Senator Mitch McConnell is not very popular in Kentucky. He's in far worse shape than six years ago, when he won with 53 percent of the vote. His job approval ratings took a real hit this fall, although likely they have recovered modestly along with the favorability of
Republicans in general. (We don't have enough recent data for McConnell to say anything more than his job approval this winter is worse than it was last summer.)
So what has happened to other incumbents that are this unpopular? Has anybody else in this position been re-elected recently? Will his popularity recover? Follow the discussion below.
Can incumbents this unpopular win re-election?
What happens to incumbents this unpopular? The short answer: They usually retire. But for the brave who press on, they have a chance. Here are the outcomes for incumbents with net job approval ratings just prior to the election of -10 or worse in elections from 2006-2012:
The candidates with the three worst ratings didn't even make it to the general election. McConnell's current ratings are right between Blanche Lincoln and Ernie Fletcher, who both flamed out terribly. But the deeply unpopular Pat Quinn and the infamous Rod Blagojevich of Illinois both managed to hang on. These two are both Democrats in a relatively liberal state, arguably analogous to McConnell's situation.
However, there's also something else going on here: third-party candidates do very well, and having a lot of votes going to a third or fourth candidate seems to help the incumbent. This apparent relationship might just be a result of the small number of cases we have to work with, though.
So that's all well and good, but it's April now, not November. Can McConnell's ratings recover before November?
How do incumbents' net approvals change during election season?
Eight months can be an eternity in politics, but it turns out that over the last several election cycles, changes in an incumbent's job approval over that time period generally stayed within a certain range:
Net approval may rise by up to 15 points, or fall by 25 points, but 95 percent of the time it will not change more than that. At first blush, this seems like fairly useless information, but it does tell us that it is highly unlikely (but not impossible) that McConnell will see positive net approval ratings before election day. Likewise, it is highly unlikely that
Senator Collins of Maine will see her net approval drop into the negative zone.
Still, that's highly unsatisfying from the point of view of making predictions. Let's throw in another variable:
What's the effect of money on incumbents' net approval ratings?
Some incumbents face a challenger-in-name-only in the form of an Unknown Somebody. Others bear the brunt of an assault only the Koch brothers can provide.
Using the fundraising numbers available on Open Secrets, we can see that 80 percent of the time there's a relationship between the amount of money a challenger spends in the general election and the change in the Senate incumbent's net approval rating. Keep in mind the many factors that aren't taken into account in this graph, including outside spending and state population.
This makes sense. If you don't have a decent challenger, you can afford to spend what money you raise on some nice positive publicity. If you're spending millions in a brutal campaign, nobody comes out looking good.
But the exceptions here are informative. For instance, sore loser campaigns are pretty rough on the image of the incumbent—see Lisa Murkowski and Joe Lieberman. And notably, we see that Harry Reid managed to increase his net approval rating despite a tough campaign that nobody would describe as positive. Perhaps it was because his winter 2010 numbers reflected his approval numbers as the demonized face of the Democratic Party in the Senate, while his fall 2010 numbers reflected his role as senator from Nevada. This, at least, is what McConnell should be pinning his hopes on, in an attempt to pull off the same trick.
Wishy-washy possibilities
So what might happen? Let's look at some informed scenarios:
A. McConnell erases the hypothesized Senate Leadership Approval Penalty via campaigning, net approval rises to -15, a little worse than 2008, and he wins by 1-3 points.
B. McConnell gets hammered by negative advertising (and hammers back), his net approval falls to -35, and he loses by 5 to 10 points.
C. Same as C, but a Libertarian candidate also runs, siphons off some disgruntled Somewhat Conservative voters who would have reluctantly voted for Grimes, and McConnell ekes out a win a la Pat Quinn.
D. Fill in the blank. Share your idea in the comments.
Data: Most polling data are from Survey USA and PPP. Line is a Loess regression. Candidate spending information from Open Secrets.