Is Mitch McConnell in Trouble?

Is Mitch McConnell in Trouble?

By Sean Trende - August 12, 2013

The world of political projections was rocked by a pair of recent polls showing Senate Minority Leader Mitch McConnell narrowly trailing his Democratic challenger, Alison Lundergan Grimes. Prognosticators split on the importance of these data points. The Cook Political Report moved the race to "tossup" status, while Larry Sabato and Stu Rothenberg stuck with calling McConnell the heavy favorite.

I’ll confess that when I started writing this piece, I intended to demonstrate why Cook was correct. I’d even said as much on Twitter. But as I worked through it, I found myself unconvinced by what I was writing. Looking at what has been going on in Kentucky politics at the federal level over the course of the past decade, we can’t call this race a tossup for now. Leans Republican seems more appropriate, or perhaps even Likely Republican.

Let’s start with why I initially thought that the race was a tossup. Kentucky is an odd state politically. It’s obviously very Republican today at the presidential level. But at the state level, Democrats are very much alive and well. Democrats hold all of the statewide offices except for agriculture commissioner. They won the 2011 gubernatorial race by 20 points. Democrats hold an 10-seat edge in the state House of Representatives.

It is also one of the few states in the union where an outright majority of registered voters are Democrats; in fact, 54 percent of the population belongs to the party of Jackson (38 percent are Republican). Using various measurements of party affiliation, only Maryland has a higher share of voters who consider themselves Democrats.

Now, obviously a lot of these Democrats are functionally Republicans, just as a lot of the supposedly independent voters who dominate Massachusetts are functionally Democrats. But the fact that they still claim these preferences is of some importance. Taking these data together, Kentucky stands out as a state where the Democratic Party isn’t in a state of near-collapse, unlike nearby Tennessee, West Virginia and Arkansas.

So there’s very much a path to statewide victory for Democrats in the Bluegrass State. In fact, Republicans had extremely close shaves in the 1998 and 2004 Senate elections; McConnell himself won by a reasonably narrow margin in 2008.

There are also three other factors. First, McConnell is facing a free-spending primary challenger who is amplifying the theme that Grimes will press in the general election: that McConnell has been in Washington too long. Setting aside for now whether Matt Bevin can defeat McConnell in the primary or win a general election (we haven’t seen enough of him to know), he will tend to “de-partisanize” Grimes’ central critique of McConnell.

Second, two pollsters show McConnell in the low 40s, with high disapprovals. That is not a good place for an incumbent to be, especially when his opponent is running ahead of him.

Third, Grimes seems well positioned to benefit from what we might call the “Herseth model.” This refers to Stephanie Herseth, a bright young female candidate who nearly knocked off former Gov. Bill Janklow in an open 2002 House race in South Dakota, and later won the seat in a special election to replace Janklow. Herseth was running in a state that is more thoroughly Republican than Kentucky, and emphasized her likeability, her deep local roots, and her disagreements with the national Democratic Party. Heidi Heitkamp in neighboring North Dakota parlayed a similar campaign into a Senate seat while Barack Obama was losing the state by 20 points.

It’s too early to say whether Grimes can fit into this model, but we can say she is at least a reasonable candidate. Add in the fact that she will be well-funded and has deep connections with the Clintons, and you can see the outlines of a successful campaign.

So why did I change my mind? Three reasons. The first has to do with the polls. It is one thing to show Grimes leading McConnell with 44 percent of the vote. Very few people doubt that she can get to that point, as it’s probably close to the floor for a state Democrat in Kentucky. The trickier question is whether she can get to “50 percent-plus-one” of the votes in a race for federal office, something no Kentucky Democrat has managed to do since 1996 (despite some close shaves).

More importantly, it isn’t clear what sort of coalition Grimes can build to notch this sort of win. As I noted, Democrats have put together winning statewide tickets with some regularity, and have come close to winning Senate seats recently. For that matter, Democratic presidential candidates carrying the state is not something out of legend -- it happened as recently as 1996. But the coalition that they used to win these federal races or make them competitive has fallen apart recently. I’m not certain Grimes can put it back together, nor am I certain that she can create something new.

So let’s take a quick look at Kentucky’s geography:

I divided the state into eight broad sections, and then labeled the 10 largest cities in the state. We need some perspective on what constitutes a “large city,” however. Louisville and Lexington are the only municipalities in the state with a population in excess of 60,000, and when we get down to Elizabethtown, we’re talking about 30,000 -- really more of a town than a city.

The divisions are as follows:

1) Cincinnati’s suburbs in Kentucky. Covington and Florence are really more “sleeper cities” today than they are freestanding metropolitan areas, although this wasn’t always the case.

2) Northern Bluegrass. This area is marked by gentle rolling hills and horse farms. It is mostly rural.

3) Greater Lexington. Part of the Bluegrass Region, but more urban.

4) Greater Louisville. Also in the Bluegrass Region, Louisville developed a substantial German presence in the 1800s, and was pretty Republican until recently. Oldham County, to the northeast of Louisville, is the wealthiest county in the state.

5) The Jackson Purchase. The northernmost reach of the Gulf Plain, this is culturally part of the Deep South (it tried to secede from Kentucky and join Tennessee during the Civil War).

6) The Western Kentucky Coal Fields. Less well-known than its eastern cousin, and more ancestrally Republican (Abraham Lincoln was born near here).

7) The Pennyrile: A sort of western Piedmont for the Appalachian chain. The western portion of the Pennyrile is traditionally Democratic (though not as Democratic as the Jackson Purchase), while the eastern portion, especially the southeast, is ancestrally Republican, dating back to Civil War days.

8) The Eastern Coal Fields. East of the Pottsville escarpment, you are basically in West Virginia. The northern portion of the region is overwhelmingly Democratic, dating back to the unionization campaigns of the United Mine Workers in the 1920s and 1930s, while the southern portion resisted unionization and remained heavily Republican.

So let’s start by looking at Mitch McConnell’s surprise 1984 win over incumbent Democrat Walter Huddleston (by a few thousand votes). Counties carried by McConnell are in red:

McConnell ran well in the traditional Republican base in south central Kentucky, in the Cincinnati suburbs, in the western coal fields, and in Louisville, where he served as judge/executive (then the top office in Jefferson County). Huddleston ran well in the rest of the state, especially in the Jackson Purchase and the eastern coal fields, but it was not enough.

This was an extremely durable map. If you look at Republican Sen. Marlow Cook’s (barely) successful 1968 map, it is nearly identical. The same map shows up in Democratic Gov. Paul Patton’s narrow 1995 win. Similarly, Jim Bunning’s narrow 1998 Senate win over Scotty Baesler features almost this exact combination.

By 2004, things had begun to change, though. Democratic strength in rural areas was beginning to recede, and the suburbs of Lexington and Louisville were beginning to spread out into the countryside. In his race against Bunning, Democratic nominee Dan Mongiardo managed to hold on to enough rural areas (he represented southeastern Kentucky in the state Senate) and to make enough inroads into Lexington and Louisville itself to make the race close. Some bizarre behaviors by Bunning contributed to the closeness of the election, but Mongiardo still lost:

Since then, however, rural support for Democrats at the federal level has collapsed. The following chart is a bit different from the preceding two. This shows the change from Bunning’s 2004 win to McConnell’s six-point win over businessman Bruce Lunsford. Every gradation of red marks a single point swing toward McConnell, to a maximum of 10 points. The opposite is true for gradations of blue: 

As you can see, the Jackson Purchase and the Democratic portions of the western Kentucky coal fields swung heavily toward McConnell, and McConnell performed better than Bunning in Lexington and Louisville. McConnell ran behind Bunning in the northern tier of the state, although Bunning’s margins there were probably inflated, since he represented the area in Congress. Note too that the trend in the southeast is underestimated by the fact that the gradations max out at a 10 percent swing; Perry County (the narrow, curved county in the mass of dark red eastern Kentucky counties) actually swung a net of 56 points toward McConnell.

These trends accelerated in the 2010 Senate race pitting Rand Paul against Attorney General Jack Conway: 

As you can see, there wasn’t a generalized movement toward Paul, such that we’d attribute the statewide movement entirely to the very good Republican year of 2010 (after all, 2004 was a pretty good Republican year as well). The collapse is concentrated around Lexington, in the Jackson Purchase, and in the eastern coal fields. Again, this understates how severe the collapse in eastern Kentucky has been; Perry County swung a net 66 points from Mongiardo to Paul. Note too that the libertarian Paul wasn’t exactly a “natural fit” for this region.

To be sure, Conway carried many of the counties in eastern Kentucky, as well as Jefferson County (Louisville) and Fayette County (Lexington). The problem is that isn’t enough. As we saw above, to be competitive in federal races, Democrats need to also run well in the Jackson Purchase and sweep the Democratic portions of the eastern coal fields.

Could Grimes reverse this trend? It’s possible. As I noted, Democrats have managed to pull this off in state-level races. But such a strategy is complicated by the Obama administration’s unpopularity in rural Kentucky. This is especially true of its coal regulations, which can easily be depicted as costing eastern Kentucky jobs. She can come out against these regulations, but that approach didn’t help Ben Chandler in his losing 2012 re-election campaign. It’s just too easy to tie her to the Obama administration when she runs for federal office.

Grimes could also try to create a new Democratic coalition in the state, most likely by making up for weak showings in the rural areas with strong showings in the urban and suburban ones. Democrats have deployed such a strategy in other states, including Florida, North Carolina and especially Virginia.

The problem is that Kentucky isn’t like these other states, where a viable urban Democratic coalition is replacing a traditional rural-urban Democratic coalition. The state is barely over 10 percent minority. More importantly, rural voters make up around 45 percent of the population. Of those living in “urban” areas, 40 percent live in entities with a population below 20,000.

The bottom line is that there simply aren’t enough upscale, urban voters in the state to get to a majority; rural voters are key. In fact, Barack Obama did do quite a bit better than John Kerry in urban portions of the state, especially around Louisville, Lexington, and Cincinnati. But it was a net loss in the face of a strong rural backlash, especially in the Jackson Purchase and coal country (the shift in Knott County, in the southeast, was a net 77 points):

 Third, borrowing from Harry Enten, we should appreciate just how unusual it would be for an incumbent like Mitch McConnell to lose. Since 1982, 226 incumbents have run for re-election in midterm elections. Only 26, or 12 percent, have lost. If we narrow our analysis to those incumbents who have run in midterm elections while the other party has occupied the White House, 109 have run for re-election. Only five, or 5 percent, have lost. Of those, two lost in 1998 and two lost in 2002. In both cases the opposing party’s president had a job approval of over 60 percent, vs. Barack Obama’s job approval in the mid-40s. If we narrow even further, to look only at Democrats running in states that leaned toward the Democrat in the previous presidential election, or Republicans running in states that leaned toward the Republican, only one candidate of 65 has lost.

Many of these incumbents running in midterms in ideologically favorable states have looked vulnerable early on. This is, incidentally, true of McConnell, who has almost always had tough races. Indeed, he only trailed in one poll in 2008, but he rarely had huge leads either: Mason-Dixon and Survey USA both showed effectively tied races very late in the cycle in 2008 before finding the undecideds breaking heavily McConnell’s way in the closing weeks.

This isn’t to say that Grimes can’t win. This isn’t a “safe Republican” seat, nor is it probably a “likely Republican” seat either. None of the factors I described at the beginning of the piece have disappeared. Just because the Democratic coalition in the state has weakened doesn’t mean it is impossible for Grimes to re-create it. If a Democrat were to do so, it would be against an unpopular incumbent, who is the face of an unpopular party and one of the faces of an unpopular city. In other words, McConnell. Likewise, if a Republican running in a red state in a midterm election when a Democrat occupied the White House were to lose, he would probably look a lot like McConnell as well. The point is just that when you lay the ledgers out side by side, especially after looking at what has been going on with recent federal races in Kentucky, it’s just difficult to call this a 50-50 proposition. McConnell has the edge. 

Sean Trende is senior elections analyst for RealClearPolitics. He is a co-author of the 2014 Almanac of American Politics and author of The Lost Majority. He can be reached at Follow him on Twitter @SeanTrende.

Female Senators Join GOP Circle, but Not the Inner One
Caitlin Huey-Burns · November 14, 2014
Obama Touts Economic News as He, GOP Leaders Meet
Alexis Simendinger · November 8, 2014
McConnell, Obama Size Up Their New Relationship & Goals
Caitlin Huey-Burns and Alexis Simendinger · November 6, 2014
The Moral Issue of Climate Change
David Ignatius · November 19, 2014

Sean Trende

Author Archive

Follow Real Clear Politics

Latest On Twitter