A lot of sense November 14, 2016
Posted by WorldbyStorm in Uncategorized.trackback
This from The Broken Elbow on how the media in particular got it wrong is very persuasive to me.
Because they, the media were appalled by Trump’s behaviour, they mistakenly assumed his grassroots would be also, when instead Trump’s people just saw the liberal media up to their usual lying tricks.
And so, they got the biggest story, arguably, in modern American politics completely wrong.
Had they got off their bums and went to the Ohio’s and Pennsylvania’s of America they would have discovered ample evidence of the alienation and anger that has propelled Donald Trump into the White House.
That point about the perception of the media can be seen elsewhere. In the IT one Trump supporter essentially said that media attacks on Trump were disgraceful. A lot of soul-searching is necessary to understand how that perception came about and how it can be engaged with. And he notes one small fact from the polls…
One standout exception [to massively incorrect polls] was the Los Angeles Times/USC polling which consistently had Trump leading Clinton in the days leading up to November 8th, much to the derision of commentators. On Monday, for instance, the day before the election, the LAT/USC poll had Trump ahead by five points; with one exception every other poll on the general election recorded by RealClearPolitics that day had Clinton winning.
It would be interesting to discover what it was that the LAT/USC did that the other pollsters did not.
It would indeed. But TBE’s point about a basic disconnect, indeed an unwillingness to engage with those on the ground is also reflected in the approach of the Democrats. Of course many of us don’t expect much from them in the first place, but it is stunning how little sense they seemed to have as to how a significant constituency of their support base was walking away. More on this during the week.
The Cedar Lounge Revolution 
It’s probably trite to say this here, but the media in Ireland, the UK and the US has both mirrored and fed the break in the representative link. Just as people can read the options on a ballot sheet and decide that none represent them, their community or their interests, so can they flick through channels on Sky or ruffle through the papers at the Newsagents and come to the same conclusion.
LikeLiked by 1 person
I don’t think its trite at all, I’m coming to the conclusion it is a truth that has to be said again and again.
LikeLike
But the polling was right -or at least more accurate than the election result. HRC is currently well over 600,000 votes ahead. Getting close to about 0.5 %
LikeLike
Yes that’s a fair point. What a screw up the electoral college is.
LikeLike
Yes.
LikeLike
No it wasn’t right, in any predictive sense. Even if they had the popular vote right they should have invested in modelling of the electoral college. Which would presumably have meant more local polling and more expense.
LikeLike
Well yes. What struck me looking at some of them was the small sample 3,000 voters in an electorate of potentially over 200m
LikeLike
Yes that struck me too, even for some of the States themselves never mind the whole US.
Would through previous employment have a knowledge of advertising targeting and the various models and “mosaics” for Targeting groups. There are a lot more than 3,000 different “mosaics” (a grouping of users by income, education, location, ethnicity and other factors) in the US, so it’s hard to see how accurate a poll could be on that basis.
LikeLike
FiveThirtyEight have a series of articles where they claim they were right the whole time and that if was the fault of dumb readers who thought Hillary was going to win and who failed to take their punditry with a pinch of salt.
Ignoring that bit of wrangling, they do point to a systematic bias in state polls. There simply were much fewer polls at the state level to give us an accurate picture of voter intention in Pennsylvania or Wisconsin. Clinton was always lookking ropey in New Hampshire, they say.
LikeLike
” the LAT/USC poll had Trump ahead by five points; with one exception every other poll on the general election recorded by RealClearPolitics that day had Clinton winning.” above.
Trump ahead by five points nationally? But Trump lost nationally. Clinton won nationally by somewhere between one and two percentage points.
‘By the time all the ballots are counted, she seems likely to be ahead by more than 2 million votes and more than 1.5 percentage points,’
What the pundits,-apart from Michael Moore-failed to see was Clinton’s weakness in the mid-west especially Michigan, Pennsylvania and Wisconsin.
LikeLike
“Over all, the national polls missed the result by only a few points: Hillary Clinton is on track to win the popular vote by around 1.5 percentage points, not especially far from her roughly four-point lead in an average of national polls.
But the state polls were a different story. They systematically underestimated Donald J. Trump’s standing in the Upper Midwest and Northeast. His strength there was enough to make him the president…
LikeLike
Yes, the national polls weren’t wrong in so far as they caught her ahead but individual state polls underestimated Trump support in key states.
LikeLike
I noticed the LA Times outlier when I began looking at the polls – were they just luckily biased, or had they a better model of the election?
LikeLike
what the LA Times did was go back to the same 2,000 people online to survey them over the whole course of the campaign rather then keep surveying 3,000 random people by phone as most of the rest did, although they predicted Trump would win by more so they were nearly as wrong in percentage terms as others just picked the winner /RTE R1 This week
LikeLike
The polls were fine. We just need to learn how to read them. Generally speaking they claim a margin of error of 3%, with 95% accuracy. That means that, by their own account, they will be out by more than 3%, on 5 occasions out of 100.
That’s a significant element of doubt, and in a close election, its enough to be of great significance.
On top of this, we have to remember that one thing polls have never been good at is factoring the likelihood to vote. That was a big issue in the brexit vote, and I suspect when then umbers are all in, we will see it was a big issue here too. a small increase in turnout for Trump, and a corresponding decrease for Clinton was always going to be enough to swing things his way in key states.
That’s all in the numbers.
LikeLike
I’m not sure I’m so against the electoral college.
If we had had decades of elections won by racking up big votes along the narrow coastal corridors to the east and west, with those votes dominating the entire combined vote of the mid-west and south, I really doubt America could have held together as a federation.
How would we feel in Ireland if all EU decision making was based entirely on majority voting?
LikeLike
Isn’t the Senate supposed to give that geographical balance? Maybe the problem is the president is too powerful.
LikeLike
Doesn’t the senate provide the necessary safeguard?
LikeLike
You can certainly argue that it does, but similarly you could say the senate balances the house. The president basically appoints the gov’t, so I can see a good argument for creating a balance in that vote too. I think that was the original argument?
LikeLike
” In a direct election system, the South would have lost every time because a huge percentage of its population was slaves, and slaves couldn’t vote. But an Electoral College allows states to count slaves, albeit at a discount (the three-fifths clause), and that’s what gave the South the inside track in presidential elections. And thus it’s no surprise that eight of the first nine presidential races were won by a Virginian. (Virginia was the most populous state at the time, and had a massive slave population that boosted its electoral vote count.)”
http://www.vox.com/policy-and-politics/2016/11/12/13598316/donald-trump-electoral-college-slavery-akhil-reed-amar
LikeLike
I was re-reading Robert Dahl’s “How Democratic is the US Constitution” recently (prompted by last week’s events). The little we know about the decision on the electoral college points to it being a system designed by committee which pleased nobody. The initial proposal was to have the house of congress vote on who should be president. This was not adopted. Pretty much every other suggestion was defeated until they lumped upon allowing a state legislature to appoint a number of electors equal to the number of senators and representatives that the state has.
There have seemingly been over 700 attempts to pass legislation to reform the electoral college. They have all failed, mainly in the Senate. Less populous states have jealously guarded their influence on the presidential elections.
That being said, how the electoral college votes is a matter for individual states. One method of reform would be to get rid of the ‘winner take all’ system most states have and have electors vote in proportion to how their state voted. This would not require a constitutional amendment, but it would need more coordination between states.
LikeLike
there’s a move afoot to pass legislation through state legislatures that would insist that the electors sent to the college would vote for whoever won the popular vote nationwide, irrespective of who won the state. The aim is to get to a wedge of states totalling 270, which would ensure the election would go to the real winner. currently the campaign is at 165, ith some biggies to go. again, would solve it without an amendment to the constitution.
LikeLike
http://www.nationalpopularvote.com
http://www.nydailynews.com/news/politics/clinton-popular-vote-win-pushes-electoral-college-change-article-1.2870784
LikeLike
Investigative journalist Greg Palast poses another possibility re voter suppression:
http://www.gregpalast.com/election-stolen-heres/
LikeLike
must be difficult to do exit polls that capture early voting properly, they would have to have people polling there for a week in places?
LikeLike
Saw this graphic on Twitter last week, dunno how accurate it is, but if it is, it fairly nails why Clinton didn’t get in:
LikeLike
watch those non-zero chart axis https://twitter.com/christinacaci/status/797871671331782661
LikeLike