Sabato's Crystal Ball

Ratings Changes — And Non-Changes

McConnell, Grimes and meaning of our ratings

Larry J. Sabato and Kyle Kondik, U.Va. Center for Politics August 8th, 2013

The Crystal Ball has a number of rating changes in Senate, House and gubernatorial races to announce, but perhaps our most notable rating is one we haven’t changed.

A couple of recent Democratic polls show Kentucky Secretary of State Alison Lundergan Grimes (D) narrowly edging out Senate Minority Leader Mitch McConnell (R-KY) in their likely matchup in November 2014. Additionally, McConnell has drawn a potentially credible primary challenger, businessman Matt Bevin. Grimes and Bevin both reportedly impressed at Fancy Farm, the annual Bluegrass State political confab held last weekend.

With all that said, we’re keeping this race as “likely Republican.” We favor McConnell to win both his primary on May 20, 2014, and his sixth term in the fall general election. Why?

Let’s start with the primary challenge. McConnell is already running ads against Bevin, leading some to ask this question: If McConnell is already on the air, he must be really worried, right?

Not necessarily. McConnell is, if anything, an aggressive campaigner, and he has a massive war chest: His most recent fundraising report showed him with $9.6 million cash on hand. Additionally, in a world of SuperPACs, McConnell effectively will have an almost unlimited amount of money behind him, whether directly through his own spending or indirectly through third parties. So if McConnell has essentially bottomless coffers, then why not attack Bevin early and often? Wouldn’t it be political malpractice not to? In politics, a candidate sometimes is damned if he does and damned if he doesn’t. If McConnell were sitting on his heels, he could project “confidence” to observers, but he might allow the Bevin insurgency to grow in strength. If McConnell acts — as he has — he shows that he’s “worried” about Bevin, but given his resources, he can put in some early work to keep his challenger at arm’s length. The latter is a better strategy than the former, especially because Bevin is already on the air himself.

McConnell supported ex-Kentucky Secretary of State Trey Grayson in his losing effort against now-Sen. Rand Paul during the 2010 Republican Senate primary. However, McConnell and Paul are now allies, and a Paulite, Jesse Benton, is running McConnell’s campaign. Presumably, Paul’s support would be a big help to a McConnell challenger, but Kentucky’s two senators are on the same page, at least at the moment. We’ll see if some of the big conservative groups — the Club for Growth, the Senate Conservatives Fund — decide to intervene on Bevin’s behalf. But for the time being, it’s hard to see McConnell losing a primary.

In the general election, the Democratic polls — one by Public Policy Polling, the other by sharp pollster and strategist Mark Mellman — show Grimes at about 45% of the vote, with McConnell a point or two behind. That the race is close is not a surprise. McConnell’s approval numbers in Kentucky are poor, and of his five Senate victories, only one — 2002 — was a real blowout. McConnell beat Bruce Lunsford (D) in his last election (2008) 53%-47%. So getting to 45% of the vote isn’t Grimes’ challenge; rather; it’s finding those last five percentage points.

The trouble with the Democratic polls is that neither of them provides information about the composition of the race’s undecided voters. Given Kentucky’s Republican tilt at the federal level, it’s probably safe to assume that the undecideds are ideologically more conservative than liberal. Are these voters — who live in a state where Mitt Romney just won more than 60% of the vote — really going to break for a Democratic challenger when their choice could help Barack Obama’s Democrats hold onto the Senate? It’s possible, but it’s not very likely.

Our “likely Republican” rating does not suggest that the race won’t be close — it might well be, as many of McConnell’s races have been. Rather, it suggests that we still think McConnell is a clear favorite to win, even if it’s only by a percentage point or two. Despite the troublesome recent news for the minority leader, he is still in decent shape in our eyes.

Ratings changes

There are a number of other ratings we are revising, though. They are described below.

Chart 1: Gubernatorial rating changes

AR-GOV: The outstanding fundraising by former Rep. Mike Ross (D), combined with the exit of ex-Lt. Gov. Bill Halter (D) from the Democratic primary, has shown that the former congressman is a strong contender to succeed term-limited Gov. Mike Beebe (D). On the Republican side, ex-Rep. Asa Hutchinson (R) remains the likely Republican nominee, but it’s worth noting that he has lost three statewide campaigns (1986 for U.S. Senate, 1990 for state attorney general and 2006 for governor). This race is now a TOSS-UP.

However, Arkansas has become so strongly Republican in the age of Obama that we have a hard time believing that both the governor’s race and the toss-up Senate race will go Democratic. We have a much easier time seeing a split, or even that both will go to the GOP. But at this early point, it’s fair to call both races toss-ups.

ME-GOV: With the entry of Rep. Mike Michaud (D) into the race, Democrats have a strong nominee with proven appeal to challenge Gov. Paul LePage (R), who won the 2010 three-way election with less than 40% of the vote. LePage is very controversial and his poll numbers are poor. Complicating matters is independent Eliot Cutler, who came within a couple points of winning the governorship last time around.

We’re switching the rating of this race from toss-up to LEANS DEMOCRATIC/INDEPENDENT. The point of our ratings change is to make clear that we believe Mainers are growing weary of the LePage act, and usually the curtain comes down on the show one way or the other in these circumstances. As of today, we suspect that Michaud has the better chance to bump LePage out of prime time, but it is early and we’ll continue to watch.

OH-GOV: There have not necessarily been any big new developments in the Ohio gubernatorial race. Gov. John Kasich (R) did significantly outperform likely challenger and Cuyahoga County Executive Ed FitzGerald (D) in fundraising, but former Gov. Ted Strickland (D) had significantly more money than Kasich did at this time in the 2010 cycle. That didn’t prevent the challenger, Kasich, from winning.

But Kasich seems to be cruising along with decent approval ratings, and national Democrats appear to have many other, better targets for their third-party spending: Republican governors in Florida, Maine, Michigan and Pennsylvania are all, to us, clearly more vulnerable than Kasich (Gov. Nikki Haley in South Carolina might be, too). Democrats also have their own incumbents to defend in places. In other words, Ohio’s gubernatorial race might be a bit down the national totem pole this cycle, which would be a relief to Kasich. Combine that with what we sense is a lack of enthusiasm among Democrats for FitzGerald — a Greater Clevelander who is little-known statewide — and one has to install Kasich as a sturdier favorite in this race. So we’re moving this race to LIKELY REPUBLICAN, from leans Republican. For more information on the race’s dynamics, we recommend these columns from The (Cleveland) Plain Dealer’s Tom Suddes and Brent Larkin. “As underdogs go, FitzGerald’s a pretty big one,” Larkin notes. We agree, and that’s why we’re changing this rating.

Chart 2: Senate rating changes

NH-SEN: We keep waiting for a credible candidate to challenge Sen. Jeanne Shaheen (D), but no one has entered yet. State Senate Majority Leader Jeb Bradley (R), a former congressman, seemed to indicate that he was entering the race earlier this week, but then he backed off. Other potentially good Republican contenders, such as a member of the Sununu family (ex-Sen. John or state Executive Councilor Chris), are taking a pass. Even with the Granite State’s increasingly Democratic lean, Shaheen could have had a real race on her hands — and she still might. But the fact that no one of note has volunteered for the job reinforces the incumbent’s already substantial advantages and solid polling numbers. This race is now LIKELY DEMOCRATIC, which lumps it in with Colorado and Minnesota as places where potentially vulnerable Democratic incumbents seem to have a clear path to reelection, at least at the moment.

SC-SEN: No, Democrats do not have much of a chance to win the Senate race here — the general election rating remains SAFE REPUBLICAN. But we wanted to highlight some potential problems for Sen. Lindsey Graham (R), who has long drawn the ire of conservative activists. The race has gotten some national attention recently with the entry of Nancy Mace (R), the first female graduate of The Citadel, into the primary field. She joins former congressional candidate Richard Cash and state Sen. Lee Bright. Whether any of these candidates emerges as a decent challenger remains to be seen, but we’re calling the primary just “LEANS GRAHAM” for now. Remember that if Graham does not get over 50% of the vote, there will be a runoff held two weeks after the regular primary in June. The calendar does work in favor of the incumbent — recall how now-Rep. Mark Sanford (R-SC) benefited from a crowded primary field followed by a quick runoff in his special election victory earlier this year.

If there’s a shocker in the GOP Senate primaries, this is currently the top candidate. The unhappiness about Graham in Tea Party ranks is always at a low boil, and Mace’s entry could turn up the temperature considerably. Graham can still win another term, but he is going to have to throw himself into the contest wholeheartedly and spend freely — and perhaps temper his tongue and very public alliance with Sen. John McCain (R-AZ) for the foreseeable future.

Chart 3: House ratings change

WV-2: Mountain State Republicans are rightly worried about Rep. Shelley Moore Capito’s (R) seat, which runs from the state’s eastern panhandle through Charleston all the way to the Ohio border. Democrats shouldn’t really have a chance at any seat that gave 60% of its votes to Mitt Romney in 2012, but West Virginia retains a split political personality: It has supported Republicans in the last four presidential elections, but its state government is run by Democrats. Capito is in a great position to take the Senate seat of the retiring Jay Rockefeller (D), but Democrats have a strong contender for her House seat: Nick Casey, a former state party chairman who can at least partially self-fund his race. Republicans, meanwhile, don’t really have a top-notch candidate; one of the contenders, Alex Mooney, is also a former state party chairman — but in Maryland. This race could edge into toss-up territory soon if a strong Republican does not emerge — for now, we’re calling it LEANS REPUBLICAN, from likely Republican.

Also of note in West Virginia is the race in WV-3, which covers the southern part of the state. Rep. Nick Rahall (D) is one of the longest-serving members of the House, but Romney did better in his district than any other currently represented by a Democrat (except for the Utah district of Rep. Jim Matheson). Republicans had struck out on recruiting here until they flipped a Democrat, state Sen. Evan Jenkins (now-R), to run against Rahall. Rick Snuffer (R), who held Rahall to a 54%-46% win last year, is another possibility. That race continues to lean Democratic.

All in all, West Virginia — which has only three House seats — has two of the 36 House seats we rate as “leaning” or as “toss-ups.”

So what do these ratings mean, anyway?

We thought readers might be interested in a brief explanation of what our ratings mean, particularly in the context of our Kentucky rating. The ratings do not really have anything to do with how close we expect a race to be; rather, they are all about how confident we are that a candidate will win. As mentioned above, when we rate Kentucky Senate as “likely Republican,” we’re not saying that we expect Sen. McConnell to win by 55% or more, or 60% or more, or really by any number at all. We’re just saying that we have a reasonably high level of confidence that he’ll win, whether it’s by 100 votes or 1 million votes.

A recent race illustrates how this works. In the Massachusetts Senate special election, we consistently rated the race as “likely Democratic.” In the lead-up to the election, we wrote that it was “a competitive race, but not one where the outcome was all that much in doubt in the weeks leading up to the contest.” The same was true of the Wisconsin gubernatorial recall in June 2012, which we consistently rated as “likely Republican”.

So, again, a “likely Republican” rating in Kentucky says nothing about how tight the race will be. It only says that we’re reasonably confident about the likely winner at this point. If we thought there was hardly any chance of McConnell losing, we’d call the race “safe Republican.” If the outcome becomes less clear to us, we’ll consider downgrading McConnell’s chances to “leans Republican” or even to “toss-up.” Explanations of our four ratings are provided in Chart 4.

Chart 4: An explanation of Crystal Ball election ratings

Other, perhaps more prudent prognosticators do not call all the races, leaving some contests as toss-ups through the election. However, the Crystal Ball calls every single general election contest: each Senate, House and gubernatorial race, and every state in the Electoral College in presidential years. So we move all of our toss-up races into at least the “leans” category by Election Day. We think you’re entitled to our best guess, and frankly, the close ones are where the fun is in this business.

As the election season comes into better focus, we’ll be adjusting our ratings to reflect how we see the races. Readers can keep track of our ratings changes through a new feature on our website. On our ratings change page — which is linked from the main Crystal Ball site — we will keep a running tally of all our changes, updated as they happen, so that readers can see how our thinking evolves during the contests.