If it's in the News, it's in our Polls. Public opinion polling since 2003.


14 From ’14: Quick Takes on the Midterm

A Commentary by Kyle Kondik and Geoffrey Skelley

After going over the results from last week, we had a number of bite-sized observations to offer — 14, to be exact:

1. The polls really were worse than usual

This cycle featured the largest average miss by the two major poll aggregators, RealClearPolitics and HuffPost Pollster, in recent competitive Senate races. This isn’t a slight toward them — after all, they simply use the data that’s available, and it seems the data may be getting worse. While the median miss has been somewhat up and down, the average has increased relatively consistently cycle-to-cycle. Why? Prior to this cycle, neither average had missed a race by double-digits, but this time at least one average missed the Arkansas, Kansas, and Virginia races by at least 10 points. Below you’ll find the median and average miss per election cycle from 2006-2014 for both major poll averages.

Table 1: Median and average miss per election cycle for RealClearPolitics and HuffPost Pollster poll averages, 2006-2014

Chart 1: Median and average miss per election cycle for RealClearPolitics and HuffPost Pollster poll averages, 2006-2014

Notes: 2014 Senate data based on margins as of Wednesday, excludes yet-to-be-determined Louisiana contest. Races included in the analysis are all contests that the Crystal Ball rated as Likely Democratic/Republican, Leans Democratic/Republican, or Toss-up just prior to Election Day.

2. Republicans still ride roughshod in Appalachia

In 2012, we noted the remarkable shift in Appalachia toward Republicans. In 1976, Jimmy Carter won 68% of Appalachian counties (as defined by the Appalachian Regional Commission). Bill Clinton remained competitive here, winning 47% of them, but Barack Obama won 13% of those counties in 2008 and just 8% in 2012. So how did Democrats do without Obama on the ballot? Not much better. In 2014 statewide races, Democrats won 45 of 428 Appalachian counties in 13 states. Although Obama won 33 in 2012, most of the Democrats’ very slight gains came from a blowout win in the Pennsylvania governor’s contest and from Alison Lundergan Grimes’ (D) loss in the Kentucky Senate race, where she ran seven points ahead of Obama in the Bluegrass State’s part of Appalachia (33.8% versus 26.8%) and won five more counties than the president.

Table 2: Democratic performance in Appalachian counties, 2014 versus 2012

Notes: *Virginia counties include independent cities in the Appalachian part of the state. In states with multiple statewide elections — Alabama, Georgia, South Carolina, Tennessee — the more competitive contest is used.

3. With some exceptions, Democrats remain dominant in urban areas

In the nation’s 50 largest counties by population, Democrats won 34 of the 47 that had statewide elections this cycle. President Obama won 46 of the 50 in 2012, meaning Republicans did make some inroads, though many gains came in huge blowout races where Democrats essentially abandoned a major statewide race (e.g., governor’s contests in Ohio and Nevada). Though it mattered little for the margin in the California gubernatorial race, Republicans can take heart that they won three of the swing counties in the state that Obama won in 2012 — Fresno, Riverside, and San Bernardino.

Table 3: Election results in the 50 largest counties by population, 2014 versus 2012

Notes: *In counties in states with multiple statewide elections — on this list, counties in Georgia, Hawaii, Illinois, Massachusetts, Michigan, Minnesota, Texas — the more competitive contest is used. “G” signifies a gubernatorial result, “S” a Senate result. To download an Excel spreadsheet of this data, click here.

Sources: Population estimates — U.S. Census Bureau; 2014 results — various state and county election boards as of Wednesday afternoon; 2012 results — Dave Leip’s election atlas.

4. Who dominates the governors’ mansions?

If the victors in the Nov. 4 gubernatorial elections all serve out their terms, what will the past 60 years (from the 1958 elections to 2018) of party control for governorships look like in the midterm states? Excluding New Hampshire and Vermont because of their two-year terms, here’s the state of control through the next midterm cycle in the other 34 states that just held gubernatorial contests:

Table 4: Governorship control by party since 1958 projected forward to 2018

Notes: D — Democrat, R — Republican, O — Other. Assumes independent Bill Walker will win uncalled Alaska gubernatorial race. Nonstandard service periods for certain reasons (e.g. appointments, resignations, etc.) are rounded as best as possible.

Source: CQ Press Guide to U.S. Elections, Crystal Ball research.

Overall, Democrats will have controlled 19 states for a longer period of time, Republicans 14, with Kansas tied 30 years apiece. Democrats’ edge can be chalked up to a couple of factors. First, over the past 56 years, Republicans have held the White House for 30 years; the party that is out of the White House usually has better luck in races for governorships. (As we saw again in 2014, many contests are decided in midterm elections, which tend to go poorly for the presidential party.) Second, Democrats once held control over Southern governorships by default, and now the reverse is true with Republicans dominating much of the South. Of course, the period we covered reaches back far enough that Democrats have been in charge in Alabama, Arkansas, Georgia, Oklahoma, and Tennessee for longer.

Note that gubernatorial control does not always correspond to a state’s federal political leanings.

5. Virginia Senate race’s meager turnout bucks recent trend

There has been ample talk about the low turnout in the Virginia Senate contest between Sen. Mark Warner (D) and former RNC chair Ed Gillespie (R). At the moment, it appears that the Warner-Gillespie race will be the first two-party competitive Virginia midterm Senate race since 1982 to have lower overall turnout than the off-year gubernatorial election the prior year. In both 1994 and 2006, which saw highly competitive races for Senate in the Old Dominion, the turnout in the federal contests greatly outpaced the total vote count from the governor’s race the year before.

Table 5: Comparing Virginia election turnout in gubernatorial and midterm Senate elections

Notes: *2014 totals are unofficial. In 1990 and 2002, then-Sen. John Warner (R) had no Democratic opponent, so those contests are not included in this table.

Source: Virginia Dept. of Elections, Crystal Ball research.

The drop in turnout is reminiscent of an older pattern in Virginia, as both the 1978 and 1982 Senate elections featured slight turnout drops from the previous gubernatorial election contest. What’s interesting is, without exception, every one of the midterm Senate races in question finished with a closer margin than the gubernatorial contest the year before.

How did the race wind up being so close? As noted by Philip Bump of the Washington Post, from 2012 to 2014 turnout in counties that Warner won dropped about twice as much as in Gillespie counties. The Virginia Public Access Project also found that turnout in the most Democratic congressional district, the black-majority VA-03, was the lowest of any of the state’s 11 districts, with obvious effects on Warner’s vote total. Hampton Roads saw a nearly 1% drop in its share of the total vote in 2014 compared to 2012. Also key to Gillespie’s performance: He did better than Romney throughout the Urban Crescent, where at least 70% of the vote resides in statewide elections. Gillespie outpaced Romney in the two-party vote in all three Crescent parts — Northern Virginia, greater Richmond, and Hampton Roads.

Because the race didn’t appear to be close, it may be that Democratic-leaning voters stayed home, believing Warner was in good shape, while Republican-leaning voters were more enthusiastic about participating this cycle. Another possibility: Democrats were more fired up to vote against Republican Ken Cuccinelli in last year’s high-profile gubernatorial race than they were to vote against Gillespie, who wisely did not mimic Cuccinelli’s stridently conservative profile.

6. Outside groups didn’t buy the election

Crystal Ball Senior Columnist Alan Abramowitz ran a regression analysis to see what effect outside spending had on the Senate races. The correlation between the Democratic and Republican outside spending difference and the Democratic margin was .23, which is not statistically significant. In contrast, the correlation between the Democratic margin and incumbency status was a more significant .76, and the correlation between the Democratic Senate margin in 2014 and the Democratic presidential vote margin in 2012 was an even more significant .89.

In other words, partisanship in a polarized era, represented by the ’12 presidential vote margin, was by far the strongest predictor of 2014’s Senate vote. Naturally, incumbency status is also significant. But the difference between amounts of outside spending by groups affiliated with both parties has surprisingly little effect, perhaps because both sides spent so much that the money from Republicans neutralized the cash from Democrats, and vice versa.

Chart 2: Regression analysis of outside spending disparity vs. Democratic performance in 2014 Senate races

7. Best 2014 candidates, blowout category

Missing in the acclaim for successful GOP Senate candidates like Rep. Cory Gardner (R-CO) and state Sen. Joni Ernst (R-IA) is a deserved salute to two blowout Republican Senate winners, Reps. Shelley Moore Capito (R-WV) and Steve Daines (R-MT), both of whom crushed their Democratic opponents in easy victories. National Republicans did not have to worry much about victories in a seat they had never won in the era of popularly elected senators (Montana) or in a state where they had not elected a senator since the 1950s (West Virginia), and that is a testament to both candidates. Without question, though, both were helped by the national environment and, in the case of Daines, from the implosion of an opponent (the withdrawal of appointed Democratic Sen. John Walsh after plagiarism revelations).

Also deserving a hat tip is the only Democratic member of this year’s crop of new senators, Rep. Gary Peters (D-MI). Peters benefited from an inept opponent in former Michigan Secretary of State Terri Lynn Land (R), and from the Wolverine State’s blue hue, but this race could have been a lot more competitive, given the environment. Peters had a bit of a shaky start but quickly righted his ship with an early-year staffing shakeup, and he won by double digits.

8. Worst candidates, Maryland category

There are plenty of bad candidates every year, but rarely do two of the worst candidates of the cycle not only run for the same governorship but are also members of the same party: Maryland Attorney General Doug Gansler and Lt. Gov. Anthony Brown. Gansler became a laughing stock in the primary season for several mistakes, such as a nonchalant approach to underage drinking and ordering state troopers to use their sirens to get him to routine appointments. After Brown won the primary, he kicked away what should have been a slam dunk general election (with the help of Gov. Martin O’Malley’s unpopularity), becoming just the third Democrat to lose an Old Line State gubernatorial election since 1966.

9. Number of crossover House districts slightly increases

After the 2012 election, just 26 members of the U.S. House represented districts won by the other party’s presidential nominee: Nine Democrats won districts carried by Mitt Romney, and 17 Republicans won districts carried by President Obama. That was the lowest number of crossover districts in any presidential election since 1920, which is a good indicator of how polarized the nation has become.

The number of crossover districts went up slightly on Nov. 4, to at least 31, though it’s hard to argue that the House is really any less polarized. Rather, it’s just the result of a good night for Republicans.

There will now be at least 26 Republicans in Obama seats, and probably just five Democrats in Romney seats. Many of these representatives will be on the parties’ target lists for 2016, and they are listed in Table 6:

Table 6: House members in districts won by the other party’s presidential nominee in 2012

Note: *Indicates a newly elected member

This number could grow as a result of as-yet-uncalled races: Rep. Ron Barber (D, AZ-2), if he holds on, would be the sixth Democrat in a Romney seat (though he is trailing Republican Martha McSally in a race that is probably headed for a recount), while GOP challengers Doug Ose (CA-7) and Johnny Tacherra (CA-16) could add to the Republicans’ crossover total.

Clearly, Republicans are now overextended in the House, and some of the seats they won this year — IA-1, NV-4, NY-24, to name a few — will be very hard to hold in the long run. But with Democrats needing to flip about 30 seats to take control of the chamber, the Republicans have built themselves a comfortable buffer.

10. Republicans win their best targets

The National Republican Congressional Committee came into this election with a focus on the “Red Zone,” seven Democratic-held seats that had voted for the last three Republican presidential nominees. They succeeded in capturing four of them: open seats in NC-7 and UT-4, and the seats held by defeated Reps. John Barrow (D, GA-12) and Nick Rahall (D, WV-3). Going by 2012 presidential results, these are the four most Republican congressional districts held by the Democrats, and they are all flipping. Of the other three, the aforementioned Barber (D, AZ-2) will probably lose, while Reps. Ann Kirkpatrick (D, AZ-1) and Collin Peterson (D, MN-7) won.

These four confirmed Red Zone victories are especially important to Republicans because, unlike their victories in Obama districts in 2014, these are seats they should not have much trouble holding for the foreseeable future.

Peterson, whose district was won by Romney 54%-44% in 2012, now has the most Republican seat held by any Democrat. As of this writing, incoming Rep. Robert Dold (R, IL-10) — who first won in 2010, lost in 2012, and then won again this year — will hold the most Democratic seat controlled by any Republican (58%-41% Obama in 2012). However, if Tacherra completes his shocking upset against Rep. Jim Costa (D, CA-16), he will have the honor (CA-16 voted 59% to 39% for Obama in 2012).

In the Senate, Democrats were defending seven seats in states won by Mitt Romney in 2012, and the GOP swept all seven.

11. Retiring Republicans did their party a favor

Another positive for Republicans in this year’s House elections — as if they needed any more — is that several of their members chose the perfect year to retire. A total of 24 House members retired this cycle, according to Roll Call’s Casualty List (10 Democrats and 14 Republicans). Of those 14, the GOP retained 13, losing only Rep. Gary Miller’s (R, CA-31) seat, which before Tuesday night was the most Democratic seat held by any Republican. Republicans held on to swingy seats like IA-3, MI-8, NJ-3, PA-6, and VA-10, all of which the party kept easily (each won by more than 10 points). Meanwhile, three of the 10 retiring Democrats will be replaced by Republicans in NC-7, NY-21, and UT-4.

That does not include the 11 Republican and six Democratic seats that were open because the current occupant ran for a different office. The GOP held all 11 of their seats, while Democrats lost two of theirs: IA-1, which Rep. Bruce Braley (D) gave up to mount an unsuccessful bid for Senate, and ME-2, which Rep. Mike Michaud (D) left to unsuccessfully challenge Gov. Paul LePage (R).

12. Check the final returns

Vote counting is still going on in a number of states, and the results you might have seen on Election Night might be quite different from the actual, finalized total.

For instance, on Election Night it was not at all clear that Gov. John Hickenlooper (D-CO) would be reelected, and Sen. Mark Udall (D-CO) was trailing Rep. Cory Gardner (R) by mid-single digits. But as votes in the all-mail state continued to trickle in, it became clear that Hickenlooper was actually going to win by a bigger margin than Gardner (Hickenlooper now leads by 3.1 points, while Gardner is up by 2.2 points). Also, we wrote last week that Gov.-elect Larry Hogan (R-MD) had won by an eye-popping nine points. Well, now it’s a less eye-popping 4.5 points — still impressive, but different than what the initial election night results indicated.

The lesson here for election watchers: Don’t just check the results on Election Night.

13. An Alaskan oddity

On Wednesday, the AP called the Alaska Senate race for Republican Dan Sullivan over incumbent Sen. Mark Begich (D), finalizing what seemed like the probable outcome on Election Night. In the Last Frontier’s gubernatorial contest, incumbent Gov. Sean Parnell (R) trails independent Bill Walker, though it remains uncalled. At this point, it appears that when all the results are tallied, Parnell will join Begich as a defeated incumbent. If that’s the final result, Alaska will be one of six states this year to have split results in its Senate and gubernatorial races: The others that picked different parties for Senate and governor are Colorado, Illinois, Massachusetts, Michigan, and New Mexico.

However, there is a difference between Alaska and every other state that featured both a Senate and gubernatorial election. As Michael Carey, a veteran journalist and one of the foremost experts on Alaska politics, noted to us earlier this week, Alaska looks to be the only state to throw out both an incumbent governor and incumbent senator this year, and from different parties to boot. That the quirky Last Frontier state is likely to be unique in this or any election should not surprise.

14. Incumbents roll once again

If Parnell loses, he’ll be part of just a small group: incumbent losers. Even in a good year for Republicans, the real winners of this election are the ones who seem to be the winners in every election: incumbents.

Table 7: 2014 incumbent performance in House, Senate, and gubernatorial elections

Sources: Crystal Ball research; Rutgers’ Center on the American Governor; Vital Statistics on Congress

For the purposes of this analysis, we’re leaving out the races featuring incumbents where there is not a declared winner yet (one Senate race, two gubernatorial races, and four House races), but, regardless, incumbents had another good year, outperforming their post-World War II batting averages in all three categories.

Just thought we’d leave the 2014 election behind on a high note — for our readers who are incumbent politicians, at least.

Kyle Kondik is a Political Analyst at the Center for Politics at the University of Virginia.

Geoffrey Skelley is the Associate Editor at the Center for Politics at the University of Virginia.

See Other Political Commentary by Kyle Kondik

See Other Political Commentary by Geoffrey Skelley

See Other Political Commentary

Rasmussen Reports is a media company specializing in the collection, publication and distribution of public opinion information.

We conduct public opinion polls on a variety of topics to inform our audience on events in the news and other topics of interest. To ensure editorial control and independence, we pay for the polls ourselves and generate revenue through the sale of subscriptions, sponsorships, and advertising. Nightly polling on politics, business and lifestyle topics provides the content to update the Rasmussen Reports web site many times each day. If it's in the news, it's in our polls. Additionally, the data drives a daily update newsletter and various media outlets across the country.

Some information, including the Rasmussen Reports daily Presidential Tracking Poll and commentaries are available for free to the general public. Subscriptions are available for $4.95 a month or 34.95 a year that provide subscribers with exclusive access to more than 20 stories per week on upcoming elections, consumer confidence, and issues that affect us all. For those who are really into the numbers, Platinum Members can review demographic crosstabs and a full history of our data.

To learn more about our methodology, click here.