E.J. Dionne in his latest column nicely depicts the sort of asymmetrical political warfare that is playing out in red states like Kentucky, where Democrats must refuse total identification with Barack Obama even as they score Republicans for their bad policies and obstructionist tactics:
The result will be an imbalanced argument. McConnell and other Republicans will go hard against Obama. Their Democratic opponents will run bank-shot campaigns, far less in support of the president than in opposition to the obstruction created by relentless Republican partisanship.
The realities that make this Democratic approach necessary are starkly illustrated by an important Alan Abramowitz article at Larry Sabato’s Crystal Ball, showing that the once-abundant share of the electorate willing to vote for a Senate candidate of the same party as a president whose performance they don’t like has steadily dropped since 1980:
[O]ver this time period, voting decisions in these contests have become increasingly influenced by opinions of the incumbent president’s performance. This relationship set a new record in 2012. Ninety percent of voters who approved of President Obama’s job performance voted for a Democratic Senate candidate while 82% of voters who disapproved of the president’s performance voted for a Republican Senate candidate.
This trend portends problems for Democratic candidates in Red states like Georgia and Kentucky. Recent polls put Obama’s approval rating at 44% in Georgia and 34% in Kentucky. Moreover, in midterm elections like 2014, voters who disapprove of the president’s performance tend to turn out at a higher rate than those who approve of his performance.
Now just because the trend is towards a functional “referendum” on the president’s performance in Senate midterms doesn’t mean it’s reached some sort of all-consuming omega point, and there can obviously be circumstances where negative feelings towards non-presidential-party Senate candidates matter more than negative feelings towards the president. It just means the odds of success for candidates like Allison Lundergan Grimes and Michelle Nunn are lower than they would have been in the relatively recent past. Add in the racial dynamics that especially matter in the Deep South–as discussed by Nate Cohn at The Upshot yesterday–and you’ve got what he accurately calls “a narrow path” to victory for Democrats:
In the racially polarized South, where white voters have been trending Republican for more than a generation, the Democratic route to 50 percent is mainly a matter of racial demographics. Democrats must wait for more nonwhite voters to overcome their disadvantage with white voters.
That wait might end soon in Georgia, but not in this November’s election. In the midterm balloting, the share of whites will be around 64 percent of registered voters, down from 72 percent in 2002, when the Democratic senator Max Cleland lost re-election by 7 points. Ms. Nunn will need nearly 30 percent of white voters to prevail. If Mr. Cleland were running today, his 30 or 31 percent of white voters would probably be enough to squeak out a win.
But most Democrats running for federal office in Georgia fall well short of that 30 percent. The next-highest tally was Jim Martin’s 26 percent in 2008, when he lost a close Senate race to Saxby Chambliss, a first-term incumbent Republican running in a disastrous year for the G.O.P.
Now the “narrow path” for both Nunn and Grimes could be widened and smoothed by a rise in the president’s approval ratings, by Republican mistakes, and less visibly by the DSCC’s well-financed efforts to change midterm turnout patterns and thus refute the kind of calculations Cohn is making. But Democrats would be foolish to dismiss the tough terrain, just as Republicans would be foolish to imagine a Senate takeover in 2014 won’t be exceptionally vulnerable to reversal just two years hence.