[ad_1]
Republicans are not very present in the Senate this year – but they are doing everything they can to help the Democrats make the most of it. The GOP entered this electoral cycle with only eight of its own seats to be won. Republican incumbents withdrew in three of those seats, while Democrats have no chance in Utah – where Mitt Romney will certainly take over from Orrin Hatch – races to replace Senator Jeff Flake in Arizona and Senator Bob Corker in Tennessee is very competitive.
In addition, the early retirement of Mississippi Senator Thad Cochran this spring has sparked a special election that will add to the role of November 6 and also give Democrats a plausible chance to meet again. Just how plausible? Continue reading. We cover each of these races – with Minnesota's special election, the only democratic retreat of the cycle following the resignation of Senator Al Franken – in the number of POLLS vs. FUNDAMENTALS, the extremely dorky series of articles in which I evaluate the conflicting outlook that polls and non-polling ("fundamental") factors provide to the Senate this year. Unlike races with Democratic holders, where the inclusion of fundamentals in our forecasts has generally helped democrats, this helps republicans in states like Mississippi and Tennessee:
A big gap between polls and fundamentals
Margin of victory or defeat in open-air races and races with nominees
| The margin of victory or defeat provided by the Republicans | |||
|---|---|---|---|
| Race | Holder | fundamentals | Adjusted surveys |
| Arizona | – | -1.4 | -3.8 |
| Minnesota Special | Smith (D) | -15.8 | -8.2 |
| Mississippi Special | Hyde-Smith (R) | 13.1 | + 5.9 * |
| Tennessee | – | 13.8 | -1.1 |
| Utah | – | +26.8 | 32.5 |
But let's start with Arizona, which is a simpler case (and first in alphabetical order). Contrary to the relatively complex fundamental calculation that our model makes for races with holders, that of open-seated races is simpler: it takes into account only the partisanship, the generic ballot, the experience of the candidate, Fundraising and not a candidate is undergoing a scandal. In Arizona, Democratic candidate Kyrsten Sinema and Republican Martha McSally are both current US representatives, which means that the experience variable is a tie. And neither one nor the other is caught in a scandal. The questions, then, are who has the most money and the lean blue of the generic ballot is enough to make up for the lean red in Arizona.
And the answer is … an open air race in Arizona should be very tight in a political climate like this. Hillary Clinton lost Arizona by just 4 percentage points in 2016, but the state was more Republican than in the past, and he was more Republican in races across the state than in the federal races. (Our partisanship variables take into account the results of the state legislature as well as the presidential vote.) Sinema has a slight lead in fundraising, so the calculation of fundamentals pushes her slightly towards it. . Sinema is currently running a little over 4 percentage points in average polls. Nevertheless, this is a good example of how races tend to gravitate towards the fundamentals: Sinem had an average of 7 or 8 points before McSally won a controversial primary last month.
FiveThirtyEight treats the races featuring appointed holders (as opposed to elected the incumbents) because some of the variables we use to assess incumbents are not available for appointees and because appointees do not have a very good re-election background. For example, the special elections in Minnesota and Mississippi still have the potential to create problems for Democrats and Republicans, respectively. do not perform particularly well in surveys.
Despite this, it is probably too ambitious to think that Republicans will win the Minnesota Special Election considering that elected Democrat MP Amy Klobuchar is ready to win a landslide victory in the other Minnesota Senate race; The so-called two-barrel Senate elections (in which the two Senate seats of a state are entered in the ballot the same year) are almost always won by the same party. But while the calculation of fundamentals predicts that Smith will gain 15 points – she has raised more money and held the most elected position – Housley has kept Smith's margin to single digits in most polls. It may be that the name Housley has a small extra currency in Minnesota, obsessed with hockey (Housley's husband, Phil Housley, is a member of the Hockey Hall of Fame and the current coach of the Buffalo Sabers). Yet Smith has about 9 chances out of 10 to win – chances the Democrats are reasonably satisfied with the unpredictable race that might have occurred if Franken had remained on the ballot.
the Mississippi special election is the most complicated race on the November newsletter – and one of the most difficult to predict. If you take the polls to their true value, it sounds like an underestimated opportunity for Democrats; based on fundamentals, however, it is longer.
Here is how it works. On November 6, Mississippi will hold a first non-partisan coverage with several candidates from each party. If no one receives the majority of votes, the first two will qualify for the second round on November 27. (Basically, it's the same thing that Louisiana does every year with its Congressional races.) Run: Hyde-Smith, the Republican nominee nominated; Chris McDaniel, the controversial Republican backed by the tea party who nearly defeated Cochran in the GOP primary four years ago; and Mike Espy, an African-American democrat who was US representative in the '80s and' 90s, and then Bill Clinton's first agricultural secretary. (There is also a second Democrat, Tobey Bartee, but he has only 2% of the votes in the polls.)
The polls for the first cover, Espy and Hyde-Smith, are pretty much in the lead, with about 30% each, and McDaniel by far, with a large number of undecided individuals. None of these numbers are too surprising. President Trump supported Hyde-Smith, although she was ahead of McDaniel even before the downline. So there will probably be a second round and it is likely that Espy will face Hyde-Smith.
There are also polls on potential clashes, and in these polls, Espy is doing surprisingly well, at least in relation to how the fundamentals might expect it. On average, it outperforms Hyde-Smith by only about 6 percentage points in direct polls and tracks McDaniel in a potential flow of about 18 points. So the Lite version of our model, which predicts the second round based on polls only, gives Espy a decent kick by defeating Hyde-Smith (and supposes he would crush McDaniel instead of McDaniel). The Classic and Deluxe versions, on the other hand, which represent fundamentals, consider that Mississippi is a red state and that both Republicans are likely to get many more votes. combined as the two Democrats (Espy and Bartee) on November 6, which historically is a good predictor of runoff results. Thus, they see Espy as a heavy outsider against Hyde-Smith, and also think he could have trouble with McDaniel despite having conducted in the polls.
Each scenario in Mississippi special senatorial elections
Probability to occur, based on 50,000 simulations as at 19 September 2018
| Scenario | Lite | classical | Luxury |
|---|---|---|---|
| Espy wins the majority on November 6th | 5.4% | 5.9% | 6.2% |
| Espy beats Hyde-Smith in the second round | 17.4 | 8.9 | 9.3 |
| Espy beats McDaniel in the runoff | 0.6 | 0.4 | 0.5 |
| Hyde-Smith wins majority on November 6 | 2.7 | 2.1 | 1.8 |
| Hyde-Smith beats Espy in second round | 71.8 | 80.5 | 80.4 |
| Hyde-Smith beats McDaniel in the second round | 1.5 | 1.4 | 1.2 |
| McDaniel wins majority on November 6 | <0.1 | <0.1 | <0.1 |
| McDaniel beats Espy in the runoff | <0.1 | 0.3 | 0.3 |
| McDaniel beats Hyde-Smith in the second round | 0.5 | 0.4 | 0.3 |
Needless to say, all of this is pretty complicated. We are doing our best to model these possibilities based on race data from Louisiana, California, and Washington, which use a format similar to the first two, but this is a case in point. 39; uncertainty. The fundamentals bet would be simply that a second round, organized three weeks after polling day, in a red state like Mississippi – which could determine control of the Senate – is very likely to favor Republicans. But the runoff surveys are more equivocal.
In comparison with Mississippi, there is nothing particularly delicate about Tennessee – polls and fundamentals have totally different perspectives on the race. The polls show basically a controversy between Democrat Phil Bredesen, former governor, and Republican Republican Marsha Blackburn. (In any case, Bredesen has the slightest advance, as he tends to do better in better quality surveys). The fundamentals, on the other hand, predict a Republican victory of 14 percentage points. Tennessee is very red, having voted for Trump by 26 points. Unlike other southern states, it is quite red during statewide elections; Republicans have a 28-5 advantage in the Tennessee Senate, for example. Blackburn has also slightly beaten Bredesen so far, making it one of the only competitive Senate races where the GOP heads the fundraiser. The model gives Bredesen credit for being a former governor, but he has a lot to do against him.
Rather than analyze Tennessee more, let's look at what happened in the past when polls and fundamentals clashed. In the table below, I've listed the races in the Senate since 1990, where there were at least 10 points between polls and fundamentals, 60 days before the campaign, based on the retroactive results of the FiveThirtyEight model. . I limited the analysis to races deemed competitive by the Cook Political Report, where the number of polls was sufficient. For example, the 2016 Indiana Senate race meets all these criteria; Former Democrat Evan Bayh, a former governor and senator, was well ahead in the polls, but the calculation of fundamentals considered Republican Todd Young as a favorite. (Young finally won easily.)
What happens when surveys and fundamentals collide?
Competitive US Senate competitions since 1990 are characterized by a gap of at least 10 points between polls and fundamentals with 60 days before the election.
| Margin of the Democratic candidate 60 days before the election | ||||||
|---|---|---|---|---|---|---|
| State | Year | Survey average | Fund-amentals | Current result | Fundamentals more accurate than polls? | The race has moved in the direction of the fundamentals? |
| HE | 1990 | 19.4 | +7.1 | 30.1 | ||
| HE | 1992 | +34.2 | +12.2 | 10.2 | ✓ | ✓ |
| California | 1994 | 16.7 | 0.9 | +1.9 | ✓ | ✓ |
| New Jersey | 1994 | 18.5 | 5.2 | +3.3 | ✓ | ✓ |
| OH | 1994 | -17.9 | -7.9 | -14.2 | ✓ | |
| South Dakota | 1996 | -0.7 | -13.7 | 2.6 | ||
| California | 1998 | 0.6 | 10.7 | +10.0 | ✓ | ✓ |
| HE | 1998 | -4.7 | 10.2 | -2.9 | ✓ | |
| New York | 2000 | -0.0 | 17.0 | 12.3 | ✓ | ✓ |
| TX | 2002 | -1.8 | -16.5 | -12.0 | ✓ | ✓ |
| THE | 2004 | -20.0 | 3.9 | -21.7 | ||
| North Carolina | 2004 | +8.3 | -9.1 | -4.6 | ✓ | ✓ |
| South Dakota | 2004 | +3.2 | 15.7 | -1.2 | ||
| MARYLAND | 2006 | 6.6 | +21.3 | +10.0 | ✓ | |
| New Jersey | 2006 | -1.2 | 19.9 | +9.0 | ✓ | |
| Pennsylvania | 2006 | 8.1 | -3.4 | 17.4 | ||
| TN | 2006 | -3.3 | 12.4 | -2.7 | ✓ | |
| Virginia | 2006 | -1.4 | -18.1 | 0.4 | ||
| Washington | 2006 | 10.4 | +24.1 | 16.9 | ✓ | |
| AK | 2008 | 2.5 | -28.2 | +1.2 | ✓ | |
| Virginia | 2008 | +22.7 | 5.7 | 31.3 | ||
| AR | 2010 | -27.8 | +4.1 | -21.0 | ✓ | |
| North Carolina | 2010 | -6.4 | -17.1 | -11.8 | ✓ | ✓ |
| NV | 2010 | +1.8 | 13.1 | 5.7 | ✓ | |
| Tc | 2012 | -1.6 | +23.0 | 11.8 | ✓ | ✓ |
| South Dakota | 2014 | -13.7 | -30.8 | -20.9 | ✓ | |
| Georgia | 2016 | -13.1 | -24.3 | -13.8 | ✓ | |
| IN | 2016 | 9.9 | -6.1 | -9.7 | ✓ | ✓ |
What happened in these races? The polls moved closer to the last margin, about two-thirds of the time (in 18 out of 28 cases). So, if you had to choose between surveys and fundamentals, you would select polls. However, the race moved in the direction of the fundamentals three quarters of the time (in 21 out of 28 cases). That is, if the Republican was better at fundamentals analysis than polls, the Republican tended to gain ground 75% of the time, and so did the Democrat.
The best prediction is therefore to take a mix of surveys (essentially) and (some) fundamentals. The exact weight of our model for each component depends on the number of polls and the duration of polling day. Essentially, the calculation of fundamentals is considered equivalent to 1 or 2 recent surveys of high quality. So, if there are 10 or 15 recent polls of a state, the calculation of fundamentals has little influence. In states like North Dakota, where polls are pretty rare, they can have more influence.
What this means for Tennessee is that any poll indicating that Bredesen is tied or even behind by 1 to 2 percentage points, is good news for Bredesen, as the model expects that Blackburn is based on fundamentals. With each new survey, it weighs less and less on fundamentals. That said, Bredesen is not out of the wood yet; this Indiana race took a long time to fall to Young in 2016 before he progressed on election day.
Finally, we come to Utah, where there is not much suspense: Romney is a huge favorite according to polls and fundamentals, and the only real question is whether he will be a reliable vote for Trump or a thorn in his side.
Source link
