the ball on his priorities but carried no deal breakers and didn't exceed his grasp.

Second, I think people understand that we have a Governor who actually has leadership ability, in contrast to the blank space that was Bob Taft. That in itself may be enough to give Strickland an extended honeymoon.

The potentially bad news is the perception of Ohio's economy. Right now the Governor is riding high in spite of the economic problems -- real and perceived -- bedeviling Ohioans. This won't last indefinitely.

Meanwhile, the numbers on the state's U.S. Senators are a study in contrasts. Sherrod Brown and George Voinovich have similar approvals - Brown 44%, Voinovich 45%. But Voinovich's disapproval rating of 34% dwarfs Brown's 19%. Moreover, much of Voinovich's strong disapproval comes from defecting Republicans. 32% of Republicans disapprove of Voinovich, while only 12% of Democrats disapprove of Brown.

Similarly, each Senator has similar approval numbers among Independents -- Brown 41%, Voinovich 40%. But for Brown only 22% of Independents disapprove versus 35% for Voinovich, suggesting that he is weak among both left-leaning and right-leaning Independents. Finally, Voinovich's disapprovals among White Born-Again Evangelical Christians are an ugly 34%.

This confirms my suspicions about a fundamental tension for Republicans. Mike DeWine's loss was evidence that an Ohio Republican needs the right wing of the party to mobilize to win. Ken Blackwell's crushing defeat indicates that the right wing alone can't put an ultra-conservative over the line. And the numbers on Voinovich are the latest indication that the party's right wing is unwilling to compromise on moderate politicians for the sake of winning elections.

All of which is good news for Democrats, but we can't be complacent. The current environment makes it difficult for Republicans to win, but Republicans losing is not the same as Democrats winning. The new crop of statewide officials needs to start notching some real successes if this is to be anything more than a temporary deviation from Ohio's rightward trend.

The B-W Education Poll -- Found

Well, actually I didn't find it so much as project investigator Tom Sutton dropped a comment in the original post giving the address. He had been alerted to my post by Ryan of BlogginRyan (formerly based in SummitCo, now in Atlanta.) And in retrospect I should have written to Tom before kvetching about the lack of online information. I was irate after like ten different searches failed to dig up either the report or the Center website, but no excuse. Bad blogger! No coffee!

So, the website for B-W's new Public Interest Research Center is here and the report pdf of the actual survey is here. From the latter we find that the question about the GIRFOF was:

15. A coalition of education groups is trying to put an amendment to the Ohio Constitution on the ballot in November. If approved, the amendment would require the State Board of Education to determine the cost of a good education for every student in Ohio, and then require the state legislature to provide the funding. The proposal also includes a cap on the level of property taxes paid by senior citizens. What is your opinion?. . .

Which is a pretty solid summary of the proponents' description of the amendment. It doesn't mean GIRFOF fans should breath easy. Given the aforementioned KnowledgeWorks findings that few people have heard of the amendment, this is an indication that when pitched in this way, you get 63% approval at first hearing.

A "No" campaign will take that as a starting point and come up with ways to either move people off the belief that the amendment will do that or move them off the belief that a system like that is a good idea or -- and this is the biggie -- make them nervous about the details. Think about how likely it is that come election day, everyone is going to go into the voting booth thinking only that the amendment is as described in the question above. Now you understand why I say that 13% isn't much room to fall.

The poll itself covers far more ground than just the issues raised in Stephens' article. If I find anything else interesting as I dig around, I'll post.