The AL-SEN Class 2 special election - polling, stats and history

The general election to elect a new Senator from the great State of Alabama to fill Jeff Sessions' old seat (Luther Strange was a temporary appointment, who then lost the primary election to Roy Moore) will be on December 12th, 2017, today in 19 days.

Here the current polling aggregate for the Senatorial race, according to RCP:



Currently at Jones (D) +0.8, this is a statistical dead-heat in a very, very deep red state where the Democratic nominee should be hopelessly behind.

You will note that the RNSC poll, which showed Jones at +12 over Moore, is not in the aggregate, and correctly so, in spite of the fact that it is within the time-frame, for the RNC poll is an INTERNAL poll and in most all cases, only PUBLIC polls are included in aggregates. I just wanted to clear that point up right away.

In the case of the FOX 10 (WBRC-TV)/Strategy Research poll currently showing Roy Moore at +2, the historical trend-line over the last 3 weeks has gone from Moore +11 to +6 to now +2. So, this latest FOX-10 poll (which is not to be confused with the FOX national poll: different pollster) is not good news for the Moore camp at all.

And on top of that, the most damning poll for the Moore campaign is the Gravis (R) poll, currently showing Jones +5. Gravis (R) is a decidedly partisan pollster just like PPP (D), but both of those pollsters use respected methodologies and their numbers have been, historically seen, not bad. In 2012, Gravis (R) was like a mini-Rasmussen, because they used Rasmussen's cheaper "Pulse" software. Since then, Gravis (R) has developed into a pretty respectable pollster on it's own and above all else, has put time and money into researching the South in order to know how to build reliable models, just as Quinnipiac has done for the NE, Acela and battleground states.

Let's put this polling in context of Alabama's electoral history (Senatorial, Gubernatorial, Presidential):


Source

Some interesting information to get from this table. Note: at Dave Leip's uselectionatlas website, red = D, blue = R.

Since 1996, 8 Sen-cycles, the Republicans have won all 8 and with the exception of Jeff Session's first election in 1996, all by big landslide margins, in 6 of 8 elections, by more than +25 (in 2014, Sessions ran unopposed). The last time we saw a "squeaker" election in this state was in the 1986 mid-term elections, where Richard Shelby, at that time a Democrat, just barely unseated Republican incumbent Jeremiah Denton, a one term incumbent who rode in on the Reagan national wave of 1980 and did slightly better in the state than Reagan himself. Pre-1980, this was a solid D-state at the Sentorial level, even in 1972, going against the Nixon (R) national re-election wave by 76 points. Now, this senatorial info only goes back to 1960 (with freak data from 1914 also in the table), but the only other election that was a squeaker was in the 1962 mid-terms (Kennedy administration), as talk of Civil Rights as really beginning to heat up.

So, in terms of electoral history, if the Moore/Jones 2017 matchup remains a dead-heat, then this election would be the first "squeaker" in Alabama in over 30 years.

On the Gubernatorial level:


Source

Since 1986, with the exception of 1998-2002, a Republican state. Note that in 1986, a mid-term election year when the voters of Alabama elected a Democrat to the Senate, they elected a Republican to the Governor's mansion for the first time in a long, long time (since 1872) in spite of a heavy 1986 D-wave, so, 30 years ago, Alabamans were more than willing to vote split-ticket.

Since 1946 (actually, since way before then) there have only been two squeaker gubernatorials in Alabama, both with Republican victors, in 1994 and 2002. In 1994, Bob James just beat out Democrat James Folsom (who also ran for Senate in the past). Interesting about that result is that it happened in the massive 1994 Republican wave that brought Newt Gingrich to the speakership of the House. This tells me that Alabama was far less partisan 30 and 20 years ago than it is now. In 2002, Republican Bob Riley just barely unseated Democratic incumbent Don Siegelman and that in a mid-term election cycle where the party in the White House (Bush 43, a Republican) did NOT take a hit as is usually the case (see: 2002, 1998, 1962). So again, in 2002, Alabamans were, according to their voting history, far less partisan, relative to the waves (or non-waves) of that time.

On the Presidential level:



Source (2 graphics, overlapping at 1936)

So, we see from the tables that Alabama, on the presidential level, has never really been a battleground state, except maybe in 1980, which was the leanest result (Reagan +1.30). It was also lean in 1928 and 1968 (like Pennsylvania 1960 and 2004 lean). Since 1984 it has been a very reliable R state. The somewhat leaner margins in 1992 and 1996 of course have to do with the Ross Perot (I) factor. Please note the Barry Goldwater (R) margin from 1964, a year in which Democratic incumbent President Lyndon B. Johnson won the highest NPV percentage of all time (61.04%), a margin only surpassed by Nixon in 1972 when we are talking about Republican victories in this state. But indeed, Trump's 2016 performance here was the best for a Republican since Nixon.

Alabama was one of only 3 states to go third party ("STATES RIGHTS") in both 1948 (Strom Thurmond) and 1968 (George Wallace, from Alabama), alongside Mississippi and Lousiana. Also, in 1960, despite Kennedy winning the state in a landslide, he only got 5 of 11 electors; the other 6 remained unpledged. This agreement was reached in the 1960 primaries because, you guessed it: dispute over Civil Rights. And of course, see the mention of Barry Goldwater 1964 above. The 3 presidential election cycles from 1960 until 1972 were turbulent years for Alabamans, I surmise, as Alabamans were politically reorienting themselves.

So, if it ends up being a very close senatorial race, then that would mean that Alabama would have moved between 26-27 points to the Left in relation to the 2016 Presidential and Senatorial results and also since the 2014 Gubernatorial.

That being said, in the current polling, both candidates are between 46-47%, which means about 6-7% undecided. Theoretically, those 6-7% could:

a.) immediately side with one party or the other, which could result in a comfortable win for either Moore or Jones, albeit not even remotely close to the usual margins in this ruby-red state. Or,

b.) those 6-7% could end up being write-in votes, or,

c.) those 6-7% could stay home.

In a nail-biter election, anything can happen. But this race wasn't supposed to be a nail-biter at all. It was supposed to be a sleeper election, a massive landslide for the R-nominee as usual. I wasn't expecting to have to write about this at all. But here we are, 19 days before a senatorial special election that could theoretically move the Senate from R52 to R51.

So, with that in mind, I will now be paying close attantion to this race up to and through election day.

Hope this information helped. Happy Thanksgiving.

-Stat

 

Alabama didn't move 25 points to the left. This one outlier race did due to the allegedly spotty past of the republican candidate. Alabama remains as red as it's always been. Nice try lol

That being said, if Moore is behind by 3% in polls on the election day, he will win by 4% or so. The stigma associated with him prevents people from admitting to pollsters they would vote for him, yet they will hold their noses and pull the lever for (R) in the voting booth.

 

Alabama hasn't moved anywhere yet, the election has not yet happened...

But the swing of a state is indeed measured from one election to the next election.

 

If you want to fool yourself and believe that it's anything other than a one time aberration caused by an unfortunate choice of a tainted candidate - by all means go ahead. I doubt anyone on the planet will buy the spin that the state of Alabama is an iota less red than it was on 11/8/2017

 

It's not a matter of belief. It's a matter of cold, hard, statistic. Either a person understands the difference between the two, or he does not.

And BTW, before the allegations of pedophilia came out, Moore was already between high-single digits and very low double-digits against Jones in a state where historically, it has been a GOP blowout pretty much every time. So, Moore was not tainted by someone - he was severely flawed from the get-go.

Education cures ignorance. Good luck to you.

 

That might be the GOP party line but the reality has changed considerably since you BLOTUS infested the Oval office.

Alabama Republicans are not exactly enthralled with the BLOTUS or Moore which depresses their motivation whereas Dems are motivated since they see an opportunity to send a message.

The motivation of the electorate is a huge factor in local elections and Moore has been an embarrassment to the state for a very long time. The opportunity exists to change the image of AL and the GOP voters are not beholden to vote in lockstep for purely partisan reasons.

 

The concept of a upset senate win in a deep red or deep blue state is not without precedent. Republican Scott Brown won Ted Kennedy's Senate seat in deep blue MA in 2010. Your historical insight above establishes that the voters of AL are not locked into voting for the GOP regardless of who is on the ballot.

The Dems in AL have been stirred up for almost a year now and they were discussing the potential for an upset victory in their primaries even before Moore became the R candidate. Jones is a solid Dem who is sticking with his message and not ashamed of his ties to the party. He makes a lot more sense as a candidate than Moore does.

It is worth noting that AL has a very strange form of state government. Apparently everything passed in the legislature automatically becomes part of the state constitution. For instance if the legislature were to pass a law require that government buildings must be painted dayglo purple then every government building would have to painted that color. That includes local municipal offices, libraries, fire stations, police, etc, etc. I know that is a little absurd but it has had a negative impact on the state as a whole because local jurisdictions can be constitutionally mandated to do things that not within their budgets and provide no benefit to the local communities.

I don't know if the above is relevant to this election but it does provide a little background into how the state operates.

Right now Moore is hanging in there and the hardcore base will vote for him regardless. What matters is how many voters will refuse to hold their noses for him. That and a motivated Dem base could be what it takes to swing this race.

 

Statistics understands the concepts of bad data (garbage in garbage out), aberrations, outliers. You are not a very good statistician if you don't.... Moore is undoubtedly "bad data" as he is uniquely unrepresentative of the political trends in the state. The election in Alabama will be a referendum on his behavior (alleged or actual) 40 years ago and nothing else. But keep trying to convince us and yourself that Alabama is turning blue, it's actually funny to watch.

CRO Statistics: How to Avoid Reporting Bad Data
https://moz.com/blog/cro-statistics-how-to-avoid-reporting-bad-data

 

Exclusive: Roy Moore rebounds in polls as accusations are discredited
Moore with 46.7 percent support to Jones’ 39.6 percent. 13.7 percent of voters polled remain undecided. With a margin of error of 3.1 percent, Moore maintains significant lead over Jones
https://altoday.com/archives/19835-...ounds-in-polls-as-accusations-are-discredited

Jones’s problem’: African American voters not energized by Alabama’s Senate race
https://www.washingtonpost.com/powe...d1e45a6de3d_story.html?utm_term=.664bfc529271

 

Single polls are meaningless, the average is what matters.

The rest of your article goes on to describe how the Jones campaign is addressing the issue of black voter turnout on 12/12.

 

Later polls have significantly more weight than the older ones. Two latest polls have Moore comfortably ahead.

In your previous post you baselessly claimed that the dems are uber-motivated.... As it turns out, it's an issue that needs to be addressed.

 

You only provided a single poll in your article and the prior poll that was mentioned in that article stated that Jones was leading.

The Dems are motivated to GOTV and you haven't proven that they are not. Your link just proved that they are motivated to address any and all potential ways that they can win.

 

Here is the second [latest] poll showing Moore ahead by 2 points
http://www.wbrc.com/story/36899832/...s-roy-moore-and-doug-jones-in-statistical-tie

Nice try shifting the goalposts. As my link shows, the dem electorate IS NOT motivated.

Jones’s problem’: African American voters not energized by Alabama’s Senate race
https://www.washingtonpost.com/powe...d1e45a6de3d_story.html?utm_term=.664bfc529271

PS Don't worry, it won't be close anyway

 

Excellent work on the OP, as usual, @Statistikhengst.

I, too, am watching this race fairly closely. I agree with @Derideo_Te that turnout is going to be critical. There seems to be an underground "repression" of votes by Republicans that say they will just stay home, rather than vote for either a Democrat or a pedophile. I personally know people down on the Alabama Gulf Coast that are working very hard on the Democratic GOTV effort. Will it be enough? I don't know.

Here's what I'm watching in this election…
1. Will the people of Alabama rationalize away child sexual molestation for political purposes? This race will be seen that way, if the race is close, no matter which candidate wins.

2. Will women actually vote for Moore? If Moore wins this election, the people of Alabama are telling every little girl or woman that has ever been sexually harassed or assaulted to shut up about it, because no one cares what actually happened to them, if it means losing a partisan Supreme Court Justice vote at some nebulous time in the future. When, nationally, 60% of women experience some sort of sexual harassment, I wouldn't bet on them voting for Moore. Will Republican women will cross over and vote for Jones or just not vote, if they choose not to vote for Moore? I think that may end up being the deciding factor in this election.

3. 89% of Alabamians qualify for health care premium supplements through Obamacare. Repeal of the mandate is in the Senate tax reform package. Will Kellyanne Conway's flipping from "no Senate seat is worth a child" to "we need Senate votes to pass tax reform" make a difference for a candidate that Trump didn't originally support? I'm not sure cutting health care sells in Alabama, especially for a tiny tax cut.

 

That poll is a statistical tie!

That is just a misleading clickbait headline but par for the course for those who support a predatory child molester.

 

Great points about ObamaCare premium supplements and women voters.

I was not aware that so many of the people of AL were earning so little as to eligible for those programs. We have already seen how upset Republican voters were about repealing Obamacare. If the GOP has managed to aggravate 89% of the voters with the repeal attempts that alone will harm their support.

Your point about women is crucial because they make up 51.6% of the population and are the largest voting bloc. The BLOTUS did not do as well amongst women in 2016 so if they either abstain or crossover that is going to a deciding factor.

 

Actual election results are not outliers nor are they bad data.
You do understand this very simple-to-grasp fact, right?

Now, one single election out of many can indeed be an outlier, but we can only classify it as such if successive elections prove it to be such. That however, would still not change the MATHEMATICS behind terms such as "margin", "swing" (which is the term to which I referred in the OP) or "trend". Those numbers are never outliers, but rather, very simple and proveable mathematical equations. In the real world, adults understand this material.

And finally, I never even once wrote that Alabama is turning blue. I DID write that even if the election is close to a tie, then that would represent a 26-27 point swing to the LEFT, which would be a true statement and very much based on accurate math.

Good luck to you on learning how math works.

 

Your statements above are contradictory but yes, an election can be an outlier, just like I said in my previous post

C-. Wrong, if a data point drastically differs from all available historical data, it IS by default classified as an aberration/outlier, unless/until future data confirms its validity. Even a high school student, let alone a professional or a scientist knows full well that if a data point is produced under extremely atypical unusual circumstances and deviates significantly from historical trends and averages, the chances are 1000 to 1 that this data point is an outlier caused by these unique circumstances and not by a change in the underlying conditions.

Good. So we can agree then that any claim or implication that Alabama is any less red than it was a year ago are baseless regardless the outcome of this outlier election. Right?

 

NO. Once again, just in case your brain is weak: the actual results of any election are, AT THAT TIME, not an outlier. They are simply current data, and for the congressional record, at that. First when they are compared to a like election that happens in the future can we ascertain whether it was an outlier, or not. So, until such a future election happens, none of us can claim that any election is an "outlier", because the sucessive election has not yet happened. It's really that simple.

Good luck working on your comprehension.

 

Now, that's just the thing. I am not trying to say that Alabama is turning blue and I have already challenged you once to point to any sentence of mine in the OP which either says or infers such. I wrote that if the election is even close, it will represent a large swing to the LEFT for this election, which would be factually absolutely correct. I am sorry if real numbers disturb your fantasy world, which you are more than entitled to have, but I will not let you lie about my writing.

I challenge you once again to show EXACTLY where I wrote that Alabama was "turning blue". Good luck with that one, you will need it.

Are there any rightist out there at all who do not spew bullcrap daily? Wow, this is really childish...

 

This is not how it works in science. If a result contradicts thousands of past results it IS considered an outlier even if this result is the latest. It needs to be researched further of course to confirm beyond any doubt that it is in fact an aberration but the default assumption is that it is, until proven otherwise. Don't take my word for it though, ask your professor or a mentor.... You have a mentor, teaching you the ropes of the trade, right? BWT This does not mean that the result is somehow invalid, of course it is but one result (caused by a particularly weak/damaged candidate) does not indicate a political trend change in the state (just like the election of republican governors in NJ and MA was not a sign of growing republican influence in those states).

The point is there is zero evidence that Alabama is turning even an iota bluer than it was a year ago and this election regardless of the outcome is hinged on the weakness of the republican candidate, not the shrinking popularity of the republican party. The party affiliation in the state and the support of republican causes in Alabama has not changed whatsoever.

PS why aren't you reporting on 2 latest polls showing Moore comfortably 5 and 6 points ahead?

 

The GOP is in a Lose-Lose scenario when it comes to Roy Moore.

If he wins in AL and becomes their next Senator then the GOP is tainted as the party embracing a predatory child molester and if he loses their Senate majority becomes tenuous at best.

There is no positive result for the GOP no matter whatever the final outcome.

 

Lol, your "we're winning by losing" approach certainly works. Keep counting your pyrrhic "moral" victories while we stick with good old fashioned majorities in all branches of the federal government.

 

We the People will decide if the GOP deserves to remain in power in 2018.

 

The bolded: no. That is a lie. Using your "logic", then Reagan's 2 point win in Alabama in 1980 over Jimmy Carter was an "outlier". But wait, ever since then, a Republican has won this state on the presidential level, so no, it wasn't an outlier. See how that works? You can try your best to troll this thread all you want, but facts are facts and adults can read past bullcrap.

As for polls, I comment on them at a time of MY choosing. I am not your secretary. Unlike others who seem to live here in PF, I have a busy life in the outside world.
But just for historical context, the final Emerson poll in VA picked the correct winner, but was off 6 points to the Right. So, yeah, feel free to tout that Emerson poll in alabama. LOL.

and finally, it's not your business to know who my mentors were or are. Nosy people bore me...