robtuck

The polls are usually pretty accurate at National level, the way they are designed to be used. The issue, especially with 2016, was the extrapolation to local level results with the Electoral College system - they essentially did not do the legwork to understand the regional variation that existed and missed the result, even though they were pretty much spot on (from memory) on the proportion of the vote. Trump's success was in being able to win a few States by small margins that had decent wedges of Electoral Votes, while also losing some others by excessive margins. So, he sneaks in a Pennsylvania but gets trounced in California, which he'd have lost as a Republican even if he was actually Obama. Getting that type of more regional voting intention data will clearly be more costly than just asking a sample of the total population.

scrubbedexpat099

I listened to the Polls and then saw the election and a lot of very surprised people. They were quite a way off and then spend a long time trying to work out why.

I can not imagine I am alone in both never being asked and if I was asked telling them what I would actually do.

It is a big country but only a few States matter,

Anian

People were so embarrassed about voting for Trump that they lied even in anonymous polls. It's not a thing they have ever had to take into account before, and the number of those who lied was different everywhere. Either that or people were easily swung just before they voted.

scrubbedexpat099

Not sure about embarrassed, there was a lot of hate and why go there?

robtuck

I know someone from work who voted for him after spending months telling me how awful he was. She apologised to me a few weeks after! Said her Dad would disown her if she had voted for Hilary, as her Sister had done and had indeed been cut off by her Dad. Madness.

The National vote went for Hilary and the % margin of error pretty much has most National polls about right. The process for getting data is pretty archaic though. A decent enough statistician should be able to model in the factors being outlined though. BARB, who do the TV audiences in the UK, use a very small sample of just over 1,000 people - an entire multi Billion pound industry in Advertisers hangs off of that data. The sample size seems low, but it's more than enough to enable them to extrapolate.

Steerpike

Since when were polls 'designed to be used' only for national level? The sole purpose of a poll in a presidential election is to predict who will win the presidency. Anyone remotely familiar with the process knows that each state's vote translates into EC votes, and that can be predicted quite easily. If you poll California and you get 55% of voters saying they'll vote Hillary, you know that the full set of EC votes from CA will go to Hillary - it's not a dark art that it is not understood or somehow unpredictable.

Apparently, people don't understand that their vote is private! I spoke to a really good friend who voted for a bunch of local 'old white Republican men' in the last election in AZ; when I challenged her on the topic, she said "but I know all these people ... I can't not vote for them!".

BEVS

robtuck

I know you can create polling to pick-up the variance, but it's more effort for them. All the polling done, that I have seen here, is pretty much at the National level when you watch the comedy trio of CNN, MSNBC & Fox "News"

Ingles

There are "predictions" from those who know much more than most off us , that the "arse" has a VG chance off falling out of the Stock Market , with the resultant problems !
Any comments ,as this if it happens will be reminiscent off the state Bush left behind him.

Steerpike

There were massive predictions back in 2016 that the stock market would collapse IF Trump were elected. I was very heavily invested in stocks, but I was convinced he would not win, so did not do anything about it. On the night of the election, when it became clear he was winning, my heart sank and I figured I'd be set back 5 years in my financial plan (ie, no retirement!). The markets even started to tank overnight. But when I woke up the next day, the markets were strong and have been so ever since. So I'm still heavily invested in the market, and I can't face pulling out due to the gains I've made. So if it tanks, I'll be very sad BUT - if that's what it takes to see Trump off, I'll put up with it.

So I suppose ... if it doesn't happen I'll be personally happy but sad for the country and the world; if it happens, I'll be personally devastated but happy for the country and the world!

But as for 'what will actually happen' - I suspect no one knows. I'm meeting with my financial adviser next week and I'll get her take on things. She typically advises against any attempts to predict the market.

scrubbedexpat099

The Dutch had Tulips

Now we have Tesla, Lyft and Uber etc

A market that is grossly over valued and a Bond market that is perhaps in an even weirder state.

State Pension Funds etc that need 8%? return to make any sense when 1 or 2 seems more logical, not a Trump or whoever situation.

Ingles

Just too illuminate the above post.
https://dailysoundandfury.com/the-poor-jobs-report-isnt-the-only-bad-economic-sign-the-bond-market-is-screaming-recession/?omhide=true&utm_source=newsletter-735&utm_medium=email&utm_campaign=tti

dc koop

But if Bernie were to get elected it would be 1929 revisited. Stockholders leaping from windows or selling candy bars on street corners

Ingles

This is interesting on many fronts
Not least off all ,Trump suggesting that a National Holiday be declared in his name !

Ingles

Not living USA ,I was wondering if any off you that do so, could confirm or deny that the photo enclosed show's a fair sample off Trumps "base"
Thank you for your indulgence
Ingles