Princeton Election Consortium

Innovations in democracy since 2004

Outcome: Biden 306 EV (D+1.2% from toss-up), Senate 50 D (D+1.0%)
Nov 3 polls: Biden 342 EV (D+5.3%), Senate 50-55 D (D+3.9%), House control D+4.6%
Moneyball states: President AZ NE-2 NV, Senate MT ME AK, Legislatures KS TX NC

Dark side of the polling moon: Indiana and North Dakota

September 12th, 2012, 11:17am by Sam Wang

Despite the fact that Senate control is on a knife edge (Politico NYT), publicly available polls are surprisingly scarce. I estimate that 6-8 seats are in genuine play, with a probability of continued Democratic control of 60%, the current likely range being 48 to 52 Democratic/Independent seats. Today, a brief review of two underpolled races – and a big reason why we know so little about them.

Six seats with probabilities in the 20-80% range are: CT, IN, MA, ND, VA, and MT (polls here). In addition, MO is likely Democratic, though Todd Akin (R) may have made up lost ground against incumbent Claire McCaskill (D). AZ was likely Republican (Jeff Flake), but Democrat/former Surgeon General Richard Carmona is making a surprisingly strong run of it, with one recent poll showing a 1-point gap.

Assuming possible correlated movement of up to +/-3 points across the board between now and November, I have a current overall outcome histogram of:

Senate projection 12 September 2012 This is rather crude because swings in single states can and do happen. A 50-50 tie is scored as being 87% likely Democratic control, following the Presidential prediction.

Two critical races for which especially little polling is available are Indiana and North Dakota, the only two states in the Union that prohibit surveys by robo-call. This leaves us grasping at other information such as political narrative, state partisan fundamentals, and campaign finance.

Indiana. RCP shows one survey in March and one at the end of July. Two more recent August surveys show Donnelly +2% and Mourdock +2%, for a median of Mourdock +1.0+/-1.1% (estimated SEM, n=4).

Mourdock is a Tea Party Republican who dislodged incumbent Senator Richard Lugar with promises of a more confrontational, take-no-prisoners attitude toward the Democrats. More recently, he has attempted to walk that back for the general election.

Of interest is the fact that Joe Donnelly, a well-liked Congressman in Indiana, has more cash on hand, as documented by data for Indiana Senate as of 12 September 2012

And Majority PAC, a 527 group, just did a $500,000 ad buy for Donnelly in late August.

North Dakota. This race fascinates. North Dakota is strongly Republican, but also has an individualist style of politics somewhat separate from the national scene. Its economy is booming, and there has been an influx of new residents. Last week Heidi Heitkamp (D) and Rick Berg (R) had their first debate, parts of which you can watch here.

Polling is positively dreary, with two polls in July giving opposite results (Rasmussen Berg +9%DFM Heitkamp +6%). Close observers think that Heitkamp has run a considerably stronger campaign than Berg, and make it a toss-up. I suspect Heitkamp’s ahead, but it’s hard to tell.

What is interesting here is that Heitkamp has a significant chance of being very outspent by Berg in the coming two months. At the end of June, Berg had more than three times as much cash on hand:OpenSecrets campaign finance Senate ND Sept 2012

And in the past, conservative groups American Crossroads and Crossroads GPS have paid for ads against Heitkamp.

A well-done survey could clarify the attractiveness of either of these races to donors. A typical cost for a live-operator survey of 700 likely voters is in the range of $15,000, or about $20 an interview. But who to pay for that, hmmm.

Donate. Joe Donnelly and Heidi Heitkamp are featured on the ActBlue site at left. For Republicans, there is Crossroads GPS.

I thank Ed Freeland and Ryan Tully at the Princeton Survey Research Center for information on polling costs and laws.

Tags: 2012 Election · Senate

13 Comments so far ↓

  • Olav Grinde

    I am puzzled by the dearth of polls on the Senate and House races. While I understand that robo-polling is illegal in some states, that can only be part of the explanation.

    I simply don’t buy the idea that Karl Rove’s American Crossroads or Majority PAC are spending tens of millions of dollars while “shooting in the blind”.

    How extensive is the state polling that is not released into the public domain? Is there any information on that?

    • Sam Wang

      If you run a multimillion-dollar operation, as those PACs do, then you have your own private polling that you do not release.

      I feel tempted to ask for donations to commission a poll, especially for ND-Sen. It would really clarify further whether readers ought to put money into that race. Donations are climbing high enough that it might be a cost-effective approach.

  • Olav Grinde

    Then, in the absence of polls, perhaps the best indication is obtained by studying the resource allocation of SuperPACs? Which is precisely the sort of thing you refer to above…

  • Spencer Huebner

    Despite being strongly Republican, the ND Senate seat in question have been held by a Democrat since 1960. That said, it has been almost as long (1964) since the a Democratic Presidential candidate won the state.

  • Zach C

    As someone who is from CT and goes to college in OH and campaigned as a fellow for the Brown this summer, I find the claim Murphy to be more likely to lose than Brown rather hard to believe. Yes Brown has more money on hand than Mandel but then the outside spending makes Brown’s chances look much worse. From the Daily Beast, “Mandel’s third-party allies have outspent and outreserved Brown’s 6 to 1, and nearly twice as much money has been spent and set aside by or for Mandel than Brown.” In 2010, McMahon spent $50m to Blumenthal’s $9m and still lost. It just seems hard to believe that McMahon would have a better showing in 2012 than 2010 and that somehow Brown is somehow safe in a swing state where is opponent has shown the ability to call upon millions in outside money. Are my concerns for Ohio simply unfounded?

  • Billy

    Sam, Montana isn’t being polled as often either. Is there a reason not to be looking much at that one, besides the fact that Tester is likely to be kicked out? That state has had two democratic senators since 2006, so the dems have some kind of history in the area.

    • Sam Wang

      Billy – compared to IN and ND, Montana is a veritable river of polling. Today there was a PPP poll showing Tester slightly ahead. Check out RealClearPolitics.

      Zach C – Indeed it is somewhat surprising to see CT so competitive. Such a close race has not been seen in OH, where Brown has a persistent, consistent lead. It might be time to readjust your intuitions. Or we can readjust polls – donate to the CT race or work for your candidate there.

  • MAT

    Changing gears for a moment, I wonder if today will prove to be our elusive Black Swan.

  • Tapen Sinha


    It could indeed be the Lehman moment.


  • Olav Grinde

    Tester ahead in Montana; McCaskill widening over Akin in Missouri; likewise Heinrich outdistancing Wilson in New Mexico, and Brown widening the gap over Mendel in Ohio.

    There’s a lot of blue…

  • Josh

    Having done some work on the ground in CT, I would say things look pretty grim for Murphy. While other Democratic candidates are illiciting a strong response in the state, voters seem decidedly lukewarm toward Murphy. Part of this can be blamed on his “insider” status (though he has only been a Congressman for 6 years), but most of this seems to relate to advertising. McMahon has done an excellent job of painting Murphy as an out-of-touch, overspending Democrat, and Murphy doesn’t have near the amount of resources to counter McMahon effectively. McMahon has absolutely flooded the airwaves in the state. This is certainly one to watch. Though polls had this within 3-4 percentage points a few weeks ago, I predict there has been a marked swing in McMahon’s favor. Perhaps the top-line draw of Obama will lift Murphy to victory, but right now that seems unlikely to me at least. Connecticut has an independent streak, although it has voted consistently Democratic in presidential races. Let’s not forget Joe Lieberman won reelection over a strong Democratic challenger in 2006, in the midst of the Iraq war, and that Connecticut has had a Republican governor for 20 of the past 21 years (if you consider Lowell Weicker, a turncoat Republican-to-Independent, a Republican).

  • BillSct

    The problem is Murphy has been slow to get his act together. Went to his website last week , clicked on events, and there was nothing there. Tonight there was something. Meanwhile, McMayonaise has been out and about like crazy.

    This past weekend McMahon was at Taste of Mystic (as in Mystic Connecticut, home of Mystic Seaport.) Murphy wasn’t anywhere to be seen. The Village of Mystic is half in the Town of Stonington and half in the town of Groton. While the Mystic that is part of Stonington has a fair number of registered Republicans, Stonington, overall, is heavily democratic because the Village of Pawcatuck, on the eastern end of town, is democratic by almost 2 to 1. Murphy is known in the western and middle of the state and not well known in the southeast corner. The southeast corner can go either way and can be very tight. When Joe Courtney beat Rob Simmons it was, I think, the closest House race in the country at the time.

    Given the heavy advantage Obama has in the state, for McMahon to win means a lot of people will have to split their vote. It should be possible to get those people to vote straight Democrat, but Murphy is now going to have to do some serious work to make that happen

  • Matt McIrvin

    It will be very, very interesting to see what happens to the presidential polling once it gets into the post-Wednesday period.

Leave a Comment