Sabatos Crystal Ball

Ratings Changes: House and Governors

Affluent suburban seats looking dicier for GOP, but their numbers in the House are not all bad; Colorado, Michigan gubernatorial races shift to Democrats

Kyle Kondik, Managing Editor, Sabato's Crystal Ball September 20th, 2018



— A slew of new House polling, mostly from the New York Times and Siena College, contains bright spots for both parties but also suggests a Democratic edge in the race for the House.

— We have seven House ratings changes, all in favor of Democrats.

— We also have two gubernatorial ratings changes, also in favor of Democrats.

Table 1: Crystal Ball House ratings changes

Member/District Old Rating New Rating
AZ-9 Open (Sinema, D) Likely Democratic Safe Democratic
Mike Coffman (R, CO-6) Toss-up Leans Democratic
FL-15 Open (Ross, R) Likely Republican Leans Republican
Erik Paulsen (R, MN-3) Toss-up Leans Democratic
George Holding (R, NC-2) Likely Republican Leans Republican
NM-2 Open (Pearce, R) Leans Republican Toss-up
Chris Collins (R, NY-27) Likely Republican Leans Republican

Table 2: Crystal Ball gubernatorial ratings changes

Governor Old Rating New Rating
CO Open (Hickenlooper, D) Toss-up Leans Democratic
MI Open (Snyder, R) Toss-up Likely Democratic

By now, those who watch the House of Representatives are aware of the New York Timesmassive House polling project in conjunction with Siena College. They are in the field daily and will release dozens of House polls from now through the November election. They are providing House analysts (and the general public) with a treasure trove of nonpartisan data about the most competitive House races, many of which might not even have received a single nonpartisan survey were it not for this project. So this House analyst is grateful for the data they are providing. And while operatives on both sides complain about certain aspects of the polling — particularly that NYT/Siena is continually updating the progress of the polling, providing partial results that can be misleading even though the site features an admirable array of cautions and caveats — we have heard from several pros who are hitting “refresh” on their web browsers every night to keep track.

Better to be a polling junkie than another kind of addict, we suppose.

Overall, the numbers are broadly suggestive of an environment where Democrats appear to be favored in the House, but there are results with which both sides can be pleased.

For Republicans, the completed polls showed state Del. Carol Miller (R) with a 48%-40% lead in WV-3 over state Sen. Richard Ojeda (D), an ancestrally Democratic open seat that has become one of the most Republican districts in the country at the presidential level, and also had former Paul Ryan aide Bryan Steil (R) up 50%-44% over ironworker and online fundraising sensation Randy Bryce (D) in WI-1, the seat Ryan is leaving behind as he heads into retirement. While both of these seats are right of center at the presidential level — overwhelmingly so in the case of WV-3 — they are vulnerable as open seats, but the GOP seems to be holding up OK in both (we call WV-3 a Toss-up still, and WI-1 Leans Republican). Other strong results for Republicans include TX-23, where Rep. Will Hurd (R) was up 51%-43%, which seemed to confirm the thought from many observers that he was well-positioned despite occupying a swing district. The GOP also got good news from a part of TX-23 on Tuesday night, when Pete Flores (R) won a traditionally Democratic state Senate seat in a special election victory over ex-Rep. Pete Gallego (D), who held TX-23 from 2013-2015. On balance, special election results since the 2016 presidential election have pointed to a positive environment for Democrats, but this particular one definitely did not.

Perhaps the best poll result for Republicans in public polls lately was not from the NYT/Siena project, but from Monmouth University, another nonpartisan pollster that has helpfully stepped up its House presence this cycle. Monmouth has former state Assemblywoman Young Kim (R) up 51%-41% based on a historical midterm turnout model over lottery winner and philanthropist Gil Cisneros (D) in CA-39, an open Southern California seat that is traditionally Republican but that Hillary Clinton carried by nearly nine points in 2016. The president’s party rarely holds such districts — open seats that the other party won in the last presidential race — but Kim’s local appeal, the district’s GOP tradition, and an aggressive advertising campaign by the Congressional Leadership Fund (the group that arguably has eclipsed the National Republican Congressional Committee as the top outside conservative spending group on the Republican side) seems to be pushing Kim ahead. We’ll see if it lasts. Also encouraging for Republicans is that KY-6, a Republican-leaning district based in Lexington that is bluer down-ballot, seems to be something of a tie based on the NYT/Siena polling. Former Marine pilot Amy McGrath (D) had apparently been leading Rep. Andy Barr (R, KY-6) there previously, but a combined Barr/CLF onslaught attacking McGrath for a variety of liberal statements may have brought Barr back from the brink.

Still, Barr is one of many Republican incumbents who is below 50% in at least some polling. If one believes the environment is Democratic-leaning and very well could become more Democratic-leaning as opposed to less as the election approaches (that’s what happened for the presidential out-party in the last three midterms, all waves), one has to wonder how GOP incumbents below 50% will increase their level of support. So that’s why NYT/Siena polls showing Republican incumbents in tied races but at only around 45% support, like Reps. Dana Rohrabacher (R, CA-48) and Peter Roskam (R, IL-6), are not good for the GOP. Both represent districts that backed Clinton after supporting Mitt Romney in 2012, President Trump’s approval is weak in both, and Democrats lead in both districts on the question of which party respondents prefer would control the House. So even though those polls are ties, their findings are probably more encouraging for Democrats than Republicans.

More alarming for Republicans is that other Clinton-district incumbents Reps. Mike Coffman (R, CO-6) and Erik Paulsen (R, MN-3) were clearly losing to their Democratic challengers (both Democrats were up by about 10 points on their Republican opponents). Even more alarming for Republicans is that the results basically make sense: One would expect Republicans in affluent, highly-educated suburban seats to be struggling in this kind of environment. Both move from Toss-up to Leans Democratic, joining another high-end suburban district incumbent, Rep. Barbara Comstock (R, VA-10), in the underdog column. A warning to Democrats: Even in a bad environment, one or more of these incumbents could still win, as several of the most vulnerable Republican incumbents did in 2006 even amidst a wave. In fact, the NRCC believes Comstock currently leads her race, but that opinion is far from universally held even on the GOP side. In any event, we give all three less than 50-50 odds at this point.

We have reiterated many times this cycle the vulnerabilities that Republicans face in open seats, and one of those where the Democrats have a decent shot is NM-2, a usually GOP seat that covers southern New Mexico. The last time this seat was open, Democrats won it, but then-former Rep. Steve Pearce (R, NM-2) came back to recapture it in the 2010 wave after he lost a 2008 Senate race. Pearce is running statewide again, this time for governor, and Democrats love their candidate, water rights attorney Xochitl Torres Small (D), who faces state Rep. Yvette Herrell (R). Trump won the district by 10, so it’s tough sledding for a Democrat, but NYT/Siena found it roughly tied (other polls have shown Herrell leading). Both party committees are engaged in the race, a tell-tale sign that both sides see a very competitive contest. We are moving NM-2 from Leans Republican to Toss-up. Another open GOP seat that appears to be a decent Democratic target is FL-15, a district that covers some of suburban Tampa and also some rural areas. Attorney Kristen Carlson (D), the Democratic nominee, recently released an internal poll showing her effectively tied with state Rep. Ross Spano (R) in another district Trump carried by 10. All caveats about internal polls aside — they often are overly rosy for the side that releases them publicly — this looks like another competitive open seat, so we are moving it from Likely Republican to Leans Republican.

While Republican open seats are going to be hard for the party to defend, some Democratic open seats in recently-competitive districts are looking like easier holds for the party on the right side of the national environment. One of those is AZ-9, a Phoenix-area seat that Rep. Kyrsten Sinema (D) is leaving behind to run for Senate. While Sinema had a challenging race in 2012, she coasted in 2014 and 2016 as the district swung toward Hillary Clinton, and former Phoenix Mayor Greg Stanton (D) seems well-positioned to hold it. We’re moving AZ-9 from Likely Democratic to Safe Democratic. For future reference, Stanton is probably someone to watch as a possible Democratic candidate for the special election that will be held for Arizona’s other Senate seat in 2020; in fact, if Sinema had not run this year he might’ve been the nominee in this cycle’s open-seat Senate race instead.

In recent weeks, internal polls (and one public poll) have shown Rep. George Holding (R, NC-2) effectively tied with former state Rep. Linda Coleman (D) in a suburban Raleigh seat that Trump won by 10 points. NC-2 moves from Likely Republican to Leans Republican, and both Holding and Rep. Ted Budd (R, NC-13) are endangered Tar Heel State Republicans in districts that might be Toss-ups at some point. A wrinkle in North Carolina is that this is what some there have called a “blue moon” election — there is no Senate or gubernatorial race on the ballot to juice turnout, which might help engaged Democrats given the intensity edge the party has demonstrated in many places nationally in the aftermath of 2016.

Finally, scandal-plagued Rep. Chris Collins (R, NY-27), who holds a Buffalo-area seat that probably is the most Republican seat in New York state, is not going to leave the ballot despite insider trading charges. It is very hard to get off the ballot in New York, and Collins may not want to leave office and forfeit the possibility of using his potential resignation as a chip to trade in as part of a future plea deal. But an indicted lawmaker on the ballot gives the GOP a headache here, just like in CA-50, where Rep. Duncan Hunter (R) also is running under indictment on corruption charges. We’re moving NY-27 from Likely Republican to Leans Republican, matching our CA-50 rating. Neither of these districts should otherwise be competitive. As of now, our sense is that Hunter’s race is closer to a Toss-up than Collins’ is.

Let’s update the big picture. Our new ratings put 208 seats in the Leans/Likely/Safe Democratic column, 199 seats in the Leans/Likely/Safe Republican column, and 28 in Toss-up. Based on those ratings, Democrats would need to win only 10 of the 28 Toss-ups to win the House. One reasonably would expect the Democrats to win more Toss-ups than Republicans given the trajectory of the election, and probably at least a couple of seats where we currently favor the Republicans. On the flip side, the Republicans very well may salvage a few seats we currently rate as Leans Democratic. As noted above, it should surprise no one if someone like Comstock or Coffman came back from the brink, and some seats that seemed like easy Democratic wins may not be. For instance, Politico’s Marc Caputo reported Wednesday that Democrats are facing some challenges in FL-27, an open South Florida seat that Clinton won by 20 points but that retiring Rep. Ileana Ros-Lehtinen (R) held for decades. One emerging House trend is that Republicans may be persevering in a number of diverse, Clinton-won districts, like CA-39, FL-26, FL-27, and TX-23, in part because of strong candidates but also because the truly fiery, anti-Trump midterm Democratic energy may be concentrated more with suburban, college-educated white voters than with nonwhite voters. If that is the case, Republicans may be able to hold the line in some surprising places, a must if they are to hang on to their majority.

Put it all together, and our best guess right now is a Democratic House gain of somewhere in the low-to-mid 30s. But there are enough very close races that something like a 30-seat gain could turn into more like a 20-seat gain, and leave the Democrats short of a majority (they need to net 23 seats to take a majority). Back in July, we said the Democrats were “soft favorites” to win the House. Their odds have likely gotten better since then, or at the very least have not gotten worse, but the GOP still has an opportunity to retain the House with some breaks.

Upgrades for Democrats in two swing state gubernatorial contests

As soon as Donald Trump won the White House, Republicans knew that holding the Michigan governor’s race would be a slog. Yes, Trump carried the Wolverine State, but the usual midterm trend against the White House combined with fatigue over eight years of total GOP control at the statewide level gave Democrats a good opportunity in the state. After the general election matchup between state Attorney General Bill Schuette (R) and former state Senate Minority Leader Gretchen Whitmer (D) was set last month, we wanted to wait and see how the dust settled. Now that it has, and Whitmer has led every poll we’ve seen (often by double-digits), and we just don’t see what changes those numbers dramatically in the GOP’s favor in the last month and a half of the campaign. We’re moving Michigan from Toss-up all the way to Likely Democratic, matching our rating in the state’s Senate race. We’ve heard from several sources that the GOP position in Michigan is poor, giving the Democrats a good chance not just to win the governor’s race but also make up ground in the state’s U.S. House delegation.

With this ratings change in Michigan, we now favor the Democrats in three of the six Midwest gubernatorial races (Illinois and Minnesota are the other two), and the other three (Iowa, Ohio, and Wisconsin) are Toss-ups. Of those three, the Republican nominee is probably only ahead in one of them, Ohio, where state Attorney General Mike DeWine (R) appears to retain a small lead over former Consumer Financial Protection Bureau Director Richard Cordray (D), although that lead is probably only a few points. The Midwest, always swingy and competitive, may break for the Democrats in 2018 after breaking for the Republicans in 2016.

The other change comes in Colorado, where wealthy Rep. Jared Polis (D, CO-2) faces state Treasurer Walker Stapleton (R). Unlike in Michigan, where there’s been a lot of public polls, there’s hardly been any in Colorado, although we have heard that Polis is leading, but perhaps only by a small amount. While Polis is an imperfect candidate — he recently snubbed a longstanding rural state political gathering for no good reason that we could tell — he has a ton of money and is running as a Democrat in a Democratic year in a state that is trending Democratic. So Colorado moves from Toss-up to Leans Democratic. If Polis wins, his victory would be the Democrats’ fourth-straight Colorado gubernatorial win.

So our ratings now show Democrats favored to net three Republican-held governorships, Illinois, Michigan, and New Mexico, while the Republicans are favored to win Alaska, currently held by an independent. There are seven remaining Toss-ups, and all but Connecticut are currently held by Republicans (and Democrats may ultimately have the edge in the Nutmeg State despite the unpopularity of outgoing Democratic Gov. Dan Malloy). So Democrats remain poised to net several governorships, although some of the biggest races — Florida, Ohio, and Wisconsin most notably — remain uncertain.

Keep on Keepin’ On: 2018 Incumbent Renomination Rates

Four incumbents lost primaries in the House, one in gubernatorial contests, and none in the Senate

Geoffrey Skelley, Associate Editor, Sabato's Crystal Ball September 20th, 2018



— As a whole, incumbents won renomination in U.S. Senate, U.S. House, and gubernatorial nomination contests at a typical rate in 2018. In all, 100% of senators, 99% of representatives, and 95% of governors who sought renomination in Democratic or Republican primaries/conventions won renomination this cycle.

— Just four members of Congress lost renomination. No senators lost renomination, but four House members fell in primaries or primary runoffs: Reps. Mike Capuano (D, MA-7), Joe Crowley (D, NY-14), Robert Pittenger (R, NC-9), and Mark Sanford (R, SC-1). Outside of the four congressional defeats, the only other incumbent to lose renomination in 2018 was Gov. Jeff Colyer (R-KS).

— Overall, in the 1974-2018 period, incumbents won renomination at the following rates: 97% for senators, 99% for representatives, and 95% for governors.

— Although few incumbents lost renomination in 2018, much turnover in the House and governorships remains possible, even likely. In all, 58 House members have retired, second only to the 1992 cycle, and including open seats via redistricting in Pennsylvania and incumbent primary losses, 64 House seats will feature no incumbent running in November. Meanwhile, 17 gubernatorial contests have no incumbent in the general election (including Kansas, where Colyer lost), nearly half of the 36 seats that are up in 2018.

Adios, primary season

With the completion of New York’s primary for state-level offices on Thursday, Sept. 13,[1] 49 states have concluded their nomination processes for the 2018 November election. Because Louisiana has a unique system whereby candidates can win outright with a majority in the state’s all-party blanket primary on Nov. 6, this means that every state with a two-step electoral process (first a primary, then a completely separate general election) has finished choosing candidates for office. This also means that any incumbent seeking another term in office in states outside of Louisiana has now successfully advanced to the general election or met defeat in a primary. As a whole, incumbents won renomination in U.S. Senate, U.S. House, and gubernatorial nomination contests at a typical rate in 2018.

Just four members of Congress lost renomination out of the 401 senators and representatives who sought another term in office in a primary outside of Louisiana. No senators lost renomination, but four House members fell in primaries or primary runoffs: Reps. Mike Capuano (D, MA-7), Joe Crowley (D, NY-14), Robert Pittenger (R, NC-9), and Mark Sanford (R, SC-1). Pittenger and Sanford’s losses truly shocked few observers. Both had faced tough primary challenges in recent times: Pittenger won his 2016 primary with only 35% and by a margin of just 0.5 points on a new congressional map, and Sanford won his 2016 primary 56%-44%. Moreover, Sanford had been critical of President Donald Trump, increasingly a no-no in GOP circles, while Pittenger faced the same opponent in 2018 whom he had narrowly defeated in his 2016 primary. Meanwhile, the Democratic losses were more surprising, particularly the one that has received the most attention: Activist Alexandria Ocasio-Cortez defeated Crowley 57%-43% in the NY-14 Democratic primary, a seismic result that turned out a potential future speaker of the House and brought in a new public face for the Democratic left. Ocasio-Cortez’s victory suggested that Capuano, another veteran, white Democrat in a diverse urban district, might be in trouble, and Boston city councilor Ayanna Pressley defeated Capuano 59%-41% in the early September primary. While the losses by Capuano and Crowley differed in some respects, they both led to a similar result: A nonwhite woman replacing a white man in a heavily Democratic urban district.[2] Given the large number of women running for office on the Democratic side in 2018, these primary losses make more sense in the larger picture of the election.

Outside of the four congressional defeats, the only other incumbent to lose renomination in 2018 was Gov. Jeff Colyer (R-KS). However, Colyer ascended to the governorship in Topeka from the lieutenant governor post when then-Gov. Sam Brownback (R) resigned in early 2018 to take a post with the Trump administration, making Colyer a “successor governor,” a historically weaker position than that of an elected incumbent. Even then, Colyer only lost renomination by a hair: His 0.1-point margin of defeat at the hands of Kansas Secretary of State Kris Kobach proved to be the narrowest primary loss for an incumbent governor in U.S. history. The 2018 cycle featured four successor incumbent governors seeking a full term in their own right, and Colyer was the only one to lose a primary, though Gov. Henry McMaster (R-SC) had to survive a primary runoff to advance to the general election.[3]

No senators lost renomination in 2018, the third consecutive cycle without such a defeat. The last Senate incumbent to lose an intraparty contest was Sen. Richard Lugar (R-IN), who lost to Indiana Treasurer Richard Mourdock in the Hoosier State’s 2012 GOP primary. That said, we did not get to find out if Sens. Jeff Flake (R-AZ) or Bob Corker (R-TN) would have won renomination had they attempted to run again; Flake in particular seemed a likely loser thanks to his criticism of the president, which probably contributed to his decision not to seek a second term. This gets at one factor to keep in mind when considering why incumbent renomination (and overall reelection) rates tend to be high: The most vulnerable pols may not seek reelection, possibly because of fears about an intraparty challenge or because of a potentially dangerous general election contest.

For all the talk of fractures in parties and political upheaval, the 2018 cycle does not stand out as particularly troublesome for incumbent renomination. In all, 100% of senators, 99% of representatives, and 95% of governors who sought renomination in Democratic or Republican primaries/conventions won renomination in 2018. Chart 1 shows the renomination rates for incumbents seeking reelection for Senate, House, or governor from 1974 to the present. The chart utilizes data from elections that had the aforementioned two-step primary-general election process (thus excluding most Louisiana elections in that period as well as a few other cases). The gubernatorial renomination rates include odd-year elections with the next regular cycle (i.e. a 2013 incumbent governor’s primary is included in the 2014 numbers). Overall, in the 1974-2018 period, incumbents won renomination at the following rates: 97% for senators, 99% for representatives, and 95% for governors.

Chart 1: Renomination rates for incumbents in the U.S. Senate, U.S. House, and governorships in presidential and midterm cycles, 1974-2018

Notes: The 1974 data include New Jersey’s 1973 gubernatorial election. The data exclude recall elections. The data exclude any incumbents who ran in blanket primary elections where a candidate could win an election outright with a majority, such as all gubernatorial elections in Louisiana for the 1974-2018 period as well as congressional elections there from 1978 to 2006 and 2012 to present. The excluded data also include some blanket House elections in Texas in the 1990s and 2000s plus three Senate races (2008 and 2018 in Mississippi, 2000 in Georgia). The data also exclude independent incumbents seeking reelection as they did not make the ballot via primaries or conventions. In three cases, incumbent senators who lost renomination in a major party primary still ran in the general election as a third-party nominee (Sen. Jacob Javits, R-NY, in 1980), an independent (Sen. Joe Lieberman, D-CT, in 2006), or as a write-in candidate (Sen. Lisa Murkowski, R-AK, in 2010).

The Senate and gubernatorial data in Chart 1 feature relatively small sample sizes from cycle to cycle, particularly for governors in presidential years because most states hold their gubernatorial contests in midterm cycles. Thus, 2018 saw a lower percentage of successful renomination bids for sitting governors than in many cycles, but that came from just one loss (Colyer) out of 19 incumbents seeking renomination (not including independent Gov. Bill Walker of Alaska). Of the 23 cycles in the 1974-2018 period, a majority had no Senate and/or gubernatorial incumbent lose renomination. The 2004 gubernatorial cycle stands out on the chart, with a 78% renomination rate based on two losses in nine attempts for incumbent governors (the 2004 total includes Mississippi’s 2003 election). In that cycle, Gov. Bob Holden (D-MO) lost the Democratic primary against Missouri Auditor Claire McCaskill, who went on to lose narrowly in the general election (McCaskill is now seeking her third term in the U.S. Senate this November). Gov. Olene Walker (R-UT) also lost renomination in Utah’s 2004 race, failing to advance to a primary out of the state Republican convention in a race eventually won by Jon Huntsman (now ambassador to Russia). Walker had succeeded to the office when Gov. Mike Leavitt (R) resigned to join President George W. Bush’s administration, but her effort to retain the governorship ended in failure.

Just three cycles — 1978, 1980, and 2010 — had renomination rates below 90% for incumbent U.S. senators. Three Senate incumbents lost in 1978 and 2010, and four met defeat in 1980. In the case of 2010, challenges from the right led to losses for incumbent Sens. Lisa Murkowski (R-AK) and Bob Bennett (R-UT) while a party-switcher — Sen. Arlen Specter (D-PA, from R) — failed to win the nomination of his new party. Murkowski famously won as a write-in candidate in Alaska’s general election that year.

Most House cycles featured few incumbent losses, with renomination rates finishing above 98% in all but four cycles in the 1974-2018 period: 1982, 1992, 2002, and 2012 (note that the 2002 rate rounded to 98.0%). All of these cycles featured redistricting in every state with at least two House seats, which sometimes threw incumbents from the same party into the same seat or placed incumbents in districts with unfamiliar territory, making it easier for an intraparty challenger to defeat them. The most tumultuous cycle was 1992, which saw about 95% of incumbents win renomination. But that cycle also had a record number of retirements (66) and a host of newly-drawn seats that sought to provide better opportunities for minorities to elect representatives, which prompted many House exits (by primary defeat or retirement).

Although few incumbents lost renomination in 2018, much turnover in the House and governorships remains possible, even likely. In all, 58 House members have retired, second only to the 1992 cycle, and including open seats via redistricting in Pennsylvania and incumbent primary losses, 64 House seats will feature no incumbent running in November. Should Democrats make sizable gains in the House — they are moderate favorites to win the chamber — that will probably spell defeat for a fair number of Republican incumbents in the general election. As for the governorships, 17 gubernatorial contests have no incumbent in the general election (including Kansas, where Colyer lost), nearly half of the 36 seats that are up in 2018. Regardless of how November shakes out, new faces will abound at future National Governors Association meetings — and in U.S. House committee meetings, for that matter.


1. New York’s primary for state-level offices usually occurs on the first Tuesday after the second Monday in September. In 2018, this would have meant holding it on Sept. 11. However, concerns about conflicts with the Jewish holiday of Rosh Hashanah and 9/11 commemorations led officials to move the primary to Thursday, Sept. 13 (Delaware and Rhode Island also shifted their 2018 primary dates due the Jewish holiday). Since 2012, New York has held its primary for federal offices in June due to a court order regarding voting access for active military service members. Since this two-primary set up came into effect, politicians in Albany have failed to agree upon a date for a consolidated primary, making New York the only state in the country with separate federal and state primaries.

2. Crowley remains on the general election ballot in NY-14 as the Working Families Party nominee but is not actively campaigning.

3. The other two successor incumbent governors to win primaries were Govs. Kay Ivey (R-AL) and Kim Reynolds (R-IA).

New Polls: Tight Senate Races Across the Sun Belt

Florida, Texas effectively tied; mixed bag for both parties in Arizona, Nevada

UVA Center for Politics September 19th, 2018


A series of new Reuters/Ipsos/University of Virginia Center for Politics polls found close races in the key Senate battlegrounds of Arizona, Florida, Nevada, and Texas.

The polls, conducted online in English from Sept. 5 to 17 with roughly 2,000 respondents per state (narrowed down to about 1,000 likely voters per state), found the following results for the Senate and gubernatorial races among likely voters in the five states surveyed:

Arizona Senate: Kyrsten Sinema (D) 47%, Martha McSally (R) 44%

Arizona Governor: Doug Ducey (R) 51%, David Garcia (D) 39%

California Senate: Dianne Feinstein (D) 44%, Kevin de León (D) 24%

California Governor: Gavin Newsom (D) 52%, John Cox (R) 40%

Florida Senate: Rick Scott (R) 46%, Bill Nelson (D) 45%

Florida Governor: Andrew Gillum (D) 50%, Ron DeSantis (R) 44%

Nevada Senate: Dean Heller (R) 46%, Jacky Rosen (D) 43%

Nevada Governor: Adam Laxalt (R) 43%, Steve Sisolak (D) 40%

Texas Senate: Beto O’Rourke (D) 47%, Ted Cruz (R) 45%

Texas Governor: Greg Abbott (R) 50%, Lupe Valdez (D) 41%

President Donald Trump’s approval rating among likely voters in each state is mixed at best, although perhaps unsurprisingly given how Democratic the state is, Trump’s approval is only very sharply negative in California:

Arizona: 45% approve-53% disapprove

California: 36%-63%

Florida: 47%-53%

Nevada: 49%-49%

Texas: 47%-53%

Full details, including toplines, tables, crosstabs, and methodological information, for all five of these polls are available at:

The finding in the Texas Senate race is a more bullish result for Rep. Beto O’Rourke (D, TX-16) than other polls. Sen. Ted Cruz (R-TX) leads by 4.5 points in the RealClearPolitics polling average, and led by nine in a Quinnipiac University poll released on Tuesday. One similarity between this poll and others is that Gov. Greg Abbott (R) is running ahead of Cruz in his own reelection bid, although other polls have shown the incumbent governor with a larger lead. The Crystal Ball rates the Texas Senate race as Leans Republican, and the gubernatorial race as Safe Republican.

On the plus side for Republicans, the finding in Nevada for Sen. Dean Heller (R) is a little bit better for him than a few other recent polls: Rep. Jacky Rosen (D, NV-3) has held nominal leads in two other surveys, but Heller is up by a small amount, three points, here. So too is state Attorney General Adam Laxalt (R) in his gubernatorial race against Clark County Commissioner Steve Sisolak (D). Both races are Crystal Ball Toss-ups.

The other state-level results reflect some trends in other surveys. In Florida, the Senate race between Gov. Rick Scott (R) and Sen. Bill Nelson (D) is about a tie while Tallahassee Mayor Andrew Gillum (D) has a lead on ex-Rep. Ron DeSantis (R, FL-6), who recently resigned his seat, in the Sunshine State gubernatorial race. The RealClearPolitics average in the Senate race is Scott leading by 1.6 points, while Gillum is up 3.7 points in the gubernatorial race, a difference between the two races that is roughly reflected in these numbers. In Arizona, Gov. Doug Ducey (R) has a double-digit lead on education professor David Garcia (D), better than his RealClearPolitics polling average, while Rep. Kyrsten Sinema (D, AZ-9) has a small lead over Rep. Martha McSally (R, AZ-2) in the Senate race. That too reflects other recent surveys, although there is some disagreement among pollsters as to who is leading in the Arizona Senate race. In California, Lt. Gov. Gavin Newsom (D) and Sen. Dianne Feinstein (D) have seemed like significant favorites in their respective races for governor and Senate, and these polls back up that belief (Feinstein is running against a fellow Democrat, de León, in the general election because of California’s top-two election system).

Reuters, Ipsos, and the UVA Center for Politics are collaborating this fall on several state-level polls. This is the first batch that has been released so far, and more releases are planned in advance of the November general election. These individual state-level polls also will help supplement the data presented on the UVA Center for Politics/Ipsos Political Atlas, a new website that uses Crystal Ball ratings, poll-based modeling, and social media data to present the state of play in this cycle’s Senate, House, and gubernatorial elections. A holistic approach is also what we at the Crystal Ball apply to polling, and we try to take many different surveys into account as we formulate our ratings.