OH-Gov/OH-Sen: Strickland Weak; SSP Changes to “Lean D”

Research 2000 for Daily Kos (7/6-8, likely voters, no trendlines):

Ted Strickland (D-inc): 44

John Kasich (R): 39

Undecided: 17

(MoE: ±4%)

R2K is now the third straight pollster to peg Strickland in the mid-to-low 40s, apparently solving the OH-Gov polling mystery. This is not a pretty chart:


One important thing to note, though: R2K has Obama’s favorables at 59-35 in Ohio, while Quinnipiac pegs them at 49-44. R2K is probably the outlier, though, as the two firms show similar favorables for Strickland, and Q’s job approval numbers for Obama now match up with PPP’s.

As a result of this recent nosedive, the Swing State Project is changing its rating on OH-Gov from Likely Dem to Lean Dem. Of course, it’s early; if Strickland can pull out of this tailspin, then we’ll be ready to adjust our rating once again as needed.

R2K also looked at the Senate race:

Lee Fisher (D): 22

Jennifer Brunner (D): 17

Undecided: 61

(MoE: ±5%)

Lee Fisher (D): 42

Rob Portman (R): 35

Undecided: 23

Jennifer Brunner (D): 40

Rob Portman (R): 36

Undecided: 24

(MoE: ±4%)

These numbers are very similar to those shown by Quinnipiac and PPP.

RaceTracker: OH-Gov | OH-Sen

19 thoughts on “OH-Gov/OH-Sen: Strickland Weak; SSP Changes to “Lean D””

  1. I don’t follow Ohio politics too closely, but I have been watching local television news from Toledo for the last few weeks, and they’ve been running fairly regular stories about various local agencies or organizations (ie. the Toledo-Lucas County Library system) that are facing major budget cuts because of cuts to state aid. Whenever they run a story like that, it’s not explicitly anti-Strickland, but it is a story along the lines of, “Due to the cuts proposed by Governor Strickland, X could be facing a budget shortfall of…”

    So I think a lot of people could just be frustrated by cuts in services and the situation in general, and Strickland’s poll numbers are suffering because of it. It’s not that people don’t like Strickland, it’s that they want to blame someone for a bad situation, and his name comes up a lot. He’s suffering from the same problem every other governor in the country is facing.

    I think in a campaign setting, with the governor actively making his case and reminding people that, hey, he’s actually done a decent job in the midst of tough times, his support will stabilize.

    At least, I hope so. If a popular and personable governor like Strickland can’t get re-elected in Ohio, then John Cherry doesn’t have a chance in Michigan.

  2. Haven’t seen June numbers for Ohio yet but May unemployment rate for Ohio was 10.4%.

    An incumbent holding around a six-point lead with double-digit jobless numbers isn’t bad at all.

    The rule of thumb that any incumbent below 50 is endangered does kick in, but, since his personal favorability is not yet upside down, Strickland has the potential to soar if the economy picks up by Nov 2010.

    And, if it doesn’t and Kassich doesn’t shoot himself in the foot, well, we will probably see the undecideds break 3-1 for the challenger.

    Under the circumstances, I think the leaner call is dead solid perfect. The economic numbers should have Strickland behind; the fact he retains a lead is a testament to the strength of his core support, yet calling likely D would be a stretch as long as he is below the magic number.

  3. Just looked at Quinni trends from January on Strickland’s personals. Yikes! Pretty steep rate of decline.

    Still, could have been worse under the circumstances. And makes retaining a lead more impressive. Love to see how his 44 compares to right/wrong track for the state.

    If anyone should be kicking himself, it should be Kassich for not being able to take a lead.

    Not that sure of the 38-30-32 DRI or whether by registration or party self-ID. Ohio has very high 55 percent undeclared. Stay that way there unless vote in a primary. Of course, they tend not to vote, primary or general, so probably not bad being light on Indies. Particularly hard to get them to do survey in off-off-year.

    And kind of hard to figure what breakdown to use when they had 1,114,740 registered voters in Cuyahoga County out of a total all-ages population of 1,283,925. Guess there are no minors in Cleveland. But the Dem numbers have been trending nicely all over the state. 50 of 88 counties now have Dem majority. And, outside the oddity of Cuyahoga, by and large Dems are gaining in counties growing in population and GOP in ones shedding people.

    Of course, this reads too much into short and medium-term trends. Every year we get a little farther away from the Taft immolation of the state GOP. Bound to see a little bounce back towards historical norms, especially when in opposition in hard times.

  4. What is the typical relationship between approval ratings, favorability ratings, and re-elect numbers?

    Iowa Governor Chet Culver’s approval numbers have ranged from the low 40s to the low 50s in recent months (SUSA found his approval at 42 percent in April, 48 percent in May and 42 percent in June, while a recent Republican poll put him at 53 percent approval).

    However, all of the polls have Culver’s re-elect numbers much lower–usually something like 35 or 36 percent say he deserves to be re-elected, while a much larger number say it’s time to give someone else a chance.

    Is this normal, or does Culver have an unusual problem?

  5. I think this could potentially stem the tide of job losses if gains could be made in cities like Columbus where major universities draw in highly educated workers who tend to favor D’s and who are less likely to lose their job in a recession.

    Those folks would seem to be more inclined to vote both for Democrats and incumbent Democrats in a down economic period because a) they are better educated and b) they aren’t upset about their employment situation.

    Just a thought. Anyone with local insight?

  6. The primary reason Strickland’s numbers have sunk is because of the budget cuts. This is happening in nearly every state.

    I do notice though, that Kasich hasn’t been able to fully take advantage of this, and I don’t expect he will. Let’s not forget that one of his last jobs was with Lehman Brothers. Plus, Kasich has the same problem as Portman: he’s too much of a Wall Street type to play well in OH at this time. In the 90s? Maybe, but definitely not now.

  7. the economy doesn’t continue to decline next year. If there aren’t at least some signs of improvement a year from now, a lot of our incumbent governors could be in trouble.

  8. that I haven’t seen any head-to-head numbers for Culver against any likely Republican challengers.

  9. As far as I know, there really isn’t a hard and fast rule on the relationship, outside of the obvious: higher is better.

    What is common, though, is for approvals to run higher than reelect. It is quite common for members of the opposite party to have favorable views of a pol, give them good job approvals, but in the end still end up voting against. A classic example is Landrieu in LA, who always gets high personals in the low 60s, rarely sees her job numbers drop below 55 yet ends up winning her general elections in squeakers.

    One way to read approval and reelect numbers is to filter them through the DRI numbers for a state or district. A pol with lower approvals in an area in which his/her party has the edge usually has a ceiling higher than the personal and job approvals.

    Now the traditional rule of thumb is any incumbent with below 50 percent job approvals is potentially vulnerable. However, this rule has sort of fallen apart in the trench warfare which has evolved since 2000. We don’t see crossover percentages of Ds voting for Rs and Rs for Ds we used to see. 15 years ago each side expected to lose a good 20-25 percent of their own registrants. Now when a candidate is losing more than 20 percent of their own it is viewed as vulnerability in competitive districts.

    As for reelect numbers, the math is simple. If you got 50, you ain’t losing. The only question is when you have 50. A year and a half out, it hardly matters what you have as long as you aren’t running so low you can’t raise money.

    Of course, numbers alone can be meaningless. In the end an election is a choice between two or more people, the herd gathering to decide who should be leader of the pack. Of late there has been so much focus on money and message we have sort of sublimated the most crucial factor: messenger. The quality of the candidate can trump everything else. Sort of the lesson of Terry Mac in VA. Forgot the 3rd M matters most.

    A good candidate can cancel out all the rules of thumb. So, in the end, there are no general rules on reelect and personal/job approvals. You just look at all three to get a gut read of how strong the candidate’s standing is with the electorate and extrapolate how the tide will affect it.

    Now there is one supposedly magic bullet: right/wrong track numbers.

    They ask something like, “Generally speaking, do you think things in Frostbite Falls are on track and moving in the right direction or off track and moving in the wrong direction?”

    It is an amazingly vague question open to multiple interpretations. What does it actually mean? Who knows? It means whatever it means to the respondent at the moment they hear it.

    But, whenever the right track stays above 55 percent, incumbents almost always win (unless caught in some scandal). When it drops around 45, incumbents are vulnerable. When it hits 40 or below, expect challengers to win. Around 50, only exceptionally talented challengers make a race of it.

    The key, though, is to right/wrong track for the city, county, state, district and not for the country overall. With rare exceptions (like our internal versions of a failed state… like Detroit or Cali right now), folks rate right track higher for their local area than the country as a whole.

    I always wondered why the heck private pollsters wasted time asking such a meaningless question. As I learned how it functions as a guidepost for them, I tipped my hat to the genius son-of-a-gun who figured this one out.

    As you watch Culver’s numbers, look to see if his personal and/or job approvals stay above the right/wrong track numbers. If they do, it indicates he isn’t getting blamed for the troubles in the state… at least not enough to be decisive. If the wrong track exceeds his reelect or approvals outside the margin, he is in deep trouble.

    Somehow (and not having seen any Iowa numbers with right/wrong track since the caucus), my bet is the right track numbers are well below 40. It is just not a good time to be an incumbent.

    How the economy breaks over the next year will be crucial. Folks don’t seem to process recoveries without a 3-6 month lag. Look at Bush the Elder in 1992, Nixon in 1960, Democrats in 1994. GDP can be clicking up at a nice rate and folks will not notice it for a spell. If the public perception hasn’t embraced the notion the economy is recovering by late summer of 2008, 2010 is going to be a bad year to hold office.

    Got a year and a half. That is an eternity in politics. My only concern is that political years are dog years in economics.

  10. This was always going to be a painful recession, and the White House lowballed the stimulus.  

  11. tend to be higher than re-elects is they often have two possible answers:

    Approve or disapprove

    Favorable or unfavorable

    Re-elect questions often have three options:

    Would vote to re-elect

    Would consider someone else

    Would vote to replace

  12. kind of forgot about NJ and VA being up this year.

    getting late to avoid economy dragging incumbents this year.  

  13. Most private pollsters ask favorability with four read options and a volunteer.

    “Is your opinion of Joe blow very favorable, somewhat favorable, somewhat unfavorable or very unfavorable/” with a precoded volunteer response for the mixed/neutrals.

    Also, if the respondent can’t rate, will ask “do you recognize the name?”

    Hence, on a favorability question you will have seven response codes besides the DK/refusal code.

    1) very favorable

    2) somehwat vaorable

    3) (VOL) neutral. little of both

    4) somewhat unfavorable

    5) very unfavorable

    6) rec name but cant rate

    7) does not recognize


    Similar for approval ratings but not rec name follow-up.

    College pollsters and a good number of media polls will just ask absolute (approve/ disapprove) or present as absolute then do a follow-up for degree.

    Private pollsters do not like the lack of nuance from asking as absolute. Miss a lot when you don’t see the degree breakdown.

    Also, regarding follow-up to establish degree instead of presenting intially as sliding scale skews numbers to extremes. Once someone has been forced to pick sides, they are psychically vested in their choice. When asked follow-up for degree, you will inflate your extremes. Can generate a ten point skew towards extreme versus moderate option.

    As for generic reelect questions, your version is one way to ask it. And it is a good one for measuring general vulnerability. What it doesn’t do, though, is measure the opposing party’s offsetting vulnerability.

    Another way is just matching the incumbent with a generic of the opposite party. Asked just like other vote reads.

    “If the election for dogcatcher were held today, would you support incumbent Demopcrat Joe Blow the Democrat or the Republican candidate?”

    More than one way to skin a cat.


  14. has a bunch of links to census data and charts and all. pouring over them, I reckon you could draw this conclusion looking at the growth counties and extrapolating on the jobs reports.

    job growth in areas with government and non-profit secotr jobs, high-end service.

    job losses concetrated in heavy industry and retail.

    Cuyahoga being the exception, of course.

Comments are closed.