The first Bluegrass poll on the 2014 Kentucky Senate race has Republican heads turning and not in the good way. Rather, instead of a Senate heavyweight cruising to reelection, Mitch McConnell is in a dogfight for another six years in Congress. More specifically, the survey finds that he trails his less well-known challenger Allison Grimes (D) 46%-42%. The only good news in the survey is that McConnell is easily beating his primary challenger, Matthew Bevin, 55%-29% among GOP primary voters.
The survey is notable because it diverges with private polls that showed the race tight but without Grimes in the lead. Bevin’s role in the primary may also be hurting McConnell. Bevin has been hitting McConnell hard for comprising with Democrats and being to establishment which has not allowed Republicans to coalesce around McConnell in a general election match-up against Grimes. Notably, Bevin does almost as well as McConnell against Grimes, trailing 43%-38%.
But before Republicans panic other factors should be considered. Democrats may crow that Grimes is outperforming Obama in the state (she certainly is) and has polled above the Democratic ceiling in the state but upon closer inspection not so much. Indeed, 2008 pre-election polls showed a tight race between McConnell and his opponent (McConnell ended by winning by 6%).
Historically, Kentucky has behaved much as many Southern states have. It moved to the GOP at the Presidential level in 1972 and minus Clinton has stayed there. But unlike other Southern states that have seen the Democratic state parties atrophy the Kentucky Democrats party is strong and they control most statewide constitutional offices and the senate (GOP holds the House). This has allowed Democratic candidates running for federal offices, mostly unsuccessfully, to have a base of support to build their party campaigns on.
Grimes benefits from this factor. But it also puts to rest the idea that Grimes is over-performing in the poll. Consider the two most recent Senate races in the state, 2008 and 2010. In 2008 McConnell’s opponent garnered 47% of the vote. According to exit polls Lunsford got the bulk of his support from 18-29 year olds (56%), Democrats (76%), moderates (54%), liberals (80%) and African-Americans (86%). He struggled among white men (40%), white women (44%) and 45 and older (44%). Contrast this with the crosstabs of the Bluegrass survey. In 2010 the Democratic candidate hit 44% and won all the above groups (though by smaller margins).
In the Bluegrass survey Grimes gets 68% of conservatives, 54% of moderates and 79% of liberals. She also wins young voters and a significant 16% of Republicans. McConnell by contrast wins 20% of Democrats but in the survey they constitute a larger share of voters than Republicans. A quick comparison of the numbers show that Grimes is performing almost exactly the same as Lunsford. Heck, regionally she is almost matching Lunsford percentagewise. Lunsford won the Louisville area (Jefferson County) with 56% (exit polls pegged it at 54%) and Grimes is getting 54%. She wins Western Kentucky with 44% as Lunsford did in 08 and also leads in the North Central region with 48%. McConnell is over-performing in the East (likely due to the Democratic War on Coal) with 51% compared to his 2008 46% showing. One can see a swing comparison in voting from 08-010 here.
None of this is to say Grimes is not on target for past Democratic performances. She is EXACTLY on target and that may be the problem. Grimes is not running past prior Democratic candidates even against an opponent facing a primary challenge and who has abysmal approval ratings 32/60 and favorable ratings 27/50. Grimes may also be benefiting from being unknown (as in not McConnell). Only 26% have a favorable opinion of her and 27% unfavorable. Also, 29% are neutral (a nice way to say I know of her but not about her) and 18% have no opinion. While McConnell is better known to voters (three decades of tenure) Grimes is a mystery. Thus it is reasonable to hypothesize that some voters in this survey compared to private surveys are expressing support for Grimes simply because she is not as bad as McConnell.
That will likely change as the campaign rolls on. McConnell looks likely to cruise in his primary and he has hoards of cash his campaign is waiting to use until after the primary in May. Grimes has cash but again remains undefined in many voters eyes. While Bevin is certainly doing her a favor currently the fact she has been unable to top the state’s recent Democratic ceiling has to be worrisome. It suggests that this race is extremely fluid and it is extremely unclear how Grimes will respond on two key issues: Obamacare and coal.
In the survey a solid plurality, 49% supported repeal while only 44% supported fixing it and implementing it as is. Of the 49% that support repeal 76% were conservative and all age groups above 18-29% support repeal. Notably 25% of blacks and 17% of liberals supported repeal suggesting tying Grimes to Obamacare and see her grapple to respond could yield dividends to the McConnell camp among non-traditional GOP voters. While the survey did not test attitudes about coal it is notable McConnell leads in the coal heavy east (as mentioned above he lost there in 08). Lumping Grimes with the “War on Coal” is sure to also have benefits for the campaign in that region.
This analysis is not to say Grimes is in bad shape. In fact, for a Democratic running for federal office in the state she is in superb shape. But to say she is out-performing expectations is a stretch. Rather, voters current preferences seem to be based on dissatisfaction with McConnell and less to do with the primary drivers of federal elections; presidential approval, ideology and challenger qualities. After the primary ends in May and we near November expect these things to loom larger and likely benefit McConnell.
Addendum 1: Over at RCP Sean Trende has an earlier analysis of the Kentucky Senate race (see it here). He has also noted that Democrats may face a steep drop-off problem in 2014. If so, that makes the survey’s Democratic friendly sample (not necessarily on purpose) very unlikely to show up on election day.
Addendum 2: Tied to turnout (above) the McConnell campaign has over-performed on turnout since 1996. As such it is not a done deal this electorate will show up even if Democrats do not suffer a drop-off problem and McConnell is deeply unpopular.