Talk Elections

http://www.npr.org/2017/08/24/545812242/1-in-10-sanders-primary-voters-ended-up-supporting-trump-survey-finds (http://www.npr.org/2017/08/24/545812242/1-in-10-sanders-primary-voters-ended-up-supporting-trump-survey-finds)

Also of note, 34% of Kasich voters went to Clinton.

That's like ... a really, really small percent, LOL.  I mean, the Republican and Democratic nominees usually pull around that number of the opposite party's voters, no?

Interesting. Sanders had that same outsider appeal as trump.

If I'm not mistaken Trump and Sanders did better among Independents (granted this group is usually just as polarized as Democrats/Republicans) and lower income voters in their respective primaries. So I can see where the crossover came from.

Yeah, most of these were not even "real" Democrats, they were independents or "ancestral Democrats".

That's not that bad when you consider a lot of these Sanders supporters were populists who have never really turned out for the Democrats. Even if Hillary were a squeaky clean candidate they probably wouldn't have voted Dem.

Still though, these people are HPs.

Let's not pretend Sanders voters were Republican leaners.

()()

Pretty sure most of these voters were registered Democrats who have been voting for Republicans since the 2000 election. It's an open question though whether they would've turned out for Hillary Clinton had she been the D nominee in 2008. There were polls showing her winning WV and KY in the general against McCain.

And those two states' totals make up what percent of Sanders' voters, smartass?

Also, most Democratic primary voters in Kentucky and West Virginia only voted for Sanders out of sheer disgust with Clinton's anti-coal rhetoric. which is somewhat ironic, since Sanders is more left-wing on coal than Clinton was, but he was smart enough not to really talk about it like Clinton did.

They certainly made up a disproportionately large percentage of Sanders -> Trump voters, which is who we're talking about here.

When you alienate your base this is bound to happen. Most Sanders voters stayed home however.

I'm dubious about the whole survey. The data is online. The survey was conducted by an academic consortium. The survey was divided up into 50 segments. There were 120 common questions, and 120 questions that a researcher could ask that were specific to their research. They could also choose which persons were in their segment. So a researcher might pay for a 1200-person national sample of old people or millennials, etc.

But the survey was conducted on line using Yougov. Respondents were self-selecting. They had to use the Internet and volunteer to be surveyed. Even if you do demographic matching so you get the right percentage of various demographic groups, you still have that selection bias.

Based on the raw numbers it appears that President Clinton was elected by a 48% to 41% margin, with 11% voting for someone else.

Also, lets say that 50% of respondents were Democrats, and 25% supported Sanders, what percentage actually voted in a primary? 10% of the total (i.e. 40% primary turnout). So that would mean that 1% of the total were Trump-Sanders supporters. But relative sampling error goes up as the number of respondents goes down.

The WV/KY number is stupid. Sanders had 13M primary voters. WV/KY made a very small fraction of it & would little influence on the overall number. The sample size of WV/KY voters will be quite small.

See here is the deal. What % of Hillary voters voted for McCain in 2008? According to most polls, McCain got 10%+ of the Democratic votes in 2008, a number higher than what Trump got. Most of them were Hillary primary voters (rather than Obama voters).

Hillary to McCain was likely a higher number than the 10%. 10% is a pretty low number & is also testament to the effort of Bernie as well. And stop the with the KY/WV BS.

I thought it would be worse, and it probably would be if Bernie-backers who stayed home were counted in this survey.  So many Republicans who weren't happy about Trump held their noses and voted for him anyway, but too many Bernie bros couldn't do the same for Clinton.

There is something that Democrats need to learn from Republicans: Party unity.

34% of Kasich voters apparently went to Hillary, try again.

Yes, and both sides continuing to blame each other doesn't help.
Also, the popular vote doesn't matter. Clinton won it, but did not win a majority.
It is possible, but rare, to win a majority of the popular vote and still lose the electoral college.
People living in safe states shouldn't feel compelled to vote for the lesser of two evils.
This is why we need something like IRV.

As far as the GOP goes, the anti-Trump Republicans blew it, honestly. They should have united early, in favor of one candidate.

So?  What percentage of the Republican electorate consisted of Kasich voters?

Actually,
I  believe that it was the independents that decided the election and many of those who voted in open primaries did vote for Sanders.
Some voted for Kasich, but not as many as those who voted for Sanders. At least that would be my guess.

That is true. Of the small Bernie/Trump voters, most of them weren't even Democrats. Most young people are also Independents.

So this party unity is BS because they were never part of the party !

The anti-Trump Republicans were split between the moderates backing Kasich/Jeb etc. and the hardline movement conservatives backing Cruz. That's how Trump won: he happened upon a directionless, deeply split party and as an outsider was able to play the unifying candidate.

The only candidate who could have stopped Trump was Cruz, and like hell were the "establishment" going to throw their weight behind someone they hated even more.

Walker did the right thing by dropping out early (he was one of my favorites, but in terms of unity, it was right).  The GOP candidates didn't do anything to attack Trump in his weak moments, and there were A LOT of weak moments.

Hahaha, what? They were always on the attack, rather indiscriminately. But as Trump pointed out, everyone that attacked him ended up going down in flames right after the fact. Arguably Kasich got as far as he did because he didn't attack much, and Cruz waited until the end.

Hillary was lucky enough to pull off the Popular Vote.

Well, you're never going to convince the "progressive" hacks that holding 90% of a primary candidate's base in a general election is perfectly normal and actually quite impressive based on recent historical observations.

The KY/WV discussion has merit and actually should be expanded to include OK, as well as portions of other states like NC where Bernie signficantly overperformed demographics due to closed primaries/registration.

Bernie got 500k votes out of KY/WV/OK, about 4% of his primary total; it'd be safe to say that 200k votes in those three states were votes neither he nor Clinton would have got in the general, based on how those states voted both in 2016 and in previous elections. Throw in what we'd expect out of NC and that basically brings you to 2% of Bernie's total primary vote being unobtainable in the general from just those 4 states.

Another area I would look at - but that hasn't been talked about too much in terms of the effect of closed primaries augmenting Bernie vote - is the Northeast. States like NY have systems that work in both directions: it wasn't just low-information independents who voted to vote for Bernie but couldn't because they didn't register in time, but also Republicans registered as Democrats in heavily-Democratic states who did so in the first place to have influence in local and statewide primaries.

Anyway, it wouldn't surprise me at all if after everything was said and done, half of the Sanders-Trump voters nationally were long-time Republican presidential voters who could never have been won.

So, I have been trying to wrap my head around this one since it many ways although it doesn't make sense intuitively, there are definitely overlaps in terms of many rural counties throughout the US that heavily backed Bernie in the Dem primaries, and also areas where there were significant swings towards Trump.

Granted, in many cases Trump appears to have captured very few of Obama '12 voters, and in the vast majority of situations the defection to 3rd Party Candidates among Obama '12 voters appears to be a much more significant factor than Obama '12 / Trump '16 voters.

Still, having a good friend of some 15 years who was a Nader/Kerry/Obama/Romney/Trump GE Pres voter, who changed his registration from Independent to Democrat in order to vote for Bernie in the '16 primaries, this study and overall discussion is actually of significant interest.

So, I started running a few numbers and generated a spreadsheet to explore the topic, set any assumptions and personal opinions at the door to see if the actual raw data supports the claim of a small number of Ivory Tower Political Scientists that made a dramatic claim.

Ok.... here are the numbers for every single 2016 Democratic Primary by State.

I omitted the Washington State beauty pageant Primary, although honestly that likely if anything would have decreased the potential % of Sanders>Trump crossover voters.

I used a more conservative 10% number rather than the 12% number posted in the study cited above, since generally when analyzing data it is best to under-state rather than overstate...

()

Ok--- so we are reverse engineering the study using actual election results versus their polling methodology... So all caucus states are omitted, since these are not Primary States, and additionally they don't represent the will of the voters, considering the extremely small number of Party activists that participate in these events.

Let's start with the top-line number....

Based upon my calculations, Bernie bagged 11.668k voters Nationally in Primary States.

Let's look at the Raw Vote total of Bernie Primary voters by State...

()

Let's look at the data from another perspective to see the distribution of Bernie Primary voters,  so we can understand exactly where his core National distribution of support was by State...

()

Ok--- so what really stands out here is the concentration of Bernie supporters in the "Industrial States" of the Upper Midwest. Almost 25% of Bernie Primary voters were located in five Primary States (WI/MI/IL/IN/OH). 18% were in Western/Mountain West States heavily dominated by CA and OR (15% of National Bernie Vote Share by Region). 9.2% New England. 18.4% NY/Central Atlantic. 28.5% States of the Old Confederacy (Including WV).

So logically, if Trump were to have captured 10% of people that voted for Bernie in the Primaries, he would have received 1.17k voters distributed throughout the United States from Bernie '16 > Trump '16 voters.

Let's do a peanut butter spread and assume that this 10% of Bernie voters were evenly distributed across every single '16 Dem Primary State.

I know what they say about assumptions, but hell how can we test the theory without first isolating variations outside of the norm from the data-set?

()

So what stands out here?

Obviously we have states like California, Illinois, Oregon, and Massachusetts where Clinton actually gained ground on Trump, despite Democratic defection to 3rd Party candidates.

If Trump had gained 10% of Bernie voters in the GE, most likely he captured less than the numbers above, meaning the votes would need to be made up elsewhere to meet the 10% number cited in the Study in order to hit the 1.7 Million Bernie/Trump voters allegedly out there...

Where would these places exist?

So next step is to look at the overall Bernie share of the vote by State. Meaning we take his 2016 GE primary numbers and simply divide it by the total votes by State for US Pres in the GE.

()

()

So basically it looks like Bernie Sanders primary voters accounted for an average of 10% of the 2016 GE voters!!!!

So, if we look at the % of Bernie GE voters by State we see the following:

1.) Vermont-               36.8%
2.) New Hampshire-    20.4%
3.) Wisconsin-            19.1%
4.) Massachusetts-      17.6%
5.) Illinois-                  17.6%
6.) West Virginia          17.2%
7.) Oregon-                 16.0%
8.) Rhode Island-         14.4%
9.) New Mexico-           13.0%
10.) Montana               12.6%

Runner ups would be Michigan, Indiana, Oklahoma, Pennsylvania, and Kentucky to round out the #11-15 slots, although MO/CA/MD/NY were above the average Bernie vote as a % of the Ntl vote for Dem primary states...

Wow--- what a weird Icelandic smorgasbord of States....

So obviously we see a high concentration of Bernie primary votes in high turnout states with more accessible ballot access laws that translates into raw numbers come GE time, compared with States where ballot access is much more difficult...

Let's look at it from another angle.... to what extent was the Clinton coalition dependent upon Bernie Sanders Primary voters?

Simple math we take the total number of Bernie votes by State in the '16 Primaries and then divide it by the '16 HRC GE numbers....

()

This is pretty fascinating, since it helps isolate the variables and look at the overall impact of the Sanders movement compared to the actual results of the Dem GE candidate....

Top Ten States in terms of Sanders Primary / Clinton Vote Share:

1.) West Virginia-     65.6% Sanders
2.) Vermont-            64.9% Sanders
3.) New Hampshire-  43.5% Sanders
4.) Oklahoma-          41.4% Sanders
5.) Wisconsin-          41.1% Sanders
6.) Montana-            35.5% Sanders
7.) Kentucky-           33.5% Sanders
8.) Indiana-             32.5% Sanders
9.) Oregon-              32.0% Sanders
10.) Illinois-             31.8% Sanders
11.) Massachusetts-  29.4% Sanders
12.) Missourri-          28.9% Sanders
13.) New Mexico-      27.0% Sanders
14.) Rhode Island      26.4% Sanders
15.) Michigan-           26.2% Sanders

So what does this show us? It helps isolate the Bernie Primary vote by State to see where there might have been the highest levels of Sanders > Trump crossover voters....

Definitely if we look at some of the States on the list it is crystal clear that there were a significant number of Sanders/Trump voters with WV, NH, OK, WI, MT, KY, IN jumping out especially when looking at the '12>'16 Obama swings.

Still I am skeptical on the 10-12% claim at a national level, since simply when we start adding up the numbers, I am hard pressed to come up with the math of where exactly the 1.2 Million voters (10% number) that voted for Bernie in the Dem Primaries and then Trump in the GE came from.

Quite frankly, there is a strong correlation between 3rd Party voting and Bernie support in many places than Bernie>Trump voters.

My next thought is to focus on a few key states where it is most likely that there may have been a significant concentration of Bernie/Trump voters such as WI/MI/NH/MO and possibly PA.

Thoughts???

Pretty cool. I am not sure how accurate it is but I like the maps & effort that you put in.

Pretty cool. I am not sure how accurate it is but I like the maps & effort that you put in.
[/quote]

Well, the numbers are accurate, the math is accurate, now the key question is if the analysis and maps represent the reality of the Sanders>Trump potential cross-over votes.

I am extremely skeptical on the 10% argument, let alone the 12% argument, since the overall math doesn't appear to support that hypothesis.

Additionally, we haven't even controlled for Closed Primary/ Open Primary, etc variables....

President Griffin referenced it regarding the WV/KY/OK discussion, and then added in areas like upstate New York to the topic.

What really stands out to me after running the numbers is Bernie's total GE vote share in Wisconsin. I would also argue that '16 GE results in NH & WI seem to have a strong correlation between Bernie Primary> Trump GE voters, not to even mention various other States and counties within States that weren't virtual toss-ups.

I strongly suspect that if we were to look at State level maps by County we would potentially see a similar phenomenon.

The Main Stream Media (MSM) to use a Fox News and 'Pub cliche angle was that Bernie's appeal was predominately based upon a ton of naive Millennial "kids" in College Towns. The story that was oft forgotten was that many of us (Including some of my Republican adult WWC friends) voted for Bernie in the Primary precisely because he was a different type of Democrat running for President (Although they might have voted for Obama 1x or 2x).

Where were some of the biggest Obama '12> Trump '16 swings by State?

They were in precisely the same places where Bernie captured the highest % of the Clinton GE Vote Share....

Fine, let's ignore WV and OK and claim that these Democratic Primary voters weren't "Real Democrats", but just registered to vote Dem despite their Republican Pres voting patterns (A bit of a patronizing argument IMHO)....

How do we explain NH and WI? We can even drop down into Montana where 36% of HRCs vote share were Bernie Primary Voters...

Then we get to KY, IN, and OR....

Hmmm.... Maybe Bernie played well among Ancestral WWC rural and Small-Town Dem voters that voted Obama in '08 and/or '12?

Did the Democratic Party candidate in 2016 appear to not fight for Ancestral Rural and Small-Town WWC Dems? Sure as hell looks like that was the case to me....

iirc more Clintonites defected to McCain in '08, more Obama-Clinton voters went Trump as opposed to Berners who defected. Most who did defect (and some who didn't; among Berners) were Republicans/Indy before the primaries.

Sure, I believe you are likely correct on these fundamental points, so thanks for keeping us in perspective. :)

Additionally, I'm not actually convinced at this point that 10-12% of Bernie Dem Primary voters supported Trump in the GE, which didn't intuitively make sense, especially when looking at where the largest concentrations of Bernie Primary voters resided.

Did 88-90% of Bernie Sanders Primary voters support Clinton?

Absolutely not.... in fact, I might suggest (Although it will take a bit of additional time and effort to examine the numbers) that defections among Bernie '16 Primary voters were higher to 3rd Party Candidates (Depending on ballot access in the '16 GE by State) than Bernie/Trump cross-over votes.

Sure we can point to WV/KY/OK, and there appears to be a strong correlation of Bernie>Trump cross-overs. Still these three States account for only 4.4% of Bernie '16 Primary Voters.

Now, if we look at some of the other states that swung heavily towards Trump where there was a significant Bernie "GE Vote Share" and "Bernie Primary / Clinton GE Raw Votes" in two places NH and WI.

These were both extremely close states in the '16 GE, and most likely Trump did capture somewhere around 10% of the Bernie Primary Vote share, so in my mind would be Ground Zero to look at County level numbers...

There's definitely a strong element of truth with the sexist racist white male berniebro voter. All who voted Sanders in the primary and didn't vote Clinton in the GE qualify as such. Throwing minorities, women, LGBT, and Muslims under the bus isn't sending any message to the system other than you're okay with bigotry.

The Bernie/Trump votes which is likely much less than Hillary/McCain votes is probably not 10%. It could be higher than 10% in states like WI or MI but not on the whole. You are right there were a lot of Sanders/Stein & Sanders/Johnson voters although both well hugely after Sanders campaigned hard for Clinton. Some of the Sanders/Johnson voters were libertarians who voted Sanders because there was no strong libertarian candidate after Super Tuesday.

WV/KY/OK is a small part, 3-4-5% of the overall numbers. Even if it was 25/30% here, the small size would mean it would not have a major impact.

I think what makes it difficult to estimate is the turnout numbers. Primary turnouts are a lot less & it varies from state to state. One of the reasons could also be some primary voters staying home/not voting at the Presidential level. That can also swing states as the pie falls & Trump's core voters are very enthused.



@ Ignoring the David Brock Super pac paid troll.

quite a few Hillary primary voters in 08 voted for McCain in the general also: https://www.cbsnews.com/news/who-were-those-clinton-mccain-crossover-voters/

Techno Timmy--- honestly this sounds like you're going "on tilt" a little bit here with some bad primary blood in your system.

My question, is what does the actual data show to support or challenge this hypothesis, and not picking old scabs from the '16 Dem Primary?

It seems like there are a few states (OK/WV) where Trump definitely captured >10% of Bernie Primary Voters....

Additionally we have a few other States where it is possible that Trump captured 10% of Bernie Primary Voters (WI, NH, MT, and possibly PA deserve more detailed County level analysis).

Again, by using the whole "BernieBro" stereotype you appear to be ignoring the reality that Sanders Primary > Trump GE defections most likely occurred among Middle-Aged Rural and Small-Town WWC voters that shifted Democrat under Obama, and didn't feel that the economic gains of "The Great Recovery" were being equally shared....

I'm sorry if you feel the need to impose party purity upon many voters whom fundamentally agree with major aspects of Democratic Party "Liberal" economic policy.

You can shoot the messenger if you want, but let's face it regardless of Bernie Sanders, Trump would have won many of these swing voters back mainly because of his campaign message was essentially economically protectionist and anti-interventionist.

This is exactly why he was able to win a plurality of Republican Party primary voters.

HRC would have faced the same assault that Trump's 'Pub Primary opponents did, regardless of Bernie Sanders.

Since you now have an Ohio avatar, do you want to chime in on the topic of Bernie '16 > Trump '16 voters in Ohio doing your Technocratic deal and analyzing county level returns?

Ohio is obviously the trickiest of the bunch because of the whole Primary system and cross-over votes there, so arguably is the hardest State to analyze on the topic at hand....

"You are right there were a lot of Sanders/Stein & Sanders/Johnson voters although both fell hugely after Sanders campaigned hard for Clinton"

Italics bold and underline for your key point regarding Sanders and 3rd Party Voters....

This is a key element.... 

If we look at various precincts in Oregon for example, there doesn't appear to have have been a major increase in Republican % of the Total Vote....

In fact, many of the areas that some some of the largest swings were places where there was a significant defection of Democratic Party Votes towards 3rd Party candidates....

The obvious areas are the most Democratic Cities in the Oregon (Portland, Ashland, Corvallis & Eugene).

Still, even in many traditionally WWC Dem strongholds in Oregon, we don't have a huge amount of data that supports the theory of a significant increase in the Republican Total Vote Share, as part of a movement towards Trump.

Now, why should one state like Oregon be especially relevant on this topic?

For starters, regardless of the stereotypes of "Portlandia", Oregon actually has one of the highest percentages of WWC voters in the Western United States....

Additionally, there is a much higher % of jobs directly tied to manufacturing, especially if we isolate Metro Portland.

In many ways rural and small-town Wisconsin and ME-02 share many more similarities than differences.

The rust belt, Auto, Steel, and precision tooling areas share a slightly different historical narrative that stretches from the Sit-Down strikes and occupations of the Auto Plants during the New Deal, in places like Flint, Michigan and Toledo, Ohio...

We could also go into the brutal strikes in the Tire Factories of Akron, Ohio and the Steel Mills of Youngstown, Ohio and Pittsburgh/ SW PA.

Although Ronald Reagan stuck and giant dagger into the back of what was once a formidable Industrial Union New Deal Democratic existence, and decimated communities across America and created wastelands where there were once thriving prosperous Middle-Class Manufacturing communities, the residual collective memories still reside, even as kids and grandkids move down the road seeking better employment opportunities elsewhere.

Democratic and Republican leading political figures alike in recent decades are both widely seen as complicit in the auctioning off of America to the highest Corporate bidder.

It is no wonder that both Trump and Sanders emerged as the "Dark Horse" candidates of '16.

Still, it does appear that 3rd Party voting was highest among Millennials (Trump's worst demographic) and lowest among 65+ (Trump's best demographic).

The 65+ likely accounts for a decent chunk of the 65+ Obama '12/ Trump '16 Swing alone.... (But lowest level of Sanders '16 support)

Millennial voters likely accounts for a huge chunk of the Sanders > 3rd Party Vote.

So really if we really saw 10% of Sanders Primary voters support Trump in the GE, then logically there must have been an insane swing among the 36-54 Yr Democratic Primary demographic....

As someone said to me some 15 years back "If it doesn't pass the sniff test...."

Hope it was worth it.

Many of the various posts on this thread remind me precisely why I rarely post on the 2016 US Presidential Election Board anymore...

It has long become the last ditch refuge and eminent of bottom feeder Catfish and partisan political hacks of all varieties, that I am dismayed that I actually spent some time trying to crunch the actual data and expected different results...

If there are any adults in the room that actually want to discuss real election results, I might pop back in, but at this point I have no time nor interest in petty squabbling of Twitter style posts of one or two barely intelligible sentences, that aim to solely accomplish something (?) in 45 characters, rather than actually spend the time and effort to examine and analyze actual data, and discuss it in a meaningful manner, that actually adds to the body of knowledge and doesn't subtract from it....

What should I expect on certain boards in Atlas, when after all we now have a "Twitter in Chief" as President of our great nation....

Gone for awhile now on the 2016 US Pres Election Board, and putting my efforts into another Atlas board where folks of all political backgrounds actually do a pretty decent job of discussing and analyzing election data....

1 in 10 is believable, perhaps low. At least in Metro Detroit, most communities that Sanders won in the primary voted for Trump in the GE, while most communities that Clinton won in the primary voted for Clinton in the GE.

34% of Kasich voters voting for Clinton doesn't surprise me either, since Kasich was for many an anti-Trump protest vote.