Statistikhengst
Diamond Member
- Banned
- #1
This is somewhat a continuation of this thread from six months ago, but there is so much new data, it's just better to start fresh.
Of the nine largest states in the nation (California, Texas, New York, Florida, Illinois, Pennsylvania, Ohio, Georgia and Michigan), which currently account for 50.26% of our total population, there can be no doubt that the closest and most fought-out electoral battleground state is Florida.
Since 2013, now into the month of July, 2014, there have been 13 polls taken of Florida vis-a-vis the 2016 election, pitting Hillary Clinton (D) against a field of potential GOP candidates. Within those 13 polls, there have now been 46 individual one-on-one matchups, and Hillary has decisively won all 46 of those matchups.
Two days ago, on July 22, 2014, SUSA released a large poll from Florida.
Today, Quinnipiac released a large poll from Florida.
The results are mostly very similar to each other, and in one case, they are absolutely identical to each other.
The Quinnipiac poll can be compared to a complete poll from May 1, 2014.
The SUSA poll can be compared to two smaller SUSA polls, from July 7th and from June 10th, 2014, but only concerning two candidates.
So, I am going to present the latest of polls in two ways.
First, Quinnipiac alongside SUSA:
Quinnipiac, July 24, 2014:
Florida (FL) Poll - July 24, 2014 - Obama In Slump, But Clinton Sc | Quinnipiac University Connecticut
Survey group: 1,251 RV. MoE= +/-2.8
SUSA, July 22, 2014:
SurveyUSA Election Poll #21475
Survey group: 836 RV. MoE= +/-3.5
Qpiac: Hillary Clinton (D) 49 / Jeb Bush 42
SUSA: Hillary Clinton (D) 47 / Jeb Bush 41
Margin: Qpiac - Clinton +7 / SUSA Clinton +6
Margin disparity: Qpiac Clinton +1
Qpiac: Hillary Clinton (D) 51 / Paul Ryan 38
SUSA: no Clinton /Ryan matchup taken
Margin: Qpiac - Clinton +13 / SUSA N/A
Qpiac: Hillary Clinton (D) 53 / Mario Rubio 39
SUSA: Hillary Clinton (D) 53 / Mario Rubio 39
Margin: Qpiac - Clinton +14 / SUSA Clinton +14
Margin disparity: NULL (identical margins)
Qpiac: Hillary Clinton (D) 53 / Rand Paul 37
SUSA: Hillary Clinton (D) 46 / Rand Paul 42
Margin: Qpiac - Clinton +16 / SUSA Clinton +4
Margin disparity: Qpiac Clinton +12
Qpiac: Hillary Clinton (D) 54 / Chris Christie 33
SUSA: Hillary Clinton (D) 49 / Chris Christie 38
Margin: Qpiac - Clinton +21 / SUSA Clinton +11
Margin disparity: Qpiac Clinton +10
So, just comparing the two polls from independent-from-each-other pollsters, there are some similarities:
-in both polls, Jeb Bush comes the closest to Hillary Clinton and he is the only GOP candidate to come within single-digits, albeit high single-digits, of her.
-in both polls, Chris Christie fares the worse and after checking, 2014 is the first time in 26 years (the 1988 Bush vs. Dukakis election) that I have seen a Florida poll showing a candidate at +20 or more over another candidate. It's now happened twice in Florida, and twice for Hillary. Hillary Clinton's margin over Chris Christie in the Qpiac poll is an astounding +21, really quite unheard of in this state.
-the Hillary Clinton / Mario Rubio toplines and margins are absolutely identical to each other: 53/39, Clinton +14.
-in 4 of 5 Qpiac matchups, Hillary is over the 50% mark.
-in 2 of 4 SUSA matchups, Hillary is over the 50% mark.
Florida is usually a state where both candidates stay in the low to mid fourties until primary season, and then, usually one of the two gets closer to the 50 mark.
There are also disparities:
-again, vis-a-vis Chris Christie, Clinton is at +21 in the Qpiac poll, but at +11 (still a landslide margin) in the SUSA poll. That is a very, very large disparity, actually, too large. The average of those two polls would be Clinton +16, which, btw, was exactly the Clinton/Christie margin of the Qpiac poll from January 2014.
-against Rand Paul, Qpiac shows Clinton +16 (a very large landslide margin) but the SUSA poll shows Clinton +4 (almost within the MoE). This is an even larger disparity, and really should not be. The average would be: Clinton +10.
Let's dig deeper into the internals, into the female vote:
Qpiac: Clinton 56 / Bush 36, Clinton +20
Qpiac: Clinton 58 / Rubio 33, Clinton +25
Qpiac: Clinton 58 /Ryan 32, Clinton +26
Qpiac: Clinton 61 / Paul 29, Clinton +32
Qpiac: Clinton 61 / Paul 28, Clinton +33
So, according to Qpiac, Hillary Clinton is beating the GOP in the female vote by a minimum of +20 and a maxium of +33.
Historical context:
In 2008, nationally, then-candidate Obama (D) won the female vote by +13, 56/43:
National Exit Polls - Election Center 2008 - Elections & Politics from CNN.com
In Florida, Obama won the female vote by only +5, 52/47:
Local Exit Polls - Election Center 2008 - Elections & Politics from CNN.com
However, female voters were 53% of the Florida electorate in 2008.
Obama won the state in 2008 by +2.81%.
Fast forward to 2012. In 2012, President Obama won re-election nationally and took the female vote by +11, 55/44:
Presidential Race - 2012 Election Center - Elections & Politics from CNN.com
(So, he lost 2% on margin in the female voter over 2008).
In Florida, Obama actually improved on his statistic in the female vote, winning it by +7, 53/46
female voters were 55% of the Florida electorate in 2008.
(So, while Obama lost 2% on margin in the female vote nationally, he gained 2% on the female vote in Florida).
Obama won the state in 2012 by +0.88%.
Sidenote: the fact that the female vote spiked in 2012 over 2008 and that Obama improved his margin among women voters but his statewide margin shrank by 2 points over 2008 can only mean that he, while having improved upon the female vote in Florida, took a major hit in the male vote at the same time.
The point I am making here should be crystal-clear: if a Democrat can win the state of Florida in a close presidential election with only between +5 and +7 in the female vote (well under the national margin in the female vote), then it is easy to see why the race for a Democrat would win a large landslide-margin in any state if she is winning by +20 or more in the female vote. It's really simple math.
This is the major warning sign for the GOP going into 2016.
------------------------------------------------------------------------------------
Back to the Qpiac poll. Let's compare it to the one before it.
BOTH QPIACS:
Quinnipiac, July 24, 2014:
Florida (FL) Poll - July 24, 2014 - Obama In Slump, But Clinton Sc | Quinnipiac University Connecticut
Survey group: 1,251 RV. MoE= +/-2.8
-and-
Quinnipiac, May 1, 2014:
Florida (FL) Poll - May 1, 2014 - Jeb Bush Is Top Dog In Florida | Quinnipiac University Connecticut
Survey group: 1,413 RV. MoE= +/-2.6
In terms of poll group size and MoE, the two polls are evenly matched with each other.
(The older Qpiac values are in parentheses)
Hillary Clinton (D) 49 (49) / Jeb Bush 42 (41)
Margin: Clinton +7 (+8)
Margin difference: latest Qpiac Bush, J +1
Hillary Clinton (D) 51 (56) / Paul Ryan 38 (36)
Margin: Clinton +13 (+20)
Margin difference: latest Qpiac Ryan +7
Hillary Clinton (D) 53 (52) / Mario Rubio 39 (40)
Margin: Clinton +14 (+12)
Margin difference: latest Qpiac Clinton +2
Hillary Clinton (D) 53 (55) / Rand Paul 37 (37)
Margin: Clinton +16 (+18)
Margin difference: latest Qpiac Paul +2
Hillary Clinton (D) 54 (52) / Chris Christie 33 (34)
Margin: Clinton +21 (+18)
Margin difference: latest Qpiac Clinton +3
Now, here is where comparing gets fun, for actually, compared to May, 2014, Bush, Ryan and Paul have actually improved their statistic somewhat, but losing by a landslide is still losing by a landslide is still losing by a landslide. Meanwhile, Clinton improved her statistic over Christie and Rubio, but a shift of +2 or +3 means that only 1 to 1.5 percent of the polling group has shifted sides, which could just as likely be some statistical noise - so, both Clinton shifts, plus the Paul and Bush shift essentially mean very little. But the Ryan shift is noticeable - he closed the margin between himself and Clinton by 7 points, and that is worthy of attention.
Now, these are just two polls, and more are to come, but they essentially continue exactly what we have seen in the 11 polls before these two: that Hillary Clinton is handily winning in one of the 5 absolute key battlegrounds in the Union, and a state that is absolutely essential for a GOP win nationally. Face it: without Florida in it's column, the GOP will not win.
Here is the electoral bio I did of Florida at the end of 2011, going into the 2012 election (it will be updated at the end of 2015, looking into 2016):
Statistikhengst's ELECTORAL POLITICS - 2013 and beyond: Rank 27 / 25: Florida
There is a lot of data there that might really interest people, but bring a cup of coffee with you, there is a lot of data to sift through.
How did the pollsters do in Florida in 2012?
Statistikhengst's ELECTORAL POLITICS - 2013 and beyond: The moment of truth: how did the pollsters do?
And very specifically, in terms of Florida:
https://docs.google.com/spreadsheet...0U3aFBuT09zQ2xXQ29fTjlJRlE&usp=sharing#gid=18
There were 113 polls of Florida in 2012, the end polling average pointed to Romney +0.61 in the final week average of polls. Obama won by +0.88, so the difference is 1.49 points, well within any MoE.
The pollsters who nailed it in Florida in the last week, with Obama +1:
PPP (D), Grove (D), Quinnipiac
The two pollsters that were the most off:
Insider Advantage: Romney +5 / Mason-Dixon Romney +6. It is interesting to note that these pollsters hide their results behind a paywall and do not release any internals.
I am pointing this out because some people could say "bah, just a couple of polls showing Hillary ahead in Florida, big deal", and were it just one poll, I would agree. And indeed, in Florida, there was one pollster in 2012 who was so ridiculously off (Foster-McCollum), I doubt we will hear much from them again.
However, in counting the 46 matchups to date in Florida, in 22 of them, Hillary won with a double-digit margin, way outside of the MoE.
For further information, here the complete presidential electoral (statewide) data from Florida, in one excel chart:
http://uselectionatlas.org/RESULTS/compare.php?year=2012&fips=12&f=1&off=0&elect=0&type=state
------------------------------------------------------------------------------------------
FACIT: We have a long way to go until election day, but the polling data coming out of Florida (and Ohio, for that matter) should be of great concern for the GOP. Right now, it's Hillary's race to lose in Florida, and if these numbers hold, then the state would not even really be a battleground in 2016.
Last edited: