All Polls Are Wrong

william.scherk

1,023 views

The phrase "all polls are wrong" was a cool hinge-point of argument last year, as the Trump train rolled on ...

Yesterday a Democrat penned an interesting article at The Hill. It didn't say that "all polls are wrong," but that surveys of President Trump's popularity in the USA are flawed and in no way indicative. In other words ... Why the polls are still wrongHere's a few excerpts from the article:

Quote

The polls that failed to detect the full strength of President Trump on Election Day continue to underestimate the president’s support for the job he is doing, paying way too much attention to the Twitter wars and ignoring the public support for many of the actions is undertaking.

This can create some serious misjudgments by organizations like the NFL and some Republican senators, who find out later that they buck the president only to their own detriment. And nothing was more devastating to Democrats than believing the election was over when it wasn’t.

Polls show the president’s approval rating all over the lot. An Associated Press poll put it at 32 percent and suggests that only 24 percent see the country as going in the right direction. This strains credulity given what happened in the special elections against the Democrats.

[...]

The methodology of some of these polls is to poll “all adults” without any qualification as to citizenship or voting intent. A lot of the nonvoters dislike politics and all politicians, and these polls also include them along with undocumented immigrants who are not screened out. Another group of polls has Trump’s approval in the low 40s, and Harvard-Harris Poll, which eliminates all undecideds, has it at 45 percent, similar to Rasmussen.

Remember, Americans liked President Obama for his way with words and his calm leadership style. They just opposed many of his policies, so Obama’s numbers gave a false sense of approval. Trump is the mirror opposite. People are put on edge by his words while favoring a lot of the positions he is taking on issues.

When it comes to rank-and-file Republican voters, Trump is the undisputed leader of the Republican Party. No poll I’ve seen puts his support from Republicans at below 80 percent and we at Harvard-Harris have it at 84 percent, which is remarkable, given his knock-down-drag-out fight with some mainstream Republicans.

[...]

The failure to understand the 2016 election was in large measure not a failure of the final polls, many of which showed a close race, but a failure to understand the powerful storyline of Trump’s appeal with his respect for cops and the military, taking a more aggressive position against our enemies, and pushing for tax and health-care reform. His style is not what won him the presidency. It was, remarkably, his substance.

I, frankly, didn’t at the time see his rise in the Republican primary as realistic. I don’t believe he has advanced his coalition from Election Day, and rank-and-file Democratic opposition has hardened. But he hasn’t lost his support either, and taking on “The Swamp” only empowers him further.

It is by watching the underlying public sentiment of what he is doing, and not his methods, that you see how polling better watch out here, as reality versus research will again be tested, and reality always wins.

Mark Penn is co-director of the Harvard-Harris Poll and was a pollster for Bill Clinton during six years of his presidency.

The Penn article also received some pushback, in this instance from Philip Bump of the fey canoes Washington Post: Why is a former Clinton pollster writing iffy poll analysis that panders to Trump supporters?

Quote

It’s true both that Trump is broadly unpopular (as measured by multiple opinion polls) and that there are positions he espouses that are popular. We will cede that since, well, it’s sort of obvious. We will also cede the point that Republicans broadly think Trump is doing a good job, something that has been noted any number of times by The Washington Post and others.

As Penn makes that point, though, he goes out of his way to include weird cultural touchstones that seem clearly intended to appeal to that Trump base.

For example, this is how Penn dismisses polling that shows Trump as unpopular:

The methodology of some of these polls is to poll “all adults” without any qualification as to citizenship or voting intent. A lot of the nonvoters dislike politics and all politicians, and these polls also include them along with undocumented immigrants who are not screened out.

There are two implications here. The first is that somehow people who don’t vote don’t get to have an opinion on the president. The second is that the number of people who reject the president is swollen by undocumented immigrants. At most, undocumented immigrants represent 3 percent of the population, a group that’s less likely to speak English and almost certainly not registered to vote. To suggest that they make up a significant portion of the responses to a poll is disingenuous.

“The president gets 65 percent approval for hurricane response and 53 percent approval for the economy and fighting terrorism,” Penn writes as he fleshes out “a more complex picture” of opinions on Trump. “He gets his lowest marks for the way he is administering the government. And he is a divider when people want a uniter.”

This is deliberately playing down how low Trump’s numbers are

Here's a snapshot from the folks at 538:

allpolls.png



35 Comments


Recommended Comments



4 hours ago, william.scherk said:

All polls are wrong, I say.  Some polls are more wrong than others. Here, former 538 contributor Harry Enten proclaims a new Narrative:

Trump's favorite pollster was the least accurate in the midterms

William,

Why that must mean that the mainstream polls are practically always right, right?

I mean, Trump doesn't have a clue, right?

It's a wonder he even got to be President with what little he knows...

:evil:  :) 

Michael

Share this comment


Link to comment

All polls are wrong.  With 279 days left until Iowans troop off to the caucus meetings, the 'horse-race' reporting is already feverish. Does anyone really think present polling is indicative?  I doubt it, but the business of sounding public opinion is not going to run out of clients.

With that in mind, this is supposedly worrisome news for the GOP in November 2020 -- five hundred and fifty three days from now. SMH.

ABCNews/Washington Post poll ...

As in 2018, Health Care Ranks Among Trump’s 2020 Challenges

Quote

[...] Americans by a 17-point margin say [Trump's] handling of health care makes them more likely to oppose than to support him for a second term.

That result in the latest ABC News/Washington Post poll marks one of many challenges for Trump as he seeks re-election. The Mueller report and his immigration policies are substantial negatives as well. As reported Friday, his overall popularity is low; just 39 percent approve of his work in office. And 55 percent flatly rule out voting for him next year.

[...]

 

Share this comment


Link to comment

I am thoroughly tired of the horse-race/Groundhog Day reporting of public opinion -- as if the horses run the race freshly each day of soundings.  There are two hundred and sixty two days until the first minivan is loaded for the Iowa caucuses. Gah.

As you can imagine the hoopla-industry is touting this meaningless poll:

Fox News Poll: Biden booms, Bernie fades

Quote

Trump slumps in head-to-head matchups - Fox News: “Biden also performs best in hypothetical 2020 matchups. Among all registered voters, he leads Trump by 11 points (49-38 percent), up from a 7-point advantage in March. Biden’s is the only lead outside the margin of sampling error in the matchups tested -- and he is the only Democrat to push Trump’s support below 41 percent. Sanders tops Trump by 5 points (46-41 percent) and Warren is up by two (43-41 percent), while Harris ties Trump (41-41 percent) and Buttigieg trails him by one (40-41 percent). The president’s reelect number holds steady at 38 percent, while 54 percent would back someone else if the 2020 presidential election were today.  In December, it was 38-55 percent.”

 

Share this comment


Link to comment

All polls are wrong, of course, but some forecasts turn out to be right. This is from the lady who successfully called the 2018 House elections, Rachel Bitecofer, in a page of material published at the Judy Ford Wason Center for Public Policy:

Quote

[..] Before revealing what my model has to say about 2020, I note one very important point of methodology. To construct predicted two-party vote shares for the Democratic Party’s nominee in each state, I use the best turnout estimate available for each state in 2018 for the Democrats. This is important because it allows me to capture the turnout surge we also saw among Republicans in 2018. Although I predicted an enormous surge in turnout among Democrats and Democrat-leaning Independents, the size of the corresponding surge among Republicans surprised me somewhat. I predicted the surge of Democratic turnout via negative partisanship, activated by the tangible threat of living under a unified government controlled entirely by Donald Trump.

What I did not anticipate was that, at least among Republicans, a threat response can be artificially generated at a mass scale and at a time when a party’s voters should be placated. Despite controlling the White House and both chambers of Congress in 2018, turnout surged nearly as much among Republicans, leading to the highest overall midterm turnout rates we have seen since 1914. Overall turnout ended up at a whopping 50.4%, tempting many analysts afterward to conduct comparisons between 2016 and 2018, a presidential-to-midterm comparison that is usually apparently absurd. Trump and the RNC accomplished this by running a base-centric mobilization campaign focused largely on stoking fear of immigration; a strategy they will replicate for 2020 while adding socialism into the mix.

Because my 2020 model relies on the 2018 vote to estimate the 2020 vote, it is naturally designed to account for this unexpected bipartisan turnout surge. As such, my expectation is the 2020 model will be better than the 2018 model, which was built with Virginia’s one-sided Democratic turnout surge as a turnout guide.

So, with no further ado:

[this is a snapshot of the interactive map from the lady:

bitCofer_Electoral_Model_2020.png]

Barring a shock to the system, Democrats recapture the presidency. The leaking of the Trump campaign’s internal polling has somewhat softened the blow of this forecast, as that polling reaffirms what my model already knew: Trump’s 2016 path to the White House, which was the political equivalent of getting dealt a Royal Flush in poker, is probably not replicable in 2020 with an agitated Democratic electorate. And that is really bad news for Donald Trump because the Blue Wall of the Midwest was then, and is now, the ONLY viable path for Trump to win the White House. [...]

 

Spoiler
On 2/2/2019 at 2:50 AM, Michael Stuart Kelly said:

My God! She sounds like a parody of an academic elitist snob. She talks down her nose with an affected "learned" voice. And progressive as all hell.

I never thought I would hear the term "morality" and think "shallow person," but that's what she lit up in my mind.

Also, I normally don't bash mystics the way Ayn Rand does because I see many of them earnestly seeking to understand the world, their lives, and why we all have to die in the end--suffering so painfully and trying so hard in their search for meaning.

But if this video is any indication of the tenor of Marianne Williamson's campaign from here on out, a silly vanity show based on clichés and bromides while wagging an elitist finger at the audience to cash in on shame, Ayn Rand's term, "witch doctor," comes to mind.

shudder...

Marianne Williamson used to serve up some decent woo-woo.

Now it's ick...

Ick!

Edited by william.scherk

Share this comment


Link to comment

"All Polls Are Wrong, and Tim Pool Knows It," but why pimp for Rasmussen?

Donald Trump's Approval Just Broke A Record High, Why Do Polls Say He Will Lose 2020?

Good question from Tim, although entailed enough to become a complex question fallacy if not scraped to the bone. Why do irrelevant poll purveyors right now try to 'horse race' this or any week's abstract mix of fluid and  fixed opinions and intentions? When do whose polls matter how where, and why? May a hundred conceptually-firm answers bloom. Name and shame, then move on.

I watch Tim Pool for his evocation of a 'sensible middle,' on behalf of the "Radically-Classical-Liberal Moderately-Cynical Realist Socially-Liberal-Libertarian Urban Studies" bloc I run with.

Photo of Los Popularos (on the left) copyrighted to Alex Waterhouse-Hayword. Story here: Art Bergmann and Los Rádicos Popularos I'm the slithery-looking  guy.

Los+Popularos+combined.jpg

There's only four of us revolutionaries left alive, which kind of wrecks our Reunion plans.

Edited by william.scherk

Share this comment


Link to comment

The Canadian federal election is scheduled for October 21, 2019.  I believe that the CPC (Conservative Party of Canada) will be able to form a government. In Canada, election law constrains the 'official' length of the campaign -- which is proclaimed when Parliament is dissolved. The actual campaign will be between 36 and 50 days from when 'the writ is dropped.'

My relatively confident prediction of a CPC government is based upon the "All Polls Are Wrong" notion. 

For those who want to make their own confident guessing ... there are two equivalents up here of 538 and RCP, sites which aggregate polling and publish projections. The first is 338Canada and the second is via the CBC: Canada Poll Tracker.

Although the polling and projections suggest a Liberal party victory and a second term for Prime Minister Trudeau ... I do believe that the polling misses something. 

-- one complication is the relative rise of another national party (one that will name candidates in a majority of the 338 seats at stake). This is the Peoples Party of Canada/Parti Populaire. It is led by a former Conservative party member who lost the leadership of the CPC by a relatively small margin -- Maxime Bernier.

I don't know if Bernier's party will "Jill Stein" or "Ralph Nader" the total 'conservative' votes, but I have noted that hardcore PP supporters reject the very notion of 'stealing' Conservative votes, allowing the Liberals to squeak out a minority ...

I follow this intriguing Youtube personality who has gone all-in for Bernier's party, holding interviews with a variety of registered candidates. Here Elle is in conversation with the PP candidate for Toronto-Danforth:

 

Share this comment


Link to comment
On 9/2/2019 at 12:22 PM, william.scherk said:

The Canadian federal election is scheduled for October 21, 2019.  I believe that the CPC (Conservative Party of Canada) will be able to form a government. In Canada, election law constrains the 'official' length of the campaign -- which is proclaimed when Parliament is dissolved. The actual campaign will be between 36 and 50 days from when 'the writ is dropped.'

Writ drop!  Elle can relax somewhat, as she foresaw Trudeau ignoring the law and possibly being subject to removal by extra-parliamentary means. A bit zany, but hey. Politics always has a circus element.

For Canadian OL lurkers and members, the photo-link below of Elections Canada (Elections.ca) gives you everything you need to know about the election. Who what when where how ... and provides the means to register if you are not already registered. In my riding, the Conservative incumbent will be returned to office in the new Parliament. I don't trust the 'horse race' polls** at the moment, of course, but with only 41 days in this campaign, the race is relatively short, and provincial-level polls will reflect at least the grossest indications. I still find it most likely -- going by 'feel' -- that the Conservatives will form government in the new parliament.

electionsCanadaIndexScrape.png

** -- national polls are usually sectioned out by region, but with larger margins of error than provincial/regional polling. It takes a kind of weighting informed by individual riding contests to interpret raw percentages in these surveys. The conventional wisdom is that the CPC and LPC are tied ... the slogans are kind of sad ... though tastes will differ. At the heart of my uncertainty is the uptick in sentiment about the future ...  "There is a malaise within the land ..."   [link]

Spoiler

Already, the parties have been jockeying for traction with voters as their unveil their campaign slogans over recent weeks.

The Liberals picked “choose forward “as their re-election pitch to Canadians, while the Conservatives are running with “it’s time for you to get ahead.”

The NDP will be using “in it for you” as their slogan.

The Green Party picked “forward together.”

Their leader, Elizabeth May, will kick off the campaign launch on Wednesday from Victoria, B.C.

Conservative Leader Andrew Scheer is also expected to launch his campaign following the official announcement on Wednesday.

NDP Leader Jagmeet Singh launched his campaign over the weekend.

The campaign comes as both the Liberals and Conservatives are effectively tied in the polls, which means the differences between the two parties fall within the statistical margin of error.

It also comes amid indications there is a growing sense of voter unease and concern about the future of the country, according to an Ipsos poll which suggested populist attitudes are gaining ground in Canada.

Fifty-two per cent of respondents to that poll said they feel that Canadian society is “broken.” That’s up 15 per cent from three years ago.

Another 67 per cent said they feel the economy is rigged to support the most powerful.

 

Edited by william.scherk

Share this comment


Link to comment

Guest
Add a comment...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...