Skip links

The House Tilts Toward the Democrats

Editor’s Note: This is a special Tuesday edition of the Crystal Ball. We’ll be back to our regular Thursday schedule next week.


— Democrats are now a little better than 50-50 to win the House. This is the first time this cycle we’ve gone beyond 50-50 odds on a House turnover.

— We’re making 17 House ratings changes this week, all in favor of the Democrats.

— One of those comes in OH-12, where the last nationally-watched special House election is taking place in a couple of weeks.

Table 1: Crystal Ball House ratings changes

Member/District Old Rating New Rating
French Hill (R, AR-2) Likely Republican Leans Republican
Vern Buchanan (R, FL-16) Likely Republican Leans Republican
Charlie Crist (D, FL-13) Likely Democratic Safe Democratic
Steve King (R, IA-4) Safe Republican Likely Republican
David Young (R, IA-3) Leans Republican Toss-up
Peter Roskam (R, IL-6) Leans Republican Toss-up
Trey Hollingsworth (R, IN-9) Safe Republican Likely Republican
Jackie Walorski (R, IN-2) Safe Republican Likely Republican
Andy Barr (R, KY-6) Leans Republican Toss-up
Mike Bishop (R, MI-8) Leans Republican Toss-up
NM-2 Open (Pearce, R) Likely Republican Leans Republican
Steve Chabot (R, OH-1) Leans Republican Toss-up
OH-12 Special (Tiberi, R) Leans Republican Toss-up
Mike Kelly (R, PA-16) Safe Republican Likely Republican
John Carter (R, TX-31) Safe Republican Likely Republican
John Culberson (R, TX-7) Leans Republican Toss-up
WV-3 Open (Jenkins, R) Leans Republican Toss-up

Table 2: Crystal Ball House ratings

Why the Democrats are now soft favorites in the House

As soon as President Donald Trump was elected, the national political dynamics immediately changed. Democrats, somnolent in off-year elections in the Obama years (and also in 2016, at least in some key places), would re-energize. The historical burden of holding the White House transferred to the Republicans, and the president’s party has lost ground in 36 of 39 House midterms since the Civil War with an average loss of 33 seats. In the more recent past, since the end of World War II, the average seat loss is 26 seats, or right on the borderline of the 23 net seats the Democrats need to elect a House majority.

That average includes years where the presidential party broke the historical trend and netted a few seats (1998 and 2002) or lost only a relative handful (1962 and 1990). But the presidents who presided over those midterms were popular and had other factors working in their favor. This president is not popular, and no one believes there is any chance the Republicans come out of this election with more seats than they hold now. Even just a single-digit GOP seat loss would be shocking, an outcome driven by late developments unforeseeable at this juncture.

Yet, to this point, we’ve hesitated to come out and make the Democrats a favorite in the House. The Republicans have strong incumbents in some of their most vulnerable seats, and the national House map has a GOP tilt: By margin, the median House district is about four percentage points to the right of the nation based on district-level results from the 2016 presidential election. The economy is good — or, rather, baseline economic indicators are good even though there is a compelling argument that the nation’s economic gains have been unevenly distributed in recent times — and the nation is not engaged in a major foreign conflict. Recessions and war can hurt the president’s party in midterms, although their absence also does not guarantee a lack of turbulence, as any glance at national headlines these days would confirm.

So what’s changed? Why do we now tilt the House to the Democrats?

Well, part of the reason is simply this: In actuality, not much has changed throughout the cycle. That, in and of itself, is a problem for Republicans.

Election Day is getting closer, and the president’s approval rating is still largely stuck in the low 40s, a big red warning sign that has bedeviled the party of similarly-situated presidents in past midterms. The House generic ballot, which has generally been at around a Democratic lead of between six to eight points, is at the higher end of that range right now. But more importantly for the House battle, for most of this election cycle the generic ballot has shown a consistent Democratic lead that suggests a very competitive battle for the majority. A high number of open seats — the highest number of any postwar election save 1992 — give Democrats many more targets than the GOP (Republicans are defending 41 seats without an incumbent, while Democrats are defending only 22).

Special elections at the state and federal level, sometimes a helpful gauge of what is to come in the midterm, have generally shown Democrats improving on Hillary Clinton’s district-level performance, often drastically. Democrats seem very likely to improve on Clinton’s margin once again in a special election in OH-12 on Aug. 7, the last House special before the midterm, although by how much is a question (an update on OH-12, a race we now call a Toss-up, is included at the bottom of this article).

There are also the specifics of this particular election. The second-quarter (April through June) House fundraising reports came out last week, and the results are alarming for Republicans. It’s not that GOP fundraising, in total, was bad: Many vulnerable incumbents had very solid quarters. Rather, it’s that Democratic fundraising was extraordinary, with dozens of Democratic candidates turning in blockbuster quarters and outraising their GOP opponents. Money isn’t everything, but one expects incumbents to have a clear financial edge on their opponents, and it’s not clear that some current GOP members will have even that with several months of buckraking to go before the Nov. 6 election.

Put it all together, and the Democrats now look like soft favorites to win a House majority with a little more than 100 days to go. The usual caveats apply: There is time for things to change, and the Democrats capturing the majority is not a slam dunk. We recently were discussing the House map with a source who recited reams of positive indicators and data for Democrats. After hearing those, we suggested that, based on what this person was saying, the Democrats should win the House with seats to spare. The source then said it will come down to just a few seats either way. By the way, such a close outcome — a House where the majority party has 220-225 seats or even less (218 is the number required for a bare majority) — remains a distinct possibility, and the presence of so many competitive House seats in California, where the vote count takes weeks to finalize, could delay the final House outcome.

This week’s ratings changes

Ratings changes continue to push more seats into increasingly competitive categories for Republicans.

The headliners are the formerly Leans Republican districts that we’re moving to Toss-up, eight in total. We’ll discuss seven here, and the eighth (the OH-12 special) in a separate section below.

The sheer weight of the Democratic fundraising advantage is a factor in some of these moves. For instance, Reps. Steve Chabot (R, OH-1) and Mike Bishop (R, MI-8) hold districts that Trump won by about a half-dozen points apiece. They have had relatively easy elections over the past couple of cycles (Chabot has been in the House since 1995, with an interruption in service from 2009-2011, while Bishop was first elected in 2014), but they face two seemingly high-quality Democratic challengers, Hamilton County Clerk of Courts Aftab Pureval (D) and Elissa Slotkin (D), an Obama-era Defense Department official. Pureval raised more than double what Chabot raised last quarter and is approaching the long-time incumbent’s cash-on-hand total, while Slotkin has been crushing Bishop in fundraising so badly that she holds a $2.2 million to $1.7 million cash on hand advantage, an unusual edge for a challenger to hold on an incumbent. Both districts have above-average college graduation rates, often a predictor of Trump skepticism that could have down-ballot repercussions.

Two other districts also feature well-funded Democratic challengers, but this time in districts that flipped from backing Mitt Romney in 2012 to supporting Hillary Clinton in 2016: IL-6, held by Rep. Peter Roskam (R), and TX-7, held by Rep. John Culberson (R). Republicans have fretted that Culberson, who basically has never had a hard race in a Houston-area district that has long been deeply Republican (George H.W. Bush won a version of this district back when it was created in the 1960s), was not taking his race seriously enough, but party leaders believe he has gotten the message. That said, he still got more than doubled up by attorney Lizzie Pannill Fletcher (D) in fundraising in the second quarter of 2018, although Fletcher had to use resources to win a primary runoff. Roskam, meanwhile, is more battle-tested: He won his first House race in 2006 against now-Sen. Tammy Duckworth (D-IL) in what was a rare bright spot for Republicans that year (Roskam replaced prominent Republican Rep. Henry Hyde, who retired in 2006). Roskam faces clean-energy entrepreneur Sean Casten (D), who won the nomination in something of an upset. Roskam and Casten both had outstanding quarters: The incumbent outraised the challenger about $960,000 to $820,000 and retains a roughly 3.5 to 1 cash-on-hand edge.

Two other Toss-ups come in Appalachia. In the Lexington-based district held by Rep. Andy Barr (R, KY-6),[1] former Marine fighter pilot Amy McGrath (D) turned heads by upsetting Lexington Mayor Jim Gray (D) in a May primary. McGrath’s victory prompted us to hesitate moving this historically competitive district to Toss-up — Gray was more of a proven commodity — but Democrats argue McGrath is leading and Republicans concede this will be a hard race. Across the border in West Virginia, state Sen. Richard Ojeda (D) has become something of a folk hero in Coal Country and is locked in a close race with state Del. Carol Miller (R) in an open seat contest for WV-3.

Both of these districts (KY-6 and WV-3) are more Democratic than meets the eye. While Trump won KY-6 by 15 points and WV-3 by an eye-popping 49 points last cycle, at the same time Gray carried KY-6 in his unsuccessful challenge to Sen. Rand Paul (R-KY) and then-Democratic gubernatorial candidate Jim Justice (who is now the governor, and a Republican) carried WV-3 by 17 points. Certainly the trend in American politics has been toward less ticket-splitting and more partisanship, but there are still many places that make different partisan decisions up and down the ticket in a single election and many others that are willing to toggle between the parties from year to year. The Democrats’ House hopes are riding on this dynamic, because practically speaking they cannot win the House with just Clinton voters and Clinton-won districts alone.

An additional new Toss-up is IA-3, a Des Moines-based swing district, where private and public surveys from Democrats show Rep. David Young (R) either narrowly leading or narrowly trailing small business owner Cindy Axne (D), and Trump’s star seems weaker in Iowa than when he won the state by about nine points in 2016. So IA-3 also moves to Toss-up. Iowa is a place where Trump’s trade wars may have some political salience depending on what impact tariffs and tariff retaliation has on crops like soy beans, the price of which is down 20% as the president prepares to visit Iowa later this week. Farm economy disruptions have hurt the president’s party in this state and region in the past, like in the mid-to-late 1980s, when this dynamic ended up hurting Republicans. Iowa was never more Democratic than the nation than it was in the 1988 presidential election, when it strongly backed Michael Dukakis over George H.W. Bush.

As long as we’re on the topic of Iowa, let’s also note the new inclusion of Rep. Steve King (R, IA-4) in our ratings as Likely Republican. King’s district is very conservative, but he is constantly in the news for extreme comments on immigration.[2] He also has been lapped in fundraising by J.D. Scholten (D), a paralegal and former professional baseball player. Other additions to the Likely Republican list are Rep. John Carter (R, TX-31), where veteran M.J. Hegar (D) is raising a lot of money even though she faces a very uphill battle; Reps. Jackie Walorski (R, IN-2) and Trey Hollingsworth (R, IN-9), who face credible opponents in districts that are historically more competitive than Trump’s lopsided vote totals would indicate; and Rep. Mike Kelly (R, PA-16), who occupies a Western Pennsylvania district based in Erie that has some similarities to the one Rep. Conor Lamb (D, PA-18) captured in a March special election. To be clear, Likely does in fact mean Likely — Republicans ultimately should hold all of these districts, although in the event of a true wave one or more could flip. In what could be a turbulent year in the House after three relatively quiet elections (the total net change in the House from 2012-2016 was Democrats gaining a single seat), observers such as ourselves are straining to identify potential upsets, and one way of flagging those districts is to put them in the Likely Republican column, which is what we have been doing. Republican operatives believe we’re doing the Democrats a favor by listing so many GOP-held seats in competitive categories and helping the Democrats bolster their argument that the playing field is very large. They may have a point, but we also think in a year like this that many GOP incumbent vote shares will be significantly lower than what they are used to, both because of the national environment and because the Democrats are running a big roster of candidates who have at least some money and credibility. That could lead to a shocker or two or three come November if some members are caught napping.

More competitive than Likely Republican are three other districts, which move to Leans Republican in this update. Rep. Steve Pearce’s (R, NM-2) decision to once again seek statewide office (he’s running an uphill battle for governor and lost a Senate race in 2008) has opened his conservative but sometimes-competitive district, NM-2. Democrats won NM-2 when it was open in 2008, and attorney Xochitl Torres Small (D) is running a credible campaign and holds a nearly five-to-one cash-on-hand edge over her opponent, state Rep. Yvette Herrell (R). However, Herrell did have to expend funds in a competitive primary and one would still rather be the Republican in a district that is decidedly right of center. The same is true for Reps. French Hill (R, AR-2) and Vern Buchanan (R, FL-16), who have faced real races in the past and likely will again this cycle in competitive but GOP-leaning districts.

Finally, Rep. Charlie Crist (D, FL-13), the party-switching former governor who won his first term in a close race last cycle, probably could be pushed by the right Republican in the right year in his St. Petersburg-based district, but he is not a real Republican target this year and he moves to Safe Democratic. The only truly vulnerable Democratic House incumbent in November continues to be the aforementioned Lamb, who because of redistricting is running in a new district against another incumbent, Rep. Keith Rothfus (R, PA-17), although Republicans hope to push a few others by November.

The OH-12 special

The final House special election scheduled before November comes in the northern Columbus suburbs, where state Sen. Troy Balderson (R) and Franklin County (Columbus) Recorder Danny O’Connor (D) are battling to replace former Rep. Pat Tiberi (R, OH-12), who resigned his seat to take a very lucrative job with the Ohio Business Roundtable.

We previously held our rating in the OH-12 special election at Leans Republican for several reasons. This is a district that Trump won by 11 points, and it’s probably more Republican than that — some parts of it are among the most historically Republican counties in Ohio. Additionally, all the polling we have seen (public and private) has shown Balderson leading. It seemed as though Balderson was positioned to hold on and win by somewhere in the neighborhood of five points. But signs of strain are showing: GOP outside groups have long been involved in the district, although not at the heavy rates of the PA-18 special that Lamb won in March, but the Democratic Congressional Campaign Committee recently jumped in too, making a small but noteworthy $240,000 ad buy. Early voting, which often has a Democratic lean in Ohio and therefore can sometimes be overinterpreted, nonetheless is very Democratic so far (Ohio doesn’t have formal party registration, but voters can be identified by party in voting statistics based upon which party’s primary they most recently voted in). Democrats might just be front-loading their votes — again, that happens in Ohio sometimes, like in 2016 when early votes tailed first on election night suggested a close statewide race for president before Election Day votes appeared and were very Republican (this is a dynamic worth remembering when tracking the OH-12 results on Aug. 7, by the way). But, as the Columbus Dispatch pointed out, the early voting in OH-12 specifically so far is much more Democratic than it was leading up to the 2016 election.

There are two competing factors at work in OH-12: The local specifics and the national trends. For a while, it appeared as though the local specifics — a decent GOP candidate in Balderson and the GOP’s historic strength in this district — were guiding the race and contributing to Balderson’s lead. Simply put, this is not a district that a Democrat should be able to win under normal circumstances. Period.

But as the election approaches, the national trends may be asserting themselves in what are not normal circumstances: Democratic enthusiasm and special election results that have often broken in their favor. Oh, and there’s another local factor: O’Connor is a decent candidate, too.

The Republican’s PA-18 nightmare certainly has a chance to repeat itself and, if it does, the race may unfold in a similar way: O’Connor, like Lamb, will lose a lot of the district’s more rural territory, but he could be powered by disproportionately strong turnout and good percentages in the part of this district closest to Columbus (Lamb was boosted by the parts of his district closest to Pittsburgh).

So we’re calling OH-12 a Toss-up now, similar to how we felt about some other big special elections this cycle: PA-18, Alabama Senate, and GA-6. Democrats ended up winning two of those (PA-18, Alabama), while Republicans won the other (GA-6).

Conclusion: The Toss-ups

This week’s changes leave a very large number of Toss-ups: 36, 34 held by Republicans and just two held by Democrats. That leaves 200 districts at least leaning to the Republicans and 199 at least leaning to the Democrats, assuming the non-Toss-ups go the way we currently project (and there are shaky seats in both “Leans” columns).

That basically means that the party that wins about half of the Toss-ups (18 of 36 for Republicans, 19 of 36 for Democrats) will be the majority party in the House. At this point, we see the Democrats with slightly better odds to get their required share of the Toss-ups based largely on the environment, but also because they appear to have well-funded and credible challengers in these districts that can capitalize on that environment. But, as noted above, it’s not a slam dunk, and the GOP has the ability to hang on even if the big-picture national indicators (Trump’s approval and the generic ballot polling) do not get better for them. The danger for Republicans, and one thing that could put the House out of reach for them, is if those indicators get worse.


1. Rep. Andy Barr (R, KY-6) is a former student of Center for Politics Director Larry J. Sabato, and thus Sabato is not involved in ratings decisions in that race.


2. Because of an error in the Almanac of American Politics, a previous version of this story misstated Rep. Steve King’s (R, IA-4) share of the vote in 2016. He received 61% of the vote, not under 55%.